| Contents |

Pesticides Research no. 58, 2002

Effects of reduced pesticide use on flora and fauna in agricultural fields

Contents

Preface and Acknowledgements

In 1995 the Danish Ministry of Environment and Energy called for projects, which should elucidate possible effects on flora and fauna of reducing pesticide dosages experimentally at large scale. The present project was successful in the tendering.

The project was first designed in 1995 and further elaborated after a pilot phase in 1996 by the partners below who also executed the practical part of the project. It included the three main areas, botany, entomology (arthropods), and ornithology and in a addition supporting disciplines such as yield estimation, economic aspects and statistics. The project was initiated with a pilot year on one farm in 1996 and had its main period of practical work on five farms 1997-98-99 The finalization of sampling and counting took place in late summer and autumn 1999, ending with assessments of accumulated weed problems and occurrence of birds on stubble fields. The treatment of data and statistical analyses were finalized in 2001.

The Partners and assisting partners of the project were:

Botanical aspects: Dept. of Physiological Ecology, Botanical Institute, University of Copenhagen (Anne-Mette M. Jensen and Ib Johnsen)

Entomological aspects: Zoology Section, Dept. of Ecology, Royal Veterinary and Agricultural University (Peter Esbjerg and Søren Navntoft)

Ornithological aspects: Ornis Consult A/S (Bo Svenning Petersen)

Yield effects: The Danish Agricultural Advisory Centre (Hans Kristensen and Poul Henning Petersen)

Economic aspects: Economy Section, Dept. of Economy, Forestry and Landscape, Royal Veterinary and Agricultural University (Clemen Rasmussen and Svend Rasmussen)

Statistical advice: Section of Mathematics and Statistics, Dept. of Mathematics and Physics, Royal Veterinary and Agricultural University (Ib Skovgaard et al.)

Project coordination: Center for Ecology and Environment / Dept. of Ecology, Royal veterinary and Agricultural University (Peter Esbjerg)

Further the project has been assisted in assessments of annual yield reductions and of final weed accumulation by "The foundation for Sugar Beet Research", Alstedgaard, (Jens Nyholm Thomsen), "Syd-Østsjællands Landboforening" (Jørgen Dabelsteen and Niels Skov) and "Køge-Ringsted Landboforening " (Morten Nygaard).

For the practical execution of the project it was necessary to find landowners with suitable fields and crops. Among a rather limited number suitable of farms with suitable fields the owners and managers of the five listed below accepted the contract based collaboration.

Gjorslev Estate, Landowner Peter Tesdorf and Chief Manager J. Klith Jensen
Lekkende Estate, Landowner Andreas Hastrup
Nordfeld Estate, Landowner Jens K. Haubro
Nybøllegård, Landowner Thomas Christfort
Oremandsgård Estate, Landowner Daniel Hage and Manager Mogens Tved

The above landowners and managers are thanked for their collaboration without which the project could not have been carried out. Thanks are also due to all other assisting organizations and persons who contributed to a multifaceted project like this.

Furthermore thanks are due to the below members of the steering committee of the project:

Henning Clausen, Inge Vibeke Hansen, Jørn Kirkegaard and Kaj Juhl Madsen, Danish Environmental Protection Agency
Niels Elmegaard, National Environmental Research Institute, Silkeborg
Chris Topping, National Environmental Research Institute, Kalø
Jørgen Jakobsen and Jesper Rasmussen, Danish Institute of Agricultural Sciences, Flakkebjerg

Summary and conclusions

This report presents the results of investigations of responses of wild flora, insects and birds in arable fields to reduced dosages of pesticides. The investigation was related to the Danish Pesticide Action Plan I (1987-96) and complied with requests from the financing Ministry of Environment and Energy as regards large scale, three dosage levels and technical readiness for practical implementation. The studies were carried out in 1996 (pilot study at one farm) and 1997-99 (main study at five farms). All farms were situated on clay soils in southeastern Denmark.

Hosting was contracted with five large farms with spring barley, winter wheat and sugar beets as crop rotation. The two cereal crops differ in structure and cover the major part of Danish arable land, while sugar beet is a row crop with built-in problems due to lack of competitiveness. All study fields were sufficiently large to include three dosage plots of 6 hectares or more. In these plots, herbicides and insecticides were applied at normal, half and quarter dosages whereas fungicide dosages were not reduced. The pesticides used and the dosage level defined as normal were at each occasion decided upon by the farmer, based on his experience. From a scientific point of view this was inconvenient, but anything else would have been meaningless due to local variations of the weed problems. In this way also the request for practical implementation was met. In the pilot study, broad swath application of reduced dosages in sugar beets led to unacceptable amounts of weeds. Therefore band spraying was used to obtain the two reduction levels in the main study. As a supplementary weeding, mechanical hoeing was carried out on the farmer’s decision.

The number of weed species present was determined before and after herbicide applications, and weed densities and phenology were recorded. Also the seed rain was studied. Sampling of insects was performed every 7-10 days from mid-May to mid-August using a 4WD-tractioned vacuum sampler and supplementary pitfall trapping in fenced sub-plots was carried out in wheat. Within the same period, birds were censused and their location within the fields recorded every five days. A combination of point counts and line transects was used. As a supplement to the biological investigations, crop yields were determined.

Weed densities after spraying differed significantly between dosages, with 30 plants/m² at normal, 48 plants/m² at half and 55 plants/m² at quarter dosage. Besides there was a considerable difference between crops: 84 plants/m² in barley, 32/m² in sugar beets and 28/m² in wheat. Species richness tended to increase with decreasing dosage (significant in barley only) and rare and scarce species occurred more frequently at reduced dosages. The proportion of flowering species increased with decreasing dosage, and there were indications of increased seed production at quarter dosage.

Insect abundance generally increased at reduced dosages. This was very clear in barley while the picture was slightly obscure for wheat and sugar beets if narrow taxonomic units were considered. An overall analysis of non-carnivores and carnivores in the three crops strongly supported the improvement at quarter dosage. More specifically, higher densities of beneficial insects were found at quarter. Aphids were also more numerous at this dosage level but did not occur in densities of economic importance. Combined analyses showed some correlation between plant and insect abundances. This was clear among others for selected herbivorous weevils and larvae of moths.

The bird counts revealed a change from uniform distribution early in the season towards concentration in plots with reduced treatments. Skylarks, Whitethroats and "small seed-eaters" (Yellowhammer, Linnet etc.) all responded significantly to dosage reductions. The effect was most pronounced for species that breed in hedgerows and search their food in the fields. In July the number of Whitethroat records doubled at quarter dosage while small seed-eaters increased by 50% and Skylarks by 20-25% relative to normal dosage. The effect of half dosage was less clear, but the estimates indicate that half of the improvement at quarter dosage was also obtained at half dosage. The effect of reduced dosages was independent of crop and year.

In cereals, yield was significantly reduced in 3 of 64 cases, always at quarter dosage. Sugar beet yields were reduced in 5 of 32 cases; three at quarter dosage and two at half. The average yield in sugar beet varied much more than in cereals and at one farm the revenue was impaired by 11-27%. Despite this, profitability analyses indicate that pesticide reductions as used here generally have a very limited economic impact, at least on short term. Effects not properly covered are the risk of accumulated weed problems by continuous use of reduced dosages and possible adjustment costs associated with new growing conditions.

In conclusion, both reductions to half and quarter dosages of herbicides and insecticides improve the "nature element" of the fields. The gain at quarter dosage is much more marked than the gain at half. However, use of half dosage will only create negligible, if any, agricultural problems, especially if supplementary control of particular weed patches is carried out. General use of quarter dosage may be more problematic in this respect. Only longer time series can support a more conclusive picture, and the possible side effects of increased mechanical weeding also call for more attention.

Sammenfatning og konklusioner

Rapporten beskriver resultater fra fire års undersøgelser af, hvordan reduceret anvendelse af pesticider påvirker flora og fauna i det danske agerland. Projektet har fra begyndelsen været knyttet til Pesticidhandlingsplan I (1987-1996) og de tilhørende ønsker om at dokumentere mulige effekter af denne plans reduktionsmål. Projektet har fulgt Miljøstyrelsens ønsker om forsøg i stor skala og med et indhold og en metodik, der gør resultaterne umiddelbart relevante for praktisk landbrugsdrift. Undersøgelserne er finansieret af Miljøstyrelsen og er gennemført i et samarbejde mellem Den Kongelige Veterinær- og Landbohøjskole, Københavns Universitet og Ornis Consult.

Formålet med undersøgelserne var at belyse effekterne på vilde planter, insekter og fugle – og deres økologiske samspil – i landbrugslandet, når doseringerne af herbicider og insekticider blev reduceret til halvdelen og en fjerdedel af det normale. Efter et pilotforsøg i 1996 blev undersøgelserne gennemført i 1997-1999 på fem store landbrug på Sydsjælland og Møn.

Undersøgelserne blev gennemført i vårbyg, vinterhvede og sukkerroer. De to kornafgrøder, byg og hvede, dominerer det danske agerland, men adskiller sig med hensyn til fænologi, ukrudts- og insektproblemer. Sukkerroer blev medtaget som rækkeafgrøde med dertil knyttede ukrudtsproblemer og særlige fuglefauna pga. det åbne areal mellem rækkerne.

På hvert af de fem landbrug udvalgtes tre marker på minimum 18 hektar, der indgik i ovennævnte sædskifte. Hver af de 15 marker blev ved forsøgets start opdelt i tre parceller på minimum 6 hektar. Disse blev gennem de tre forsøgsår konsekvent behandlet med henholdsvis normal, halv og kvart dosering af herbicider og insekticider, men blev herudover som grundprincip behandlet ens. De anvendte midler og den "normale" dosering blev i hvert tilfælde valgt af den stedlige driftsleder ud fra vurdering af det konkrete behov. Dette var umiddelbart problematisk ud fra videnskabelige standardiseringshensyn. Imidlertid gav kun denne fremgangsmåde mening i praksis, da især ukrudtsproblemerne kunne variere meget fra sted til sted. I sukkerroer blev reduktionen af doseringerne gennemført ved båndsprøjtning med henholdsvis halv og kvart båndbredde. Herudover blev radrensning foretaget i det omfang, driftslederen anså for påkrævet.

Analyserne af ukrudtsfloraen omfattede registrering af arter og deres tæthed i prøvefelter før og efter pesticidbehandling. Desuden blev floraens fænologi undersøgt. Endelig foretoges undersøgelser af frøregnen fra ukrudtsfloraen i to efterår (1998 og 1999). Forekomsten af planter blev primært opgjort som antallet af vilde plantearter og deres tæthed, og kun i pilotåret blev der foretaget bestemmelse af dækningsgrad og biomasse. Frøregnen blev bestemt ved standardiseret sugning af jordoverfladen efterfulgt af frøbestemmelse under stereolup.

I insektdelen blev undersøgt, i hvilket omfang reducerede doseringer af pesticider påvirker mængden af insekter. Der blev specielt fokuseret på naturlige fjender af skadedyr og på vigtige fuglefødeemner. Insektmængden blev bestemt ved, at der i hver afgrøde indsamledes prøver med en 4WD-trukket maskine til sugeprøvetagning (figur 3.2). Der blev i hvede desuden benyttet faldgruber (plasticbægre nedgravet i niveau med jordoverfladen) til indsamling af især løbebiller på afgrænsede arealer. Herudover blev bladlusangreb optalt i marken. Efter indsamling blev insekterne identificeret, optalt og totalvægten beregnet. Antal og totalvægt blev brugt som mål på forskelle mellem doseringer.

De ornitologiske undersøgelser skulle belyse, om fuglene foretrak parceller med reducerede pesticidbehandlinger. Fuglene i forsøgsparcellerne blev i alle årene optalt ca. hver femte dag i perioden maj-juli. Ved hvert besøg gennemførtes optællinger formiddag og eftermiddag med to forskellige teknikker, rettet mod forskellige arter. Herudover blev fuglenes fordeling på de afhøstede marker undersøgt i efterårene 1998 og 1999, ligesom forekomsten blev undersøgt i vinteren 1997-98.

I byg og hvede gennemførtes standardiserede udbytteforsøg i lille skala for at belyse, i hvilket omfang reduktionen af pesticiddoseringerne påvirkede udbyttet. Med samme formål blev der i roer foretaget standardiserede prøveoptagninger med efterfølgende bestemmelse af udbytte. Efter afslutningen af det sidste forsøgsår blev samtlige forsøgsmarker gennemgået med henblik på at vurdere, i hvilket omfang der var sket en akkumulation af ukrudtsproblemer i perioden med reducerede doseringer.

Som storskala-forsøg har projektet arbejdet med meget store datamængder, men også med betydelige variationer mellem lokaliteter, marker, afgrøder og anvendte pesticider. Fordelen ved dette design er, at resultaterne må antages at have stor generaliseringsværdi. Ulempen er, at der kun i ringe omfang opnås en dybere forståelse af de processer, der ligger til grund for resultaterne. Ændringerne i forekomsten af planter, insekter og fugle samt de mulige årsager hertil er derfor behandlet på et ret overordnet plan. Dette gælder også konsekvenser for udbytter og driftsøkonomi.

De botaniske undersøgelser viste, at ukrudtstætheden efter sprøjtning ændrede sig signifikant med aftagende dosering, fra 30 planter pr. m² ved normal dosering til 48 planter pr. m² ved halv dosering og 55 planter pr. m² ved kvart dosering. Der fandtes store forskelle i ukrudtstæthed mellem afgrøderne, med gennemsnitligt 84 planter pr. m² i vårbyg mod henholdsvis 32 og 28 planter pr. m² i roer og vinterhvede. Ukrudtstætheden før sprøjtning over de tre år var uforandret uanset dosering, så der kunne ikke ses nogen generel, akkumuleret effekt af dosisreduktionerne. Følgende arter/slægter var de hyppigste i markerne: Ager-stedmoderblomst, Enårig Rapgræs, Ærenpris sp., Vej-pileurt og Tvetand sp. De hyppigste arter var samtidig de dominerende med hensyn til tæthed. For to tredjedele af de dominerende ukrudtsarter steg gennemsnitstætheden med aftagende dosering. Sjældnere arter som Liden Vortemælk og Nat-limurt viste også en større tæthed med aftagende dosering.

I alt 85 ukrudtsarter blev fundet i forsøgsfelterne. I så godt som alle tilfælde øgedes antallet af ukrudtsarter signifikant med aftagende dosering (figur 2.10), og sjældne og fåtallige ukrudtsarter var mere talrige ved reducerede doseringer. I gennemsnit øgedes artsantallet med 16% ved reduktion fra normal til halv dosering og med 28% ved reduktion til kvart dosering. Der skete ingen ændring i antallet af arter pr. parcel i løbet af forsøgsperioden, hverken før eller efter sprøjtning. De dominante ukrudtsarter var de samme for de enkelte marker uanset afgrøde.

Andelen af arter, der blomstrede, steg med aftagende dosering, og der var signifikant forskel på normal og kvart dosering. Også andelen af blomstrende individer af de enkelte arter tiltog med aftagende dosering i kornafgrøderne. I roer var det modsatte tilfældet, utvivlsomt pga. radrensningens supplerende effekt på ukrudtsplanterne. Studierne af frøregn i kornafgrøderne antydede en øget frøsætning ved kvart dosering, men der var ingen signifikante forskelle mellem doseringerne på antal og biomasse af de fundne ukrudtsfrø. Ligeledes sås ingen kumulativ effekt fra 1998 til 1999.

Insektlivet udviste generel fremgang som effekt af nedsatte pesticiddoseringer. Der var dog en betydelig forskel på effekten i de tre afgrøder. De største forskelle mellem doseringer fandtes i byg, hvorimod effekten i hvede og roer var væsentlig mindre. Samtidig observeredes et noget varierende udslag på specifikke insektgrupper. Imidlertid viste en sammenfattende analyse af rovlevende og ikke-rovlevende insekter meget entydig fremgang ved kvart dosering (tabel 3.11). Der blev bl.a. fundet en større mængde af nyttedyr, f.eks. mariehøns og løbebiller ved mindre sprøjtning, men samtidig blev der observeret en stigende mængde bladlus ved lavere doseringer.

Der er klare tegn på, at effekten skyldtes en kombination af nedsatte doseringer af insektgifte og en øget og mere varieret forekomst af ukrudt. Korrelerende analyser mellem planter og insekter viste en vis sammenhæng mellem de to biologiske niveauer. På trods af nogen usikkerhed fandtes flere signifikante korrelationer mellem forekomsten af ukrudt og flere planteædere, bl.a. sommerfugle og snudebiller. Den øgede ukrudtsfauna har, udover en effekt i form af en øget fødemængde, sandsynligvis også en effekt på mikroklimaet til gavn for specielt tørkefølsomme insekter.

De ornitologiske undersøgelser viste, at fuglene i maj var nogenlunde jævnt fordelt, men i løbet af sommeren koncentreredes i de parceller, der var behandlet med reducerede doseringer, især kvart dosering. Tre arter/grupper (Sanglærke, Tornsanger samt gruppen "små frøædere") var langt de hyppigst registrerede og de eneste, hvis forekomst blev analyseret statistisk. Alle tre viste signifikante forskelle mellem doseringer; men de klareste effekter fandtes hos de arter, der yngler i de omgivende hegn og primært udnytter markerne til fouragering.

I juli blev der registreret 20-25 % flere Sanglærker, 50% flere små frøædere og 100% flere Tornsangere i parcellerne med kvart dosering end i parcellerne med normal dosering. Effekten af halv dosering var mindre klar, men der var stadig signifikante forskelle. Estimaterne antyder, at i hvert fald halvdelen af gevinsten ved kvart dosering også opnås ved halv dosering. Analyserne af fuglenes forekomst i forhold til mængden af insekter og ukrudt gav ikke noget entydigt billede. Der var dog en klart signifikant sammenhæng mellem total-biomassen af insekter og antallet af Tornsangere i marken.

Afgrøden havde stor indflydelse på antallet af fugle, og fuglenes afgrødepræferencer ændredes i takt med afgrødernes tilvækst. Analyserne afslørede dog ingen vekselvirkning mellem afgrøde og dosering, så effekten af de reducerede doseringer på fuglene var den samme uanset afgrøden. Ligeledes var doserings-effekten den samme i alle tre forsøgsår. Analyserne af fuglenes forekomst i efterårs- og vintermånederne viste ingen signifikante forskelle mellem doseringer.

Undersøgelserne af udbytter og driftsøkonomi var sekundære i forhold til de biologiske undersøgelser, og resultaterne skal derfor tages med forbehold. I kornafgrøderne fandtes en signifikant udbyttenedgang i 3 ud af 29 markforsøg med kvart dosering, men aldrig ved halv dosering. Generelt blev udbyttenedgangene mere end opvejet af de lavere omkostninger til pesticider. I roer var udbyttet signifikant reduceret i 3 ud af 16 parceller med kvart og 2 ud af 16 parceller med halv dosering. Gennemsnitsudbyttet var kun svagt nedsat; men variationen var langt større end i kornafgrøderne, og på en enkelt bedrift reduceredes dækningsbidraget med 11-27 %.

De driftsøkonomiske beregninger viser, at de foretagne reduktioner af herbicid- og insekticiddoseringerne generelt har kunnet gennemføres uden nævneværdige økonomiske konsekvenser. Vurderingerne efter forsøgets afslutning – samt den eksisterende viden på området – indicerer imidlertid, at anvendelse af reducerede doseringer af herbicider på længere sigt kan medføre en akkumulation af problemukrudt, i hvert fald pletvis. Dette forhold, og de dertil knyttede omkostninger, har kun i begrænset omfang kunnet inddrages i beregningerne.

Sammenfattende kan det konkluderes, at en halvering af de normalt anvendte doseringer af herbicider og insekticider har en målelig, men dog begrænset, positiv effekt på det dyrkede lands vilde planter, insekter og fugle. Under forudsætning af at der suppleres med målrettet bekæmpelse af pletvist forekommende problemukrudt, vil en sådan reduktion have yderst få, om overhovedet nogen, negative konsekvenser for driftsøkonomien i landbruget. En yderligere reduktion af pesticiddoseringerne til en fjerdedel medfører mere markante forbedringer for plante- og dyrelivet, men er i den foreliggende undersøgelse ledsaget af signifikante udbyttenedgange i ca. 10% af tilfældene i korn og ca. 20% af tilfældene i roer. En endelig konklusion fordrer dog undersøgelser over længere tid, ligesom en vurdering af den økologiske betydning af alternativ mekanisk bekæmpelse af ukrudt bør foretages.

1 Introduction

(Esbjerg, P. & Petersen, B.S.)

1.1 Background

The present project is a derivative of the governmental Pesticide Action Plan I of 1986. The goal of this plan was a reduction in two steps from 1987 to 1997 of pesticide use both in terms of amount of active ingredients used and in terms of treatment intensity expressed as treatment intensity index.

The treatment intensity index is the theoretical number of pesticide treatments per hectare, as calculated by dividing the amount of pesticides used on the land of current interest with the dosage indicated on the approved label of the particular products (Danmarks Statistik 1992). The Pesticide Action Plan used the treatment intensity index for all arable land in Denmark as a practical measure. Within the first five years of the plan, the first step of a 25% reduction in sold amount and in treatment intensity index should be achieved, and finally in 1997 via the second step a total of 50% reduction should be achieved. The overall aim of the reductions was to protect the groundwater against pollution and the flora and fauna against further degradation.

A few years before the end of the 10 years period of the plan, however, it became increasingly clear that the reduction in treatment intensity index was unlikely to be reached. Parallel with this it was debated which effects the reductions aimed at would have in practice. The so-called Bichel Committee was appointed with the task of evaluating the consequences of a wholly or partly phasing out of agricultural pesticide use in Denmark. Based on their report, a Pesticide Action Plan II was approved in 2000. In general terms, the goal of this action plan is to reduce the use of pesticides as much as possible without significant economic losses. As an interim goal, the treatment intensity index (summed for all pesticide classes) shall be less than 2.0 by the end of 2002.

The public debate, and later on the work within the Bichel Committee, disclosed that scientific evidence of the effects of different levels of pesticide use on flora and fauna in general was sparse. In earlier investigations of pesticide free field margins (Hald et al. 1988) some positive effects on flora and fauna had been demonstrated, and Braae et al. (1988) had found higher densities of birds at organic farms than at conventional farms. An experimental documentation of effects on flora and fauna by reducing the pesticide dosages in general was, however, lacking. Furthermore, there was a need for a project at such a scale that its results might be regarded of general value and immediately relevant for practical farming.

Large-scale studies had previously been carried out in the English Boxworth Project (Greig-Smith et al. 1992). In this project, running from 1982 to 1988, the effects of three different pesticide regimes (Full insurance, Supervised and Integrated) on flora, invertebrates, mammals and birds were investigated. However, the low-input regimes at Boxworth by and large correspond to normal farming practice in Denmark today, rendering the results of limited value in the present context.

1.2 Overall project design

1.2.1 Aim and conditions

In accordance with the tender documents the aim of the project has been to demonstrate possible effects on flora and fauna of a reduction of pesticide dosages. The effects in this context should be and have been investigated in a broad sense with changes in occurrence of a wide range of plants and animals as the measure. It follows from this that the project aim has in no respect included investigations of dose-response relationships of specific organisms to specific chemicals or groups of chemicals.

Throughout the report, the term "biodiversity" is used in a rather inexact sense, an increased biodiversity referring to an increased species richness or to increased frequencies/densities of one or more species of plants or animals. This is in line with current terminology of the Danish Environmental Protection Agency.

From the beginning of the project a series of conditions had pronounced influence on the selection of study areas, pesticides and possible technologies.

	An economical frame, which despite wide did not allow the inclusion of all groups of pesticides or the inclusion of zero treatment plots.
	A demand for large areas to enable bird observations and to demonstrate possible consequences for flora and fauna at a scale which was relevant to agricultural practice and economy.
	A request for, if possible, including a minimum of three dosage levels and observing both direct and indirect affects across trophic levels.
	A request for working with widely cultivated crops with high levels of pesticide use as well as to problems anticipated in case of reductions in pesticide use.
	A request for using current farming technologies or, if deemed necessary, to introduce only changes in farm technology which could be brought into practice almost immediately.

Besides the above points it has been an obvious desire to use methodologies which would ease comparison with earlier results of related investigations on flora and fauna of the arable land.

In order to respect the management by the farmers and to minimize interference with normal farming practice, the experimental fields were in all aspects, except those closely related to pesticide dosages, treated in accordance with normal practice at the site. Hence all the variation which would otherwise be minimized as much as possible to ensure transparency and unambiguousness of scientific results has been included. Apart from practical considerations, the advantage of this approach is that the general value of the results is probably increased.

1.2.2 Selection of pesticides

In the setting of priorities concerning pesticide types and their effects, herbicides were the first to be included. The reasons were their pronounced dominance in practice and their known and presumed effects: directly on plants (1^st trophic level) and indirectly on food availability and behaviour of insects (2^nd and higher trophic levels) and on birds (2^nd and higher trophic levels). Insecticides were included because of their status as the generally most toxic plant protection chemicals to animal life (Candolfi et al. 1999, Samsøe-Petersen 1993, 1995a,b) and because of their wide-ranging presumed effects: directly on target organisms, on other herbivorous insects (2^nd trophic level), on predators and parasitoids (3^rd and higher trophic levels), and indirectly on parasitic and predatory insects and birds (3^rd and higher trophic levels).

Fungicides were left out because major pesticide effects of principal interest could be demonstrated through the use of herbicides and insecticides. Furthermore, the number of potential direct and indirect effects would be high, but also very difficult to demonstrate at field level. Interesting topics include effects on insect pathogenic fungi on both pests and beneficials as well as effects on fungal food sources of arthropods serving as a food reserve for many predatory insects. Also effects on fungi antagonistic to plant pathogens could be of interest but very difficult to demonstrate in the field. The economic burden of producing unambiguous results of this nature, however, proved prohibitive already in the planning phase.

1.2.3 Selection of crops and farms

Winter wheat and spring sown barley together account for the vast majority of the Danish cereal area (82%) and in addition for a remarkably large proportion (46%) of the total arable land in Denmark (Landboforeningerne 2001). As regards pesticide applications winter wheat is a relatively insecticide intensive cereal crop because of its sensitivity to cereal aphids. In addition to this, winter wheat is often treated with herbicides in both autumn and spring.

Sugar beets cover a relatively small area but can act as a key representative for row crops with their inherent weed problems due to a limited vegetation cover until late in the season. Furthermore, sugar beet for contract-based delivery is a row crop of significantly higher economic value than cereals, but still not with the same room for more costly operations as vegetable row crops.

Look here!

Fig. 1.1.
The location of the five study sites. A: Gjorslev Estate, B: Oremandsgård Estate, C: Lekkende Estate, D: Nøbøllegård, E: Nordfeld Estate ( ã Kort og Matrikelstyrelsen (A. 42-02)).

In order to meet the area requirements of the ornithological studies it proved necessary to work with experimental plots of a minimum of 6 hectares. With three dosages this implied a need for fields of at least 18 hectares for every crop included at every farm. Because of the additional demand for a 3-years crop rotation of spring barley, winter wheat and sugar beet, the number of suitable farms was limited, as mentioned in the preface. Already in the planning phase it became clear that travelling time would be a major constraint because of the number of operations to be performed. Taking this into account it was decided to work on the five estates / large farms listed in the preface and shown in Fig. 1.1. Despite considerable soil variation at the local level, all study sites are placed on relatively heavy soils with a high fraction of clay.

1.2.4 Pesticide dosages

In accordance with the requests mentioned earlier it was decided to work with a defined normal dosage, and with 50% and 25% of this. Zero treatment plots, however attractive from a scientific point of view, were not included for practical and economical reasons. In terms of standardization and scientific practice it would seem desirable to use cropwise exactly the same pesticides and dosage levels at all five farms. This would, however, have been of limited sense in practice, primarily because of the occurrence of particular weed problems and accordingly special herbicide requirements at four of the five farms. Furthermore, it does not make sense to use the same insecticides if aphids are the main problem in one season but caterpillars in the next season.

With these complications and the demand for a close connection to practice, it was decided to request each farmer to make at every instance his own decision, based on local experience, about which chemical(s) to use, when to apply, and at which dosage(s). By definition, the farmer’s choice in the particular instance was then the normal dosage, despite all the scientific trouble implied. The amount of liquid applied had to be the same at all dosage levels of a particular treatment, with the exception of sugar beet (cf. below).

It might appear attractive to use the politically oriented measure: the treatment intensity index, calculated separately for herbicides and insecticides, as a common yardstick for dosage levels. However, it should be noticed that because, e.g., different herbicides contain different active ingredients with different effects on plants, this measure is of limited scientific value.

All chemicals applied, and the normal dosage of each, are listed in Appendix A. It should be carefully reminded that in accordance with the above-mentioned constraints, the normal dosage of a certain chemical may differ between farms and years. In Table 1.1 the farmers’ use of herbicides and insecticides during the study is summed up as treatment intensity indices. The table shows that the treatment intensity indices varied a lot from farm to farm and from year to year (field to field), reflecting the very different products and dosages chosen. For each of the three crops the farmers involved in this study had, on average, a higher treatment intensity index of both herbicides and insecticides than the average treatment intensity index in Denmark for the same crops.

Table 1.1.
Summed treatment intensity index for herbicide and insecticide applications at normal dosage. Herbicides are divided in products against broad-leaved species and products against grasses (Elymus repens and Avena fatua). "Target 2002" indicates the goals set up in Pesticide Action Plan II.

Look here!

The use of herbicides in the crops was 24% higher than the average use in Denmark, while the use of insecticides was 119% above the average. These higher treatment intensity indices may be due to the location of the farms on high quality soil, where the average treatment intensity index normally is higher than on low quality soil (Jensen 2001). This is supported by the fact that the smallest deviations are seen in sugar beets, which are grown much more often on high quality than on low quality soils. Also the location of the farms in eastern Denmark might increase the use of insecticides, since aphid attacks are more frequent in eastern than in western Denmark (Poul Henning Petersen pers. com.). In addition, large farms/estates (as the experimental farms in this study) in general have a higher treatment intensity index than small farms (Jensen 2001). Due attention must be paid to these differences if the results of this study are extrapolated.

In sugar beets a further peculiarity proved necessary in order to combine the scientific project aims and the applied aspects. During the pilot phase in 1996 at Gjorslev, the sugar beet plots were treated with broad swath at all dosage levels. This practice, however, left a cover of weeds at half and quarter dosages which, irrespective of surprisingly small immediate yield effects, was unacceptable to farmers. The problem had to be solved if the project was to carry on for the subsequent three years, and again practicability influenced the solution despite some scientific shortcomings.

The method chosen was broad swath application at normal dosage, and band spraying (with normal dosage) in the half and quarter dosage plots, so that only 50% and 25-28% (minimum band width), respectively, of the area were treated in these plots. In addition to this, supplementary mechanical hoeing was carried out in all half and quarter dosage plots with sugar beets and also in the normal dosage plots depending on the desire of the individual farmer. (At Gjorslev the combination of broad swath spraying and mechanical hoeing had already been practiced for a number of years).

A table of field operations performed at each farm in each study year is enclosed as Appendix 1.5.

1.2.5 Field plots

At all farms, the fields were included in a crop rotation scheme as illustrated on Fig. 1.2. In practice, there could be some distance between the fields, and in rare cases even between dosage plots on fields that might be considerably larger than 18 hectares.

Fig. 1.2.
Schematic example of crop rotation between the three experimental fields at one farm.

Each field was subdivided into three plots, preferably in a regular fashion with the longest axes along the longest side of the field. The plot size of minimum 6 hectares should ensure that a sufficient number of birds were observed, while the rectangular shape allowed cultivation operations to be performed without substantial inconvenience.

Dosages were allocated to plots in a random way with one crucial exception: if the field was surrounded by hedgerows, the middle plot was always used for half dosage while normal and quarter dosage were randomly allocated to the outer plots. The reasons for this decision were as follows: Hedgerows have profound effects on the distribution of most species of birds and some species of arthropods within the fields (cf. section 4.1.1). Thus, plots that are bordering on hedgerows are not fully comparable with plots that are not. One solution to this problem is to randomize the assignment of treatments to plots; this avoids introducing any systematic bias but increases the variation. However, because the dosage-related differences in bird and arthropod densities might well be small, it was feared that such an increase in random variation would prevent the detection of any differences. Therefore, to maximize the chance of detecting any differences at all, it was decided to give priority to a comparison of normal and quarter dosage by allocating the dosages as described. Although the interpretation of the ornithological (and to some extent also the entomological) results from half dosage might be difficult, the results of the botanical investigations (the primary production level) should still be fully valid and thus provide a basis for some general conclusions about the effect of a halving of pesticide dosages.

The selected dosage plots were fixed throughout the project period in order to allow for possible accumulative effects of the reduction of pesticide dosages.

1.2.6 Study sites

Maps of the five study sites are shown in figs. 1.3-1.7, and a short description of the field surroundings is given in each figure caption. Further field data are presented in Appendix B.

Fig. 1.3.
Aerial view of the experimental fields at Gjorslev with the dosage plots indicated. All fields surrounded by dense hedges with scattered trees. A single marl-pit with reeds in Field 1 and three small marl-pits, surrounded by trees, in both Fields 2 and 3. Farm buildings and park immediately N and NW of Field 1. Fields 2 and 3 situated between the park and the lake Gjorslev Møllesø. Deciduous forest on the other side of the lake.
Photo by courtesy of Kampsax.
    

Fig. 1.4.
Aerial view of the experimental fields at Oremandsgård with the dosage plots indicated. Large fields in a fairly open landscape. Single marl-pits, surrounded by trees and scrub, in Fields 2 and 3. Deciduous hedgerows, partly quite open, N and W of Field 1, between Fields 2 and 3, and N of Field 3. Alley with old, broad-leaved trees S of Fields 2 and 3. Old, deciduous wood E of Field 2 and farm buildings towards the SE. Photo by courtesy of Kampsax.
   

Fig. 1.5.
Aerial view of the experimental fields at Lekkende with the dosage plots indicated. Fields located in an open, rolling landscape, sloping towards the SE and towards the strip of meadow and lakes to the NE. Field 1 with hedgerows along the NW and parts of the SW border and scattered trees towards the SE. Hedgerow between Fields 2 and 3 and W of Field 3. Covert with deciduous scrub in Field 1 and a small marl-pit, surrounded by trees, in Field 2. Photo by courtesy of Kampsax.
   

Fig. 1.6.
Aerial view of the experimental fields at Nøbøllegård with the dosage plots indicated. Fields located in an open landscape along the coastline. Field1 bordered by a fairly open hedgerow towards W and with a small village towards SE. Field 2 divided into two parts, c. 250 m apart. Reduced dosage plots with well-developed hedgerow towards the sea and less dense hedgerows towards E and W; normal dosage plot with hedgerow towards Field 3. Field 3 also bordered by hedgerow towards W and partly towards N where a fairly steep slope leads to the seashore. Single marl-pits, surrounded by trees or scrub, in Fields 1 and 3; farmsteads in Fields 2 and 3. Photo by courtesy of Kampsax.
   

Fig. 1.7.
Aerial view of the experimental fields at Nordfeld with the dosage plots indicated. Field 1 undulating, lying between two deciduous woods and surrounded by hedgerows on three sides. Farm buildings towards NW. Fields 2 and 3 situated W of the wood, close to the beach, with well-developed hedgerows on all sides. Both fields divided into two parts by gravel road bordered by scattered trees (Field 2) or hedgerows (Field 3). Single, small marl-pits, surrounded by trees and scrub, in all fields. Photo by courtesy of Kampsax.

1.2.7 Data sampling

The project has comprised a botanical part, an entomological part and an ornithological part. These parts aim at demonstrating effects at different trophic levels and thereby allow for the demonstration of indirect as well as direct effects. This is indicated by the arrows on Fig. 1.8, which also in principle illustrates why the work should focus more on some arthropods than on others. Thus a larva of a particular species which is linked to a particular weed and also is an attractive food item to birds is an insect of focus interest. The population density of such an insect may be affected by herbicide caused food limitation but may also be reduced by insecticide treatments. A further interesting element is added to this web if the herbivorous larva in focus is also prey for some ground beetles which are themselves food items for birds. Hence the entomological part of the project has been oriented as much as possible towards connections to the botanical and ornithological parts.

Very few direct effects are seen in birds, but birds may be affected indirectly by herbicide effects (via plant and arthropod food depletion) and by insecticide effects (via arthropod food depletion). Whereas plant and arthropod responses to reduced pesticide dosages were assessed by density measures, bird response could only be assessed in terms of occurrence (aggregation of birds in plots with reduced dosages). Population density studies would of course have been desirable, in particular if studies of productivity and survival in response to food availability could have been included as well. However, the appropriate scale demands for such studies would be rather different for plants, arthropods and birds. Also, detailed population studies would have to be limited to one crop and/or one site, and to one or two model species at each trophic level.

Fig. 1.8.
Diagrammatic presentation of the role of the organisms included in the study. The arrows indicate which types of animals or plants are consumed by which other types.

The detailed methodologies of the three parts are described within the specific sections in chapters 2 to 4.

1.2.8 Weather conditions

The pilot season 1996 was preceded by a relatively cold winter during which January was very dry. Also March and April were much drier than normal. During the growing season, June and July were rather chill and dry while august was warm (mean 2° C above normal) (Friis et al. 1996).

In the first year of the main study, 1997, February was warm (mean 3° C in contrast to a normal of 0° C). The weather in spring was close to normal, although with a fairly cool and wet May. July was rather warm and August very warm (mean 20.2° C in contrast to normally 15.6° C). The precipitation of the growing season was fairly normal. In the post harvest period, October was rather cold and the first snowfall occurred early (Jensen et al. 1997).

1998 started with a mild winter, with February temperatures as much as 5° C above normal and with a considerable amount of precipitation. The spring was somewhat peculiar, with almost twice as much precipitation in April as normal, followed by a warm and dry May. June was very much like Danish average while July was chill and wet (mean precipitation 92 mm, 40% more than normally). The weather situation in August was quite normal, but September had the lowest number of sunshine hours and October the highest amount of precipitation (171 mm) recorded for more than 100 years. At many locations the moist conditions troubled the autumn seed drilling (Hansen et al. 1998).

The early winter part of 1999 was mild and humid, and with a high precipitation in March plants in larger patches of autumn-sown fields were drowned regionally. This agriculturally problematic situation was eased by fairly normal conditions in April-May, but in June again very high amounts of precipitation occurred (mean 120 mm, normal mean 52 mm) resulting in weed problems at many locations. During July the weather gradually changed, with high temperatures (up to 30° C) occurring towards the end of the month. In August the weather was close to normal whereas September was the warmest for years with a mean of 16.2° C (normal mean 12.7° C) (Sørensen et al. 1999).

1.2.9 Yield effects and economy

As the focus of the project has been on the biological effects, less emphasis was given to effects on yield and agro-economy. However, the need for a scale to justify possible compensations because of reduced yields was met by establishing an array of very small treatment plots of 9 or 12 m x 2.5 m in the cereal fields, according to guidelines of The Danish Agricultural Advisory Centre. For practical reasons these field trials were placed in the normal dosage plots. Because of the small size of the plots it was not possible to assign a particular dosage level to the same area of ground every year. Therefore these plots have not been able to demonstrate the possible accumulation of weeds over the three project years. To some degree, however, the final assessments of accumulated weed problems carried out after the last season compensate for this.

In sugar beet, mini-plots were also used during the pilot phase but obvious difficulties in harvesting questioned the validity of results. Therefore a manual sampling procedure was developed in consultation with "The foundation for sugar beet research" which also took care of the final yield assessment.

Despite these shortcomings of the yield estimates they were also used to obtain a minimum of evaluation of the economic effects of reducing pesticide dosages. This evaluation also leant on the final assessment of weed problems carried out by the project and the corresponding economical compensations based on qualified estimations of the costs of weed removal.

1.2.10 Statistical analyses

Fundamentally, the whole experiment may be looked upon as a fully balanced Latin Square design with five replicates (farms), and three crops rotating between three fields in three years. Due to the cultivation history at each farm, the allocation of crops to fields in the first year was not random, but it is unlikely that this has introduced any notable bias to the results. Each field was divided into three dosage plots which constituted the basic experimental units (45). As described above, the allocation of dosages to plots was not fully random. The implications of this depend on the organisms studied and are discussed in the individual chapters on the botanical, entomological and ornithological studies.

The main focus of the analyses has been on the demonstration of differences between dosages, taking possible crop effects and the (random) variation between farms, fields and years into account. In some cases tests for dosage effects could be performed across all crops, in other cases each crop had to be tested separately. In general, the statistical analyses were based upon general linear models (anovas etc.) with the dependent variable (number of plants, density of birds etc.) being transformed as appropriate to achieve an approximately normal distribution. Sometimes, however, the nature or distribution of the response variable demanded that an analysis based on a Poisson distribution (a generalized linear model) or a non-parametric test was used.

Tests of effects of differences at one trophic level on another (higher) trophic level were performed by including covariates describing density/diversity at the lower trophic level(s) in the analyses of response variables related to the higher trophic level (e.g., arthropod biomass was included in analyses of the occurrence of insectivorous bird species).

The statistical methods are described in more detail in the appropriate sections of chapters 2 to 6.

2 Wild flora

(Jensen, A.-M.M. & Johnsen, I.)

2.1 Pilot studies

The purpose of the initial studies was to select, evaluate and adjust methods for the analyses of weed vegetation in a large-scale field study with different levels of pesticide application. The pilot study was used to find the right methods and a dimensioning of the main study that was adequate to make the collection of a sufficiently large amount of data possible despite large variations in weed populations and pesticides.

During 1996 – the pilot year – investigations were performed at one farm: Gjorslev. Two fields were selected for the pilot studies, one grown with spring barley and one with sugar beets. Time of sprayings and normal dosage of pesticides used on the fields appear in Appendix A.1.

2.1.1 Methods (pilot study)

Three different vegetation studies were performed during the pilot year (1996):

1. Biomass determination of selected weed species in combination with determination of plant developmental stages (phenology).
2. Observations of weed density, cover and number of weed species in permanent sites.
3. Preliminary investigations of the seed bank and the seed rain.

Biomass is a well known variable responding on herbicide use (e.g. Kudsk 1989, Salonen 1992b, Olofsdotter et al. 1994). In addition, the plant biomass is positively correlated to the seed production (e.g. Thompson et al. 1991, Hald 1997). Thus a high biomass produces more seeds available for seed eating birds and insects. The development stage is determined from the numbers of leaves, presence of buds, flowers, seeds etc. on a plant. The presence of flowers and seeds are important for the fauna that eats pollen, nectar or seeds. Thus the developmental stage indicate where in the development phase from seedling to seed setting the plant is, not whether the biomass of the individual plant is high or low. Evidently, the developmental stage and the biomass are positively correlated in a population. The cover of the vegetation is positively correlated to the biomass (e.g. Smartt et al. 1974) within a more or less homogeneous vegetation community. However the growth forms of the species influence cover markedly. The density of plants is also assumed proportional with the biomass, provided each plant has the same average biomass despite the density of plants. However this is often not true as several studies have revealed that the average biomass per plant decreases at high plant densities (e.g. Watkinson 1980).

It has been shown that high weed biomass, density and cover support high faunal density (Chiverton & Sotherton 1991, Moreby 1997). The number of plant species in a vegetation community reflects the diversity of the community, and high numbers of plant species often lead to high numbers of animal species.

2.1.1.1 Determination of biomass and phenology

Only weed species with high densities were selected for this study: Viola arvensis in the spring barley field, and Aethusa cynapium, Atriplex patula and Bilderdykia convolvulus in the sugar beet field. For each species, twenty plants from three to five random quadrates in each plot were sampled at four different days during the growing season (May to September). This resulted in a minimum of 60 individual plants for analysis per dosage per collection. The twenty plants were collected in the order of observation within a quadrate. The developmental stage of the collected plants was determined after the BBCH-scale (Hansen et al. 1995) immediately after collection by counting the number of true leaves, branches, buds, flowers and fruits. To measure dry biomass, plants were dried at 80^o C for 24 hours. Biomass data were log-transformed before means were calculated (=geometric means) and analyses of variance were performed. Means of biomass were analysed parametricly and medians of development stages were analysed non-parametricly by a Kruskal-Wallis test.

2.1.1.2 Observations of density, cover and number of weed species

In every plot, 5 permanent subplots of 25 m x 25 m were chosen at random. The distance to other plots, hedges, habitat islands etc. was always larger than 12 m to avoid impact from farming operations on the field headland, which may differ considerably from those practised on the experimental field (Fielder 1987). Vegetation in headlands and field margins often differ from the inner part of a field with respect to plant density and species composition (Marshall 1989, Wilson & Aebischer 1995).

In each of these 625 m² subplots a number of smaller sampling sites were selected for non-destructive vegetation observations. Hence, four random sampling sites were selected in the spring barley fields and 10 in the sugar beet field. Each sampling site measured 0.6 m x 0.4 m = 0.24 m² and was marked, so it could be visited several times during the season. At each visit, the weeds (seedlings, vegetative or generative plants) were identified to species according to Haas & Laursen (1994), Hanf (1990) or Hansen (1981). In addition, the cover of each species was estimated on a scale from 0 % to 100 % cover of the soil surface. The same person performed the identification of plants and the subjective estimates of cover during the whole season. Values of density, cover and number of weed species were summed up to the level of subplots. Weed density was log-transformed and cover was square root transformed to improve approximation to normal distribution before data were analysed for effect of dosage. Data from each visit were analysed separately.

2.1.1.3 Seed bank and seed rain

The viable seed bank population was studied in both fields in August by sampling five soil cores at six random places per plot, each core was 2.0 cm in diameter and 10 cm in depth. In September, samples of the seed rain were obtained from another six random places per plot. Each seed rain sample was extracted from the soil surface of five 0.1 m² squares using a house vacuum cleaner connected to a transformer placed in the boot of a van. The number of seeds in the seed bank and seed rain samples was estimated and identified to species by allowing the viable seeds to germinate. After approximately a month the seed samples were spread as a 5 mm layer upon a 3 cm thick sterile substrate of vermiculite and permaculite and placed in a unheated greenhouse with presumably very little influx of weed seeds. Germinating seedlings were identified to species, counted and removed. The germination trays were observed for 16 months and watered when necessary. Trays without seed sample addition were left in the greenhouse for control of sterility and any possible seed influx. This germination method has been used with satisfactory results during most of the last century to reflect the amount of viable seeds in the soil seed bank of agricultural land (e.g. Jensen 1969, Roberts 1981).

2.1.2 Results (pilot study)

2.1.2.1 Determination of biomass and phenology

From the fourth collection in September (Fig. 2.1), the mean dry biomass of Atriplex patula was significantly higher in the plot sprayed with quarter dosage than in the plots sprayed with half or normal dosage (Tukey-Kramer tests, p=0.013 and p<0.0001, respectively). No significant effect of dosage was seen in biomass of Aethusa cynapium or Bilderdykia convolvulus at this collection.

Fig. 2.1.
Mean dry biomass (g) per plant for Aethusa cynapium, Bilderdykia convolvulus and Atriplex patula collected mid-September (fourth collection) in a sugar beet field treated with three different dosages of pesticides. Error bars represent the 95 % confidence limits.

The higher biomass of Atriplex patula at quarter and half dosages than at normal dosage was significant from second collection already.

Only at second collection in late June, plant dry weight of Bilderdykia convolvulus showed an effect of dosage. Plants at quarter dosage had a significantly higher biomass than at half dosage (Tukey-Kramer test, p=0.005) and the plants at half dosage had a significantly higher biomass than plants at normal dosage (Tukey-Kramer test, p=0.001).

At third collection in late July a significant effect of pesticide dosage was observed for dry biomass of Aethusa cynapium, mean dry biomass was higher at quarter dosage than at normal dosage (Tukey-Kramer test, p=0.002).

Biomass of Viola arvensis in spring barley

Very few Viola arvensis plants survived spraying with normal herbicide dosage in spring barley, not enough for statistical analyses, thus no observations on biomass and developmental stages were performed in the spring barley plot receiving normal dosage. Immediately before harvest of spring barley in August, significantly smaller V. arvensis plants (mean 13 mg) were found in the plot receiving half dosage in comparison with plants in the plot receiving quarter dosage (mean 23 mg) (t-test, p=0.002). No difference, however, was observed at any other time of collection.

Phenology of Aethusa cynapium

Five weeks after spraying, A. cynapium showed a significant difference in developmental stages at different herbicide dosages in the sugar beet field (Kruskal-Wallis test, p<0.001). Plants treated with quarter dosage were significantly more developed than plants treated with normal dosage (Fig. 2.2). This pattern was found at the first three collection days, while at last collection the differences in developmental stages among dosages were no longer significant (Kruskal-Wallis test, p=0.08).

Fig. 2.2.
Distribution of developmental stages for Aethusa cynapium in the sugar beet field mid-June. A higher figure on the x-axis corresponds to a higher plant developmental stage.

Phenology of Atriplex patula

The developmental stages for A. patula showed significant differences only at the second seasonal collection, plants sprayed with half or quarter dosage were at a higher developmental stage than plants sprayed with normal dosage (Kruskal-Wallis test, p<0.0001).

Phenology of Bilderdykia convolvulus

The developmental stages for B. convolvulus showed no differences between the three levels of pesticides at any of the seasonal collecting times.

Phenology of Viola arvensis

The developmental stages of V. arvensis were significantly higher at second and fourth collecting day for the spring barley plot receiving quarter dosage compared with the plot receiving half dosage (Wilcoxon tests, (p=0.17), p=0.02, (p=0.26) and p<0.001, respectively).

2.1.2.2 Observations of density, cover and number of weed species

Total weed density at quarter dosage rose from week 30 to week 40 (Fig. 2.3A) because of germinating seedlings originating from seed shedding plants in the field (especially Poa annua). The confidence limits for weed density at quarter dosage were huge mainly because one of the subplots had a three times higher density than the other four. The total cover of weed increased from under 0.5 % in the beginning of the season to 4 % on average at the end of July (week 30) (Fig. 2.3B). The cover decreased to 3 % on average in week 40 because some species defoliated. During the season, changes in total cover were greater than changes in total density.

Look here!

Fig. 2.3.
Development in total weed density (A) and weed cover (B) during the growing season in a sugar beet field sprayed in week 19 and 20 with normal or reduced dosages of herbicides. Every dot is a geometric mean of five values. Error bars represent the 95 % confidence limits

Dosage had a significant effect on total density in week 23, 26 and 30 (anovas) giving higher plant density at quarter dosage than at half dosage. The cover was only significantly affected by dosage in week 26 (p=0.014) (Fig. 2.3B).

The most numerous species in the beet field was Aethusa cynapium, whereas more than 50 % of the total weed cover was contributed by Bilderdykia convolvulus.

Fig. 2.4.
A) Density of Viola arvensis in spring barley before and after spraying with normal or reduced dosages of pesticides. B) Change in density of Viola arvensis after spraying, where density before spraying is set to 100 %. Each column represents the mean of five values. Error bars represent the 95 % confidence limits.

Density of Viola arvensis in spring barley

Before spraying, the average density was very low in the quarter dosage plot and much higher and nearly equal in the half and normal dosage plots (Fig. 2.4A).

After spraying with normal dosage the average density of V. arvensis decreased markedly and the average densities in plots sprayed with half and quarter dosage increased (Fig. 2.4B).

The highest number of species was present in the beginning of the growing season (Fig. 2.5). Number of species was decreasing at the same time as the total density was decreasing. No significant effects of dosages were found (anovas).

Fig. 2.5.
Development in number of weed species during the growing season in a sugar beet field sprayed with normal or reduced dosages of pesticides in week 19 and 20. Error bars represent the 95 % confidence limits.

2.1.2.3 Seed bank and seed rain

In the control trays seedlings of Chamomilla sp., Conyza canadensis, Epilobium sp., Erophila verna, Poa annua, Sonchus sp. and Taraxacum sp. germinated. These seeds may have been in the substrate or more likely must have spread from the surroundings into the greenhouse. They are all wind dispersed or grew near the greenhouse. Three of the species (Chamomilla sp., Poa annua and Taraxacum sp.) were observed in the field vegetation. Therefore, it is not possible to clarify whether the seeds of these species were in the soil samples from the field or whether they have arrived from outside the greenhouse. Fig. 2.6 shows the distribution of germinated seeds from the samples grouped according to the proportion of seeds from species also found in control trays, species from the fields, grains and unidentified seeds, respectively. In total 15,921 seedlings were identified from the seed bank and the seed rain. More than 50 % of the seedlings were of species also found in the control trays. Only 8 % of the seedlings were identified as weed species originating from the field.

Fig. 2.6.
Total number of germinating seeds in seed rain and seed bank samples distributed on species from control trays, grains, unidentified seeds and weed seeds from the field.

2.1.3.1 Determination of biomass and phenology

The results of the determination of biomass provided good information on weed response after spraying with reduced dosages in the fields (which have also been shown by several others (e.g. Landbrugets Rådgivningscenter 1999b)). Five of twelve analyses showed a significant effect, in all cases with a higher biomass at quarter dosage than at normal dosage. The changes in weed developmental stages among dosages (and collection days) seem to reflect well the impact of pesticide application on phenology. Six of sixteen analyses showed a significant effect, which in all cases with a higher developmental stage on average at quarter dosage than at normal dosage.

For measurements of biomass and developmental stages, many plants of the same species at each treatment are needed, because of a very large variation between plants and a small variation between dosages. In addition, the investigation demands a very even distribution of plants of the same species in the three plots in a field. In many fields, this would not be possible to achieve except for the most abundant species. Furthermore, the method is very time-consuming, and would be impossible to perform on a sufficient area within each plot in the large-scale field study. In conclusion, the determinations of biomass and developmental stages were abandoned in the main experiment 1997-1999.

2.1.3.2 Observations of density, cover and number of weed species

Density and cover were measured four times through the growing season. Density was more stable than cover after week 26, where the mortal effect of the herbicide treatments had stopped. Thus, measurements of cover vary much over the growing season, as also found by Hill et al. (1989) for several species. However, greater variation in the measurements of cover than densities results in less significant differences between treatments. In this experiment, significant effects of reduced pesticide dosages were found three times in the analysis of total density and only once in the analysis of cover. Furthermore, cover is a subjective personal estimate and differs from person to person (Kennedy & Addison 1987). Therefore, variation would be greater in the main study, where several persons would evaluate the cover, than in this one-year study, where only one person estimated the cover. If a more exact analyse of cover was performed (e.g. a pinpoint analyse) the time used would at least be trebled. However, the advantage would be that exact estimations of cover might give a better estimate of the competitiveness of the weed species (Silvertown & Dale 1991) compared to density measurements. Densities tell nothing about the size of the plants and, hence, how well they compete with the crop for light and nutrition. Plant biomass might still be a better indicator of the food resource available than density of plants of unknown size, even though the two measurements are positively correlated. However, the density is more stable during the last half of growing season than the biomass (which may follow the same flutuations as cover) and may therefore be more useful in analyses of fauna densities, since the large-scale study only allows one registration during each season. On balance, it was decided to abandon the measurements of cover and retain the measurements of density in the main experiment 1997-1999.

Repeated observations of the vegetation through the growing season are time-consuming and contribute little additional information because data are not independent (counting the same plants several times). Figs. 2.3 and 2.5 show that density and number of species do not change after week 26 and until new seedlings germinate. Thus, one registration after June gives sufficient data to determine differences in vegetation density and richness, in relation to the use of pesticides. This agrees with Hald & Lund (1994), who measured weed densities in unsprayed fields at two periods during springtime and did not find any great difference in densities between the periods.

The development in density of Viola arvensis in spring barley (Fig. 2.4) illustrates that it is very important to know the differences between plots before spraying; else false conclusions may easily be drawn. This knowledge can be achieved by observing the vegetation before spraying and use that observation as a covariate in the analyses of the vegetation after sprayings. In the main study, the density and number of species were registered before spraying (in April-May) and once after spraying (in July-August).

No significant effect of dosages on species number was found in the sugar beet field. This might be due to the small area of investigation in the sugar beet subplots. The weed density in the sugar beet field was much lower than the density in the spring barley field, suggesting that the occurrence of different species would also be more scattered in the sugar beet field. If the recorded number of species is low compared to the maximal number of species found in that habitat, the variation between samples will be greater than if the number was close to the maximal number of species possible. Consequently, the sampling sites in sugar beet fields were doubled in size and a higher number of samplings sites were investigated in the main experiment.

2.1.3.3 Seed bank and seed rain

The germination method for seed bank samples proved unsatisfactory, because of the very high contamination of the samples with many arriving seeds from the environment germinating in the trays. Contamination under germination experiments in greenhouses is usual; Jensen (1969) found a contamination of 4 % of the total number of germinating seeds. Nevertheless, in the present experiment the amount of incoming seeds was unacceptablely high. Thus, another method for measuring the seed rain had to be used in the main experiment.

2.2 Vegetation studies

During the main phase of the study (1997-1999), two different kinds of studies were performed, a vegetation study (section 2.2) and a seed rain study (section 2.3).

The objectives of the vegetation study were to detect any effects of reduced pesticide dosages on the wild flora, measured by several vegetation parameters such as density, species richness (=weed diversity), species composition and ability to flower. Data were obtained from 15 experimental fields: five of each crop (spring barley, winter wheat and sugar beets) over a period of three years. The aim was not to study the effectiveness of the herbicides to control (kill) the weed flora and correlate this effectiveness with the individual treatments of individual herbicides, such studies have been performed in many small field trials (e.g. Landbrugets Rådgivningscenter 1999b). Rather, it was aimed to measure changes in the weed flora parameters that might have an impact on the existence of the fauna on the fields.

The objective of the seed rain study was to detect effects of reduced pesticide dosages on the number of species and density of seeds in the seed rain in early autumn. Data was obtained from 4 winter wheat and 5 spring barley stubble fields in 1998 and 4 winter wheat and 4 spring barley stubble fields in 1999.

Fig. 2.7.
One example of locations of plots (Normal, Quarter and Half), subplots and sampling sites (in magnified subplot) in a field grown in a three years rotation with spring barley, winter wheat and sugar beets. Punctuated lines indicate the 12 meter zone to hedges etc.

2.2.1 Methods

2.2.1.1 Sampling design

2.2.1.3 Number and timing of field observations

Eight of the winter wheat fields were sprayed with herbicides both in autumn and in spring, four only in autumn and three only in spring. For a better comparison of effects of spring spraying with sugar beet and spring barley fields, only data from the eleven winter wheat fields sprayed in spring entered the analyses and figures. Observations of arthropods and birds were not carried out in winter wheat fields before spring, so only effects of spring sprayings were analysed in the vegetation. The weed observations before spring sprayings were used as a covariate in the analyses of weed variables after spraying in that way the effects of soil cultivation and autumn sprayings on the weed populations were included in the covariate. The exclusion of winter wheat fields only sprayed in autumn was due to an expectation of very few and only minor changes in the weed populations between spring and summer caused by dosage differences in the autumn sprayings.

2.2.1.4 Flowering status

2.2.1.5 Habitats in sugar beet fields

2.2.1.6 Identification and counting of plants

2.2.1.8 Statistical analyses

Response variables

The basic experimental unit in the analyses of all variables was the plot (n=123, the four wheat fields without herbicide sprayings in spring being excluded). Values from subplots within a plot were summed or averaged to one value for each plot each year. Statistical analyses were performed on the following response variables:

1. Mean density per plot for all weed plants
2. Development of mean total density per plot during the study years
3. Mean density per plot for the most common species in cereals, separately
4. Density of some rare species, separately
5. Total number of species per plot
6. Development of species number per plot during the study years
7. Abundance of rare species
8. Abundance of scarce species
9. Proportion of flowering plants per plot
10. Proportion of flowering species per plot

In the rest of this section, variable number refers to the above response variables.

Fields, where a species was not present in any of the plots after spraying was excluded in the analyses of density for that particular species (variable 3 listed above), because lack of the species was regarded as non-informative with respect to dosage dependency. A plot was omitted from the data analyses of proportion of flowering species (variable 10) if the number of species found in the plot was less than six, due to high stochastic variation in proportions based on small numbers.

Tests

The null hypothesis was that the means of the response variables were not significantly influenced by pesticide dosages (variables 1, 3, 4, 5, 7, 8, 9 and 10) or year (variables 2 and 6). For the variables 1-6 and 9-10, the hypothesis about differences between dosages or years was tested by means of analysis of variance, taking into account the effect of differences between explanatory factors such as crop, farms, fields, years and weed status before spraying (Table 2.1).

The variables were transformed in order to improve the approximation to a normal distribution and make the variance independent of the mean. The data transformations were accepted after running the model, if the plot of residuals against predicted values did not show any apparent trends.

Effects may have accumulated during the study period because the relative dosages applied to a certain plot were the same in all three years. This was taken into account in the analyses of variable 1-6 by using density or species richness before spraying as a covariate. The total density of plants might influence the number of species found, therefore the density of weed plants was used as a covariate in analyses of variable 5 and 6 and the proportion of flowering plants was used as a covariate in analyse of variable 10.

Notice that dosage was treated as a class variable, so no assumptions were made about the effects of half dosage falling in between those of quarter and normal dosages.

The explanatory factors (main factors and interactions) used in the full models are shown in Table 2.1.

Model reduction was performed using an iterative procedure to remove the variables with p > 0.10 until the model consisted only of variables with p £ 0.10. If a significant effect of dosage was revealed, the differences among dosages were tested by a Tukey-Kramer adjustment for multiple comparisons of least-squares means. The analyses were performed using the GLM procedure in SAS (SAS Institute 1999).Because not all winter wheat fields were used in the statistical analyses, the design was unbalanced, and the F-tests had to be modified, this was done using the Random/Test statement in the GLM procedure.

The proportion of variation explained by a certain factor was calculated as the sum of squares for that factor divided with the total sum of squares.

Dosage effects on the abundance of rare and scarce species (variables 7 and 8) were analysed using chi²-tests. Species or groups of species with very low abundance were summed before tested.

Table 2.1.
Explanatory factors included in the variance analyses. An asterisk indicates that the explanatory factor is included in the full model for that particular response variable. df = degrees of freedom.

2.2.2 Results

2.2.2.1 Density

Total weed density

Weed densities after spraying changed markedly with dosage, so the density at normal dosage (mean 30 plants per m²) was significantly lower than at half (mean 48 plants per m²) and quarter dosage (mean 55 plants per m²) (Tukey-Kramer tests, p=0.016 and p=0.008, respectively). No significant difference in densities was seen between half and quarter dosage (p=0.96).

Density differences between crops were highly significant (anova, p=0.007) and explained most of the variation in densities. Total density after spraying was on average 84 plants per m² in spring barley, significantly higher than in sugar beets (32 plants per m²) and winter wheat (28 plants per m²) (p<0.0001 for both comparisons).

Development of weed density

Densities before spring spraying averaged over three crops (mean 98 plants per m²) were significantly higher than after spraying (mean 43 plants per m²) (anova, p_season<0.003) (Fig. 2.8). The density before spring spraying was significantly lower in winter wheat (mean 63 plants per m²) than in spring barley (mean 141 plants per m²) and sugar beets (mean 110 plants per m²), this difference was mainly due to autumn sprayings, winter mortality and a higher competitiveness in winter wheat than in spring sown crops. No significant dosage difference was found in the weed density before spring spraying, though the density at half dosage was higher than at normal dosage with quarter dosage in between.

There was no effect of year on the densities before spraying, but the densities after spraying increased with year independently of dosage (anova, p<0.0028 for year as a continuous explanation factor and p=0.65 for dosage´ year). Thus, no between year differences in the effect of dosage were found, so no traceable accumulation of effects could be detected over the three years.

Densities of seedlings counted prior to implementation of treatments (1997 before) were not significantly different (anova, p=0.72) on plots designed to receive normal, half and quarter dosage of pesticides. Thus, the experimental basis was not biased before the beginning of the study. No effect of dosage was seen on the weed seedling densities before spring spraying in 1998 and 1999, hence after one or two years with reduced pesticide dosages.

Fig. 2.8.
Development of total weed densities before and after spraying over three crops during three growing seasons.
Each bar represents the geometric mean of 12 to 15 values of mean density per m². Error bars represent the 95 % confidence limits.

Viola arvensis, Poa annua, Veronica sp., Polygonum aviculare and Lamium sp. were the most frequent species/genera, occurring in 75-85 % of the plots after spraying. The ten species/genera that occurred in most plots were also the ten species/genera with the highest overall densities; thus, the frequent species were also the dominant species. Table 2.2 lists the twelve most frequent species/genera, and the effect of dosage on their densities in cereal fields (anovas).

Densities of four species/genera showed a significant effect of dosage. Of those, the density at quarter dosage was higher than at normal dosage in all cases, and in one case, the densities at half dosage were higher than at normal dosage (Table 2.2). The estimated mean density of more than two thirds of the species was higher at reduced dosages than at normal dosage despite the density was only significant different in one third of the species.

Table 2.2.
Species/genera found in more than 50 % of the 123 plots. Effect of dosage on the density of each species/genus in cereal fields analysed by variance. Each species is defined as a target (T) or a non-target (NT) species (cf. section 2.2.1.7). Statistical significant difference from normal dosage is indicated as follows:
+: *: 0.05£ p<0.10, *: 0.01£ p<0.05, **: 0.001£ p<0.01 and ***: p<0.001.

Density of rare species

Of the rare species (definition in section 2.2.1.7) found in this field study only Euphorbia exigua and Silene noctiflora were found in sufficient numbers for statistical analyses.

Look here!

Fig. 2.9.
A) Densities of Euphorbia exigua after spraying with different dosages of pesticides in three subsequent years. Every bar represents the geometric mean of three plot means from fields at one farm. B) Density of Silene noctiflora before and after spraying with different dosages of pesticides in 1998 at the spring barley field at one farm. Each bar represents the geometric mean of four values from different subplots. C) Change of S. noctiflora density after spraying compared to density before spraying. Error bars represent the 95 % confidence limits.

Fig. 2.9A shows the densities of E. exigua at different dosages during three years. In 1997 and 1998 there was a significant effect of dosage on density of E. exigua, with a higher density at quarter dosage than at normal dosage. In 1999, no significant dosage effect was found, which might be due to chances. The density of S. noctiflora plants before and after spraying is illustrated on Fig. 2.9B. Because of a huge variation between subplots, no significant effect of dosages was found (anova, p=0.18). However, the figure illustrates that density at normal dosage was lower after spraying than before spraying in comparison to density at half and quarter dosages, which did not decrease (Fig. 2.9C).

2.2.2.2 Species richness

A total of 85 weed species were found within the study plots during the three-year study period (see Appendix C.1). Four of the species were only found before spraying. Of the 81 different species found after spraying 69 of them were broad-leaved species, 11 monocotyledons and 1 pteridophyte. The species grouping by life cycles gave; 4 trees, 18 perennials and 59 annuals of which 30 were purely summer annuals.

The dosage had a highly significant effect on the number of species present in a plot (Fig. 2.10). The variation in the covariates (species richness before spraying and total weed density) explained a significant part of the variation in number of species after spraying (Table 2.3). Crop type explained most of the variation in number of species and had a highly significant effect on the number of species present (anova, p<0.0001). The number of species can not be directly compared between the sugar beets and the cereals, because of the different sizes of the investigated areas. There was a much higher number of species in spring barley than in winter wheat.

Fig. 2.10.
Number of plant species after spraying with normal, half and quarter dosage of pesticides in three crops. Cereals and sugar beets are depicted individually since the areas investigated were of different sizes. Each bar represents the mean of 15 (spring barley and sugar beets) or 11 (winter wheat) values. The error bars represent the 95 % confidence limits.

In spring barley, the effect of dosage was significant and explained 17 % of the variation in species number. The number of species at quarter dosage was 14 % higher than at half dosage (p=0.002) and the number of species at half dosage was 16 % higher than at normal dosage (p=0.004). Most of the variation in number of species was explained by the farm´ year interaction, probably reflecting differences between fields.

Dosage had an almost significant effect on species richness in winter wheat (anova, p=0.054), giving 36 % higher species richness at reduced dosages than at normal dosage. Most of the variation in species richness after spraying was explained by variation in total weed density.

Table 2.3.
Schematic summary of the analyses of species richness. Statistical significance is indicated as follows: +: 0.05£ p<0.10, *: 0.01£ p<0.05, **: 0.001£ p<0.01 and ***: p<0.001. Grey areas indicate explanatory factors not included in the full model.

An effect of dosage on species richness in sugar beets existed but varied between years (Table 2.3). No significant dosage effect was detected in 1997 and 1999, but in 1998 a significantly higher species richness was present at quarter dosage than at normal dosage (Tukey-Kramer test, p=0.032).

Fig. 2.11.
Development of species richness before and after spraying over three crops during three growing seasons. Each bar represents the mean of 12 to 15 values. Error bars represent the 95 % confidence limits.

No significant effect of year was found on species richness neither before nor after spraying (Fig. 2.11).

The species richness found prior to implementation of treatments (1997 before spraying) was not significantly different on plots intended to receive normal, half and quarter dosage of pesticides, giving the optimal base for the experiment. There was a significant difference in species richness before and after spraying (anova, p=0.01). The species richness after spraying was on average 12 % higher than before spraying. This is mainly due to an artefact caused by a more exact identification of plants after spraying than before spraying. Before spraying, some seedlings could be identified to genera only, whereas most plants could be identified to species after spraying.

The dominant weed species were the same in a particular field from year to year, despite the different crops grown on the field, but varied considerably from field to field. Appendix C.1 lists the occurrence of particular species in plots sprayed with normal, half or quarter dosage. Most species, occurring in more than a few plots, did grow at all three dosages, but often with the lowest occurrence at normal dosage. A few species were most common at one or two levels of dosages: Ranunculus repens was found seldom at half dosage compared to plots sprayed with quarter or normal dosage; and Atriplex patula was found seldom at normal dosage compared to reduced dosages. Only one of the species was found exclusively at one dosage: barley as weed was only present in half dosage plots.

None of the species found in this experiment is mentioned in the Danish Red List (Stoltze & Pihl 1998). However, a few of the agricultural weed species found are quite rare in the Northern and the Western parts of Denmark according to Hansen (1981) and Mikkelsen (1989) viz. Chaenorhinum minus, Euphorbia exigua, Kickxia elatine, Silene noctiflora, Stachys arvensis and Veronica hederifolia. Table 2.4 shows the number of plots where the six species were observed at least once during the three years of study. A species was only counted once in each plot to avoid dependent observations, coming from identical populations year after year.

Table 2.4.
Number of plots where six rare species occurred at least once during the three years.

The occurrence of the rare species was not significantly influenced by differences in dosages (chi²-test, df=6 (the three species with low occurrence was summed), p>0.1).

The scarce species (found in less than 6 of the 123 investigated plots) can be divided into different groups based on habitat preferences:

Woodland species: Acer pseudoplatanus, Fraxinus excelsior, Salix sp. and Sambucus nigra.

Crop species: Beta vulgaris ssp. vulgaris, Hordeum vulgare, Medicago lupulina and Secale cereale.

Species from roadsides and meadows: Achillea millefolium, Artemisia vulgaris, Carduus crispus, Cerastium fontanum ssp. triviale, Cirsium vulgare, Festuca rubra, Ranunculus acris ssp. acris and Rumex crispus.

Species growing in dry or wet soils: Arabidopsis thaliana, Arenaria serpyllifolia, Bidens tripartita, Epilobium parviflorum, Filaginella uliginosa and Juncus bufonius.

Arable species: Alopecurus myosuroides, Avena fatua, Chaenorhinum minus, Galeopsis tetrahit, Geranium pusillum, Papaver dubium, Papaver rhoeas, Raphanus raphanistrum, Stachys arvensis, Thlaspi arvense, Veronica hederifolia and Viola tricolor ssp. tricolor.

Table 2.5.
Number of plots where at least one of the species in a group has appeared, at least once in three years.

A higher number of scarce species were found at quarter dosage than at normal and half dosages. This has to be seen in connection with the general higher species richness at quarter dosage than at normal dosage. However, the relative proportion of scarce species might be higher at quarter dosage than at normal dosage.

There was a significant effect of dosage on the group occurrences (chi²-test, df=8, 0.01<p<0.05). A higher occurrence of woodland species than expected was found at normal dosage, moreover the occurrence of crop species was higher than expected in plots sprayed with half dosage (Table 2.5).

The occurrence of woodland species at normal and quarter dosages and not at half dosages might be explained by the experimental design, since a higher proportion of plots receiving normal and quarter dosage were located near hedges (see section 1.2.5). The dispersal of seeds from woody species is expected to be more frequent near hedges. In contrast, crop species occurred most often in the half dosage plots more distant from hedges.

2.2.2.4 Ability to flower

The proportion of flowering plants was on average 44 % ranging from 5 % to 76 %. Fig. 2.12 illustrates the proportion of flowering plants in relation to crop and dosage. Table 2.6 lists the result of the variance analysis for proportion of flowering plants.

Fig. 2.12.
Mean proportion of flowering plants after spraying with normal, half or quarter dosage of pesticides in three crops. Each bar represents a mean of 8-10 values. The error bars indicate the 95 % confidence limits.

A significant effect of dosage was found in the analysis of the proportion of flowering plants, but it varied between crops (Table 2.6). In the cereals, the proportion of flowering plants increased with decreasing dosage. In contrast, the proportion of flowering plants in the sugar beet fields decreased with decreasing pesticide dosage (see sections 2.2.2.5 and 2.2.3.5 for further explanation). Moreover, the proportion of flowering plants was highly affected by the variation between fields.

The proportion of species recorded flowering was on average 62 % but varied between 17 % and 100 %.

Fig. 2.13.
Mean proportion of flowering species after spraying with normal, half or quarter dosages of pesticides in three crops. Each bar represents a mean of 8-10 values. The error bars indicate the 95 % confidence limits.

Dosage had a highly significant effect on the proportion of flowering species in a plot (Table 2.6 and Fig. 2.13). The proportion of flowering species was significantly higher at quarter dosage than at normal dosage (p=0.0032). Moreover, the proportion of flowering plants affected the proportion of flowering species positively. More plants in flower increased the possibility that a higher proportion of different species was represented.

Table 2.6.
Schematic summary of the analyses of proportion of flowering plants and flowering species, respectively. Factors not included in any of the models have been omitted from the table. Statistical significance is indicated as follows: +: 0.05£ p<0.10, *: 0.01£ p<0.05, **: 0.001£ p<0.01 and ***: p<0.001.

2.2.2.5 Habitats in sugar beets

Fig. 2.14 shows occurrence of weeds in the sugar beet fields grouped by their habitat: within the rows or between the rows.

Total weed density and species richness in sugar beets were not significantly influenced by the habitat (Fig. 2.14A and C) (anovas, p>0.1). Nevertheless, both were strongly affected by dosage (see sections 2.2.2.1 and 2.2.2.2).

Fig. 2.14.
A) Total weed density, B) proportion of flowering individuals, C) number of species per plot and D) proportion of flowering species in sugar beet fields after spraying with normal or reduced dosages of pesticides. For each dosage the weed plants are grouped by their habitat in the field: within the beet rows (dense hatching) or between the beet rows (light hatching). Each bar represents a mean of ten plots and error bars indicate the 95 % confidence limits.

The proportion of plants flowering was significantly affected by habitat, although the effect varied between dosages (p=0.036 for the dosage´ habitat interaction). The response of habitat was strongest at the reduced dosages (Fig. 2.14B). No significant effect of dosage was found on proportion of flowering plants within the beet rows, whereas between the rows 44 % of the weed plants were flowering at normal dosage, 26 % at half dosage and only 23% at quarter dosage.

Equally, the proportion of flowering species was close to significantly affected by habitat (anova, p=0.074), with more species capable of flowering within rows than between rows (Fig. 2.14D).

2.2.3 Discussion

2.2.3.1 Weed density

Investigation of the effect of pesticide application on weed densities showed a reduction of 45 - 70 % after spraying (Fig. 2.8), dependent on the dosage applied. This reduction can be related to the effects of the herbicides and to other natural changes in the population occurring between the time of spraying and the registration three months later. This period covers most of the growing season for annual plants, in which some seeds still germinate, seedlings become established, plants flower and set seeds, if the conditions allow it. In addition, many of the seedlings could have died as an effect of intra- or interspecific competition. The yield trials in this study have shown that the densities of weed plants in unsprayed areas of spring barley decreased with 6 % from seedlings to mature plants, whereas the densities in areas sprayed with quarter dosage decreased with 25 % (see section 5.1). Therefore, the reductions in weed densities are mainly due to the herbicide sprayings. No significant differences in weed densities between half and quarter dosages were found, whereas the density at normal dosage was significantly lower.

The biomass reduction after spraying is often higher than the density reduction (Salonen 1993a) indicating that the growth of each surviving plant may be reduced too. Therefore the relationship between weed density and weed biomass is probably dosage-dependent, making density just a rough measure of biomass.

For four of the twelve most dominant species (Aethusa cynapium, Bilderdykia convolvulus, Lamium sp. and Stellaria media), it was possible to show a significant effect of dosage, with fewer plants killed at quarter dosage than at normal dosage (Table 2.2). The responding species were both target and non-target species, indicating that densities of both categories increased at strongly reduced pesticide dosages. Thus, even though, the herbicides were chosen mainly to control the target species, densities of non-target species also decreased at spraying with normal dosage.

Elymus repens is a weed species farmers want to control, but most of the herbicides used in spring against broad-leaved species are not effective against grasses. E. repens is usually controlled by glyphosate in the autumn. Thus, E. repens and Poa annua do not show any significant dosage response to herbicides used in spring. The rare non-target species Silene noctiflora and Euphorbia exigua were negatively influenced by the dosage of herbicides used (Fig. 2.9) though only significant for Euphotbia exigua in two of three years. Many herbicides are broad-spectrum herbicides, affecting the density of target species as well as non-target species, and may result in a local or temporary loss of non-target species (McLaughlin & Mineau 1995), as was confirmed by this study. Use of reduced herbicides dosages may increase the population size and thus improve possibilities of long-term survival in the arable fields.

The different interspecific responses observed at reduced dosages might be due to different susceptibility towards the herbicides (e.g. Cashmore & Caseley 1995).

This study covers very different herbicide products and spraying situations, which increases the statistical uncertainty. However, the results include the great variations found between farms and years and may therefore be of more general value, than studies of responses to one herbicide on one farm in one year.

Many dose response trials have been performed, where plants of one weed species have been treated with many different dosages of one herbicide and the biomass measured afterwards, resulting in a mathematically described dose response curve (e.g Streibig 1992, Streibig et al. 1993, Olofsdotter et al. 1994). Despite only three points on a dose response curve were revealed in this study, it is the first time that large-scale dose response trials have been performed over different weed communities of several weed species, different herbicide products and different years, resulting in much variation. Therefore it is without much value to compare directly with known dose response trials.

The increase in weed densities after spraying over the three years is difficult to explain but might reflect variations in growing and spraying conditions across the different years, resulting in a higher percentage of individuals surviving at all dosages in 1999 than in 1997. Populations of short-lived plant species often vary in number of individuals between years (e.g. Milberg et al. 2000). Moreover, reductions in total weed number caused by herbicide application may vary considerably. One year Derksen et al. (1995) found a 90 % reduction in total weed number due to a herbicide spraying, next year the reduction was only 39 % despite the field, the product and the dosage being similar.Accumulation of dosage effects are described in section 2.3.3.1.

2.2.3.2 Species richness

This study has found a higher species richness in spring barley than in winter wheat (Fig. 2.10). The spring cereals in Denmark have more weed species than winter cereals (Andreasen et al. 1996, Hald 1999), because the weed flora in Denmark has been selected over many years towards the ecological conditions prevailing in spring sown crops (Hald 1999). This is confirmed by the dominance of pure summer annuals in spring barley and by the fact that all except one of the winter annual species present in this study were also capable of germination in spring. Eight of the eleven winter wheat fields sprayed in spring were also sprayed in autumn. The number of species was thus reduced twice, which might be another reason for the lower species number in winter wheat than in spring barley. Furthermore, winter cereals have a higher degree of cover at springtime than spring cereals. The competition in spring from the crop against the weed species is thus stronger in winter cereals than in spring cereals.

Species richness was affected by the dosage of pesticides used; the lowest dosage gave the highest species richness (Fig. 2.10). A reduction in pesticide dosage from normal to quarter dosage resulted in 28 % more species, and a reduction to half dosage resulted in 16 % more species on average over all three crops. Thus, reduced pesticide input promotes higher weed diversity as suggested by Clements et al. (1994). The increase in richness at reduced dosages was not solely an effect arising from the fact that species diversity increases with an increase in plant density (the species-area relationship). This effect was accounted for in the analyses by including weed density as a covariate (with a significant and positive effect on richness). To sum up, dosage affects richness both directly and indirectly through density. A total cease of herbicide use would presumably increase the richness even more as found by Boström & Fogelfors (1999). In this experiment, knowledge of the size of increase in weed richness with no use of herbicides would have allowed us to conclude whether a 28 % increase was high compared to the maximal possible.

It is worth noticing, that it was possible to detect an increased species richness at reduced pesticide dosages, even though the arable fields are poor in plant richness and the potential species pool has been strongly diminished the last decades (Jensen & Kjellsson 1995) and there were large variations in weed communities between fields.

No clear tendencies towards an increase in species richness could be detected between years, even after two subsequent years with reduced dosages of herbicides (Fig. 2.11).

2.2.2.3 Species composition

This study has like others (Andreasen et al. 1991, Wilson et al. 1994) demonstrated that species abundance and species composition also vary considerably from field to field. Despite the fact that all fields are placed on clay soils in the same region in Denmark. Differences between fields may be the result of different agricultural histories before the start of the main experiment and different agricultural practices during the study years, including among other things differences in the pesticide products used and differences in the dosage chosen as normal.

It proved impossible to find indicator species in the sense, that the species exclusively occurred in fields receiving only reduced dosages of pesticides, due to the huge variation in dominant weed species from field to field. Scarce species were found more often at quarter than at normal dosage, which may be due to a higher possibility of being established in quarter dosage plots as seeds from hedges and habitat islands.

2.2.3.4 Ability to flower

No general effect of dosage was seen on the proportion of flowering plants, but dosage had different effects on the ability to flower in cereals and in sugar beets. The expected inverse relationship between proportion of flowering species and dosages was seen in cereals, whereas in sugar beets the proportion of flowering plants was higher at normal than at reduced dosages (Fig. 2.12). This was a result of the mechanical control of weeds between the beet rows (see section 2.2.3.5). A higher proportion of flowering plants at the reduced dosages in cereals is probably accompanied by a higher plant biomass on average, since flowering annual plants has more biomass than vegetative annual plants - in general. Debaeke (1988) showed a positive relationship between dry weight per plant and number of seeds produced per plant. Therefore, it is possible that reduced pesticide dosages will result in a higher seed production per surviving plant as shown for some species (Hald 1993, Rasmussen 1993a, Rasmussen 1993b).

Dosage had a strong impact of the proportion of flowering species (Fig. 2.13), both directly and indirectly through the proportion of flowering plants, resulting in more reproductive species at quarter dosage than at normal dosage. This is to our knowledge the first time it has been demonstrated that reduced pesticide dosages increase a weed community's ability to flower. These results imply that the fitness of surviving plants is higher for plants exposed to reduced dosages than for plants exposed to normal dosage.

2.2.3.5 Habitats in sugar beets

Differences between the two habitats for weeds in the sugar beet fields had a strong impact on the proportion of flowering plants. Hoeing between the rows reduced the proportion of flowering plants with more than 50 % compared to plants exposed to spraying within the beet row. Although the density of plants was not significantly different between habitats, the ability to flower was highly affected. Hoeing operations were often performed later in the growing seasons than the sprayings, and harmed plants at a higher developmental stage. In addition, hoeing dried out the roots of weed plants between rows and promoted new weed seedlings to germinate. Most of these seedlings had not the time to reach flowering and could therefore never set seeds. Although plants are very plastic, they need time to reach a size where the plant has the energy necessary for flowering. This time is longer after spraying than after hoeing, because hoeing is performed later in the growing season than spraying. Hoeing in combination with 25 cm band spraying compared with broad spraying have not increased the weed density significantly, whereas 12.5 cm band spraying in combination with hoeing resulted in significantly higher weed density than at broad spraying (Fig. 2.14). The hoeing in contrast to the spraying gives each surviving weed plants a lower ability to flower than at broad spraying, probably resulting in a lower seed set. However, the overall proportion of species flowering was still higher in low dosage plots than at normal broad-spraying.

2.3 Seed rain study

2.3.1 Methods

2.3.1.1 Field work

2.3.1.2 Laboratory work

2.3.1.3 Data description

2.3.1.4 Control of methods

2.3.1.5 Statistical analyses

The following response variables were analysed statistically:

1. Mean number of weed seeds per m²
2. Mean biomass of weed seeds per m²
3. Mean number of spilt grains per m²
4. Mean biomass of spilt grains per m²
5. Number of weed species per plot
6. Proportion of damaged spilt grains
7. Proportion of damaged weed seeds
8. Proportion of damaged seeds per species

The basic experimental unit in the analyses of all variables was the plot (n=51). In order to improve approximation to a normal distribution and make the variance independent of the mean, the mean number of seeds and seed biomass (variables 1-4) were log_e (y+1) transformed before further analyse. The proportions of damaged grains (variable 6) were square root transformed and the proportions of damaged weed seed (variable 7) were log (arcsine (y)) transformed. The data transformations were accepted, if after running the model the residuals plotted against the predicted values were without apparent trends.

The proportion of damaged seeds was calculated as the number of damaged seeds divided by the number of damaged and whole seeds per plot. Data for variables 1-5 included damaged seeds. The factors used to explain the variation in seed rain were: dosage, crop, farm, year, density of generative plants, the sample weight and the interactions dosage´ crop, dosage´ farm, dosage´ year, crop´ year, dosage´ crop´ year and crop´ farm´ year. The log-transformed sample weights were used in the analyses, to allow for possible effects of sample size on the number of seeds found.

Response variables 1-7 were analysed separately in a general linear model with the explanatory factors mentioned above. The number of explanatory factors was reduced during the full-scale model calculations using an iterative procedure removing the variables with p > 0.1 until the model consisted of variables with p £ 0.10 only. The analyses were performed using the GLM procedure (with Random/Test statement) in SAS (SAS Institute 1999).

The proportion of damaged seeds of a given species (variable 8) was tested with a Kruskal-Wallis test with regard to differences in median values between dosages. Only plots with more than ten seeds of a given species were tested, because a rather large stochastic variation exists in proportions calculated from small populations.

2.3.2 Results

2.3.2.1 Seed number and biomass

Look here!

2.3.2.2 Species richness

2.3.2.3 Damaged seeds

To detect if dosage had an effect on the proportion of damaged seeds a variance analysis was performed on the proportion of damaged weed seeds and damaged spilt grains. The mean proportion of damaged seeds (22 %) and spilt grains (24 %) were almost identical (Fig. 2.17). Most damaged grains were pieces of grains, whereas most of the damaged weed seeds consisted of hollow seeds and to lesser extent of intact seeds and seed shells.

Fig. 2.17.
Mean proportion of damaged seeds (hollow seeds + seed shells + pieces of seeds) at different dosages of pesticides. Error bars represent the 95 % confidence limits.

Dosage did not have any general significant effect on the proportion of damaged spilt grains. Only the interaction between farm, year and crop could explain a significant part of the variation in proportion of damaged spilt grain (Table 2.10).

The proportion of damaged seeds varied between species from zero in Galium aparine and Veronica agrestis/persica to 90 % in Atriplex patula/Chenopodium album. When the proportion of damaged seeds was analysed no significant effect of dosage was seen in the nine most abundant species (Kruskal-Wallis and Wilcoxon tests, 7<n<28, p>0.10 in all nine cases).

The effect of pesticide dosage on the proportion of damaged weed seeds varied between farms. On three farms, the proportion of damaged weed seeds was highest at normal dosage, while the remaining two farms had the highest proportions either at half dosage or quarter dosage. Furthermore, the farm factor had a significant effect on the proportion of damaged weed seeds. Two farms (Nøbøllegård and Gjorslev) had a very low proportion of damaged seeds (10 %), while the three other farms had a high proportion of damaged seeds (42 %). In this context, it is interesting that different weed species were dominant in the seed rain at different farms. By number, Matricaria perforata dominated on Nøbøllegård and Aethusa cynapium dominated on Gjorslev and both Matricaria perforata and Aethusa cynapium had a very low proportion of damaged seeds.

Table 2.10.
Schematic summary of the analyses of proportion of damaged seeds in the seed rain. Factors not included in any of the reduced models have been omitted from the table. Statistical significance is indicated as follows: +: 0.05£ p<0.10, *: 0.01£ p<0.05, **: 0.001£ p<0.01 and ***: p<0.001.

From the trays with discarded parts 78 seeds germinated. In the same 26 samples 2706 whole seeds were found under the stereoscope after flotation (Table 2.11).

Table 2.11.
Total number of whole seeds in the 26 control field samples found under stereoscope or in the discarded parts by germination. - : Indicates that the material has not been investigated.

The flotation method gave a success rate around 97 % for both weed seeds and spilt grains extracted from a sample. No germination occurred in the sediment from the flotation. This might be due to the fact, that there were no seeds in the sediment, but more likely, that some remaining potassium carbonate might have prevented germination (Tsuyuzaki 1993). Thus, the actual success rate may be lower than indicated.

2.3.3 Discussion

No direct effect of dosage was found on the number of seeds, biomass or richness. However, a highly significant correlation was found between the density of generative species in the vegetation and the biomass and number of seeds. In addition, the number of generative species in the vegetation was positively correlated with the species richness in the seed rain (Fig. 2.16). Both weed density and species richness in the vegetation were highly affected by dosage (sections 2.2.2.1 and 2.2.2.2); so reduced dosages have indirectly a positive effect on the seed rain. The great variation found between farms makes it very difficult to detect dosage differences.

No clear tendency towards an increase in germination of weeds could be detected, even in the third year with quarter dosage of herbicides (Fig. 2.8). Ploughing turns around the soil, so seed rain from the first year may not reach the soil surface before spring the third year. These findings are in accordance with Salonen (1993c), who did not find any increase in weed densities during four years without the use of herbicides. The soil seed bank may work as a buffer (Wilson & Lawson 1992), so it takes several years before a possible increase in germination can be detected. A continuous reduction of inputs of herbicides over a period of years has been found to result in an increase of seeds from weedy species accumulating in the soil seed bank, followed by an increased weed problem in subsequent crops (Hill et al. 1989). However, this study does not support those results. No significant response can be due to the low density of weeds at the beginning of the experiment and the lack of problematic grass weed species. In a longer period with use of reduced dosages the effect may be more pronounced. An experiment with low inputs of herbicides through 6 years compared to recommended inputs of herbicides revealed a significant positive effect on the number of seeds in the soil seed bank on one of two farms (Jones et al. 1997).

The higher density of seeds found on winter wheat stubble fields than on spring barley stubble fields may be due to a longer growing season in winter cereals than in spring cereals, especially for annuals germinating in autumn and winter. Thus, weed plants in winter cereals may have a higher biomass when they reach the seed setting phase and may then be capable of a higher seed set (Thompson et al. 1991).

A weed seed rain of approximately 200 seeds per m² right after crop harvest may seem rather low in comparison to soil seed bank estimates of 128,000 seeds per m² in fields in Denmark (Jensen & Kjellsson 1995). However, if seeds of Juncus bufonius are excluded from this figure the soil seed bank estimate becomes only 12.000 per m² (Jensen & Kjellsson 1995). The use of geometric mean (this study) compared to arithmetic mean (Jensen & Kjellsson 1995) make the difference seems larger than it is. The soil seed bank constitutes seed rain from several decades why the content of soil seed banks might be expected to be much higher than the seed rain from one year. In the same period, the farmers have intensified their use of herbicides considerably, their use of more competitive crop cultivars, their use of more "effective" ploughs etc. resulting in a general low weed density (Andreasen et al. 1996). It must be assumed that most seeds in the seed bank are very old, which the low seed viability found by Jensen & Kjellsson (1995) infers.

The dosage effect on number of spilt grains was only observed in fields sprayed with Roundup a few weeks before harvest. The number of spilt grains per m² was lowest in plots sprayed with normal dosage of Roundup. This could be explained by at least two factors, maybe in combination: 1) A homogenous ripening of grains at normal dosages of Roundup, but not at reduced dosages. This may result in lowest escape during harvest at normal dosage because the size and hardness of the grains will be more homogenous. 2) Reduced dosages could result in a higher green weed biomass, which could reduce the effectiveness of the combine harvester and result in a higher amount of spilt grains. Sheppard et al. (1984) showed that pre-harvest spraying with Roundup compared to no spraying resulted in lower moisture content of the grains and lower grain loss from the combine harvester. This might be the case with reduced dosages too, however more experiments are needed before clear conclusions can be drawn.

2.3.3.2 Damaged seeds

This study has shown that a least 24 % of the weed seeds shed in a particular year are hollow or so badly damaged that they are not viable. Investigations of the soil seed bank in Western Denmark (Jensen 1969, Jensen & Kjellsson 1995) have also revealed a very high proportion of non-viable seeds (73 % and 79 %). The pool of dead seeds in the soil seed bank consist of both hollow seeds and whole seeds not capable of germinating, in comparison to the 24 % damaged seeds in this study, that mostly consisted of hollow seeds. The whole seeds found in this study have not been tested for viability, and many might be non-viable. In addition, some seed shells are very robust and might persist for years in the soil seed bank before decomposition although the embryo is dead (Roberts & Ricketts 1979).

On three of five farms, the ratio of damaged seeds was highest at normal dosage. This result might arise by chance or be due to the reduction in herbicide dosages. Surviving plants at normal dosage might be more stressed and more delayed in development than at quarter dosage thus having less energy resources to complete seed formation and a degeneration of the embryo and the endosperm takes place. The existence of huge differences in ratios of damaged seeds between weed species is interesting - a high production of hollow seeds (empty seed shells) seems a waste of energy. Abortion of seeds was expected to happen when seeds were immature and before the seed shells were totally developed and looked 100 % like mature seeds. A plausible explanation for a high percent of empty seeds could be that some annual plants with high seed setting capacity was able to produce a lot of seed shells early in the season and, if the resources are provided, fill them later in the season. Each seed may not represent very much energy for annual plants with high seed setting capacity. This hypothesis is supported by findings of Ogunremi (1986) who found that the percent of empty seeds in Helianthus annua was highest at early harvest time. The phenomenon with intact but empty seeds has been observed but not explained in some other dicotyledonous plant species (Ebadi et al. 1996, M. Philipp pers. com.), but to the authors knowledge the percent of empty seeds has never been recorded as high as in this study, where 90 % of the Chenopodium album seeds were empty. More knowledge is necessary to understand the biological mechanisms behind these results.

2.3.3.3 Species composition

Seed from four weed species made up 46 % of the number of weed seeds in the seed rain: Atriplex patula/Chenopodium album, Stellaria media, Aethusa cynapium and Polygonum aviculare. Leguizamon and Roberts (1982) investigated the topsoil of fields without crops in England and found four species accounting for 87 % of the seeds. This difference could be explained by the different calculations of means: Geometric means are used in this study, decreasing the influence of extremely high values, whereas Leguizamon and Roberts (1982) used arithmetic means. Using the arithmetic mean, the four species with most seeds in this study would account for 72 % of the total amount of seeds.

The species composition of generative plants was different from the species composition in the seed rain. Species like Stellaria media and Chenopodium album (very frequent in the seed rain) had a very high seed set per plant compared to species like Viola arvensis/tricolor ssp. tricolor and Aethusa cynapium (frequent in the vegetation). This is in accordance with previous findings of seed setting capacity of those species (Korsmo 1926) and can explain the differences in species composition between the vegetation and the seed rain. Also the time a weed plant requires from germination to seed set may varies between species, which influence the amount of seeds produced at harvest time.

Seeds from three weed species made up nearly 50 % of the wild seed biomass: Bilderdykia convolvulus, Stellaria media and Atriplex patula/Chenopodium album. Seeds from these species are known as food items for birds (Christensen et al. 1996). It is worth noting that more than 90 % of the total seed biomass on stubble fields consisted of spilt grains. Therefore, grains are the main sources of food available for birds foraging on stubble fields in autumn. In addition, many birds prefer eating grains rather than weed seeds (Christensen et al. 1996, Berthelsen et al. 1997).

This study showed a good qualitative accordance between the species found in the vegetation in July and in the seed rain in September. The lower total richness found in the seed rain compared to the vegetation has many explanations: 1) Some of the seeds can not be identified to species but only to genus, therefore the number of species present is possible higher than indicated (Table 2.9). 2) Some of the species found in the vegetation do not reproduce by seeds or have not reached the reproductive age (e.g. Equisetum sp., Sambucus nigra or Salix sp.). 3) Vegetation was studied on 3.6 m² per plot, whereas the seed rain was just collected from 0.72 m² per plot. The species-area relationship implies that more species would be found in the vegetation than in the seed rain samples. This is supported by the fact that most of the generative species had a scattered distribution and were only found few times in the vegetation study. 4) It is possible that one or two species in the seed rain have been overlooked (e.g. seeds from Kickxia elatine, which looks like a placenta from Anagallis arvensis). 5) Moreover, seeds less than 0.5 mm were excluded from the identification (e.g. Epilobium montanum or Juncus bufonius).

2.3.3.4 Evaluation of methods

The methods used in this study were suitable to reveal the qualitative and quantitative characteristics of the seed rain on stubble fields with high efficiency. The methods were also suitable to determine low seed densities over large areas. However, the methods had some disadvantages.

Random errors

The weight of a sample had a positive significant effect on the number of spilt grain in the sample. This suggests that the sampling method had an impact on the results. The weight of a sample depends among other things on the soil structure, the amounts of straw on the fields and the humidity of the soil at the time of suction. These factors varied a lot from field to field but they did not vary much between plots within a field.

Systematic errors

The estimate of the size of the seed rain was certainly an underestimation, because not all seeds have been sampled by the C-vac method. Firstly, numerous seeds might have germinated after rainfall. Genera like Lamium, Veronica and Viola spread seeds many weeks before harvest of the crop, and some of these seeds might have entered the soil seed bank as a result of earthworms or rainwash (Hurka & Haase 1982). Secondly, some species like Aethusa cynapium set seeds several weeks after crop harvest and the seeds are therefore not shed at the time of sampling. Finally, not all seeds lying on the soil surface will end up in the C-vac, since the effectiveness of the sampling by suction lies around 80 % (Jensen unpublished data ) and varies from species to species. Another, systematic error happens during the identification and counting under a magnifier on a white background, where it is common to miss small and light seeds (Gross 1990). This might also had happened in this study.

All these random and systematic errors implies, that the results found in this study should not be considered an exact measure of the seed rain on stubble fields after harvest. However, the results reveal the proportional distribution of seeds between dosages.

2.4 Summary of dosage effects

In Table 2.12 the dosage effects on all dependent vegetation and seed rain variables are summarised.

Table 2.12.
Summary of dosage effects on the weed vegetation and the seed rain. Percentage increases (+) and decreases (-) in response variable at reduced dosages compared to estimated mean values at normal dosage. Significance is indicated as follows: +: 0.05£ p<0.10, *: 0.01£ p<0.05, **: 0.001£ p<0.01 and ***: p<0.001.

This large-scale field study has illustrated that reduced pesticide dosage affects the weed vegetation in many ways and with profound impacts. A reduction in dosage, from normal to quarter dosage is followed by a considerably higher species richness and a higher density of weed plants. Furthermore, species have a higher probability to flower and set seeds at quarter dosages, so the fitness of surviving species is higher than at normal dosage. All these effects of reduced pesticide dosages on the vegetation may affect the fauna by improving the living conditions at low dosages.

3 Arthropods

(Navntoft, S. & Esbjerg, P.)

Farmland crop pests like Aphididae (aphids) and Oulema melanops (Cereal leaf beetles) as well as their natural enemies, often referred to as beneficials, have received much attention in pesticide research. The term "beneficial" when applied to crop-dwelling arthropods is usually associated with the predatory species of crop pests or with pollinators. There are however many other groups of species whose role within the crops may be termed beneficial. These include the so-called chick-food group; several orders or guilds of insects that are important components in the diet of other farmland species, especially birds (Sotherton & Moreby 1992). In this chapter attention will be paid to this group as well as on the traditional target species.

3.1 Pilot studies

Pilot experiments were conducted primarily to evaluate the time consumption and suitability of different sampling methods under the actual conditions and also to get an overview of the arthropod composition in general and in relation to reduced pesticide dosages.

Sampling of arthropods on plant foliage and on the ground below is an ongoing challenge to entomologists and each method has advantages and shortcomings. No single method of density or abundance estimation is suitable for all circumstances and different sampling methods are therefore relevant in order to obtain a reasonably efficient sampling depending on the target species.

3.1.1 Methods (pilot study)

The pilot experiments were carried out on the farm "Gjorslev" in 1996. Four sampling methods were evaluated in all three crops:

1. Coloured water traps, consisting of circular plastic bowls (diam. 200 mm, height 85 mm). A set of three traps in different colours was placed together on the ground and in each 1 litre of trapping fluid (water added 70% alcohol and ethylene glycol (3:1:1) with one drop of non-perfumed dish soap) to ensure drowning and preservation. In each plot five sets of traps were randomly placed avoiding the outer 20 metres of the plots to limit interference from boarder zones. The traps were serviced once a week from 9 May to 1 August in the cereals and 9 May to 10 October in beets. All catches were preserved in 70% alcohol until further treatment. The target arthropod groups were primarily flying insects and insects living in the canopy.

2. Pitfalls (plastic cups, diam. 110 mm, depth 135 mm) were buried pairwise in the ground 1 meter apart, connected by a 150 mm high steel barrier to improve the efficiency of the traps. The traps were partly filled with trapping fluid (see above). The number, distribution as well as sampling of the traps was performed as described above. The target group was epigaeic arthropods, in particular the adult carabids, a predominant and important arthropod family in the arable land.

3. Suction sampling. Different machinery has been used to sample by suction arthropods living on plant stems, in canopy and on the soil surface. The most widely used is the D-vac (Dietrick et al. 1959, Dietrick 1961). The advantages of vac-sampling is that it provides density estimates in contrast to water traps and pitfalls, which provide relative estimates since their sampling area is not defined. For the pilot experiments a modified hand carried vac-sampler was used. A few samples were taken in spring, but the machinery was not sufficiently powerful for this purpose and the sampling was cancelled.

4. Direct countings were conducted in all crops primarily to estimate Aphididae populations, but in beets also to monitor a severe attack of Autographa gamma (Silver Y’s). Sampling comprised mainly the Aphididae Rhopalosiphum padi (Bird-cherry oat aphid) and Sitobion avenae (English grain aphid) in the cereals and Aphis fabae (Black bean aphid) in beets. On each assessment day and in each plot, 100 randomly selected wheat and barley ears on a diagonal transect (Danielsen 1992) were inspected and the number of cereal ears with Aphididae was counted. In beets, sampling was done on weekly intervals in the period early July to late August. 50 plants per plot were inspected following in principle the same methodology as in the cereals. Four times during the season 21 randomly selected sugar beets plants were inspected for larvae of A. gamma (this was an ad hoc methodology).

All arthropods collected were identified to at least order. Coleoptera (beetles) were all identified at least to family with most emphasis on the pitfall collected Carabidae (ground beetles) among which imagines of larger species were identified to species. In the water traps the most abundant larger Diptera (two-winged) were identified to family. Lepidoptera (butterflies and moths) and Symphyta (sawflies) were not identified further with A. gamma as an exception. The water trap catches could be huge which automatically restricted the sorting to the higher level taxi.

3.1.2 Statistical analyses (pilot study)

Data for all common groups of arthropods, depending on taxonomic level of identification, were analysed for possible dosage effects on the populations. Each crop was analysed separately since each type of crop was considered unique. Also, the data from water traps and pitfalls were analysed separately because of their different mode of action and their different target species, but the statistical method used was the same. The data on each relevant arthropod group from each set of traps on each sampling date were log_e(x+1) transformed to normalise variance. Subsequently the data from the various sampling dates from each trap set were pooled to avoid that repeated measurements on the same spots provided dependent data (Stryhn 1996). The dependent variable (total number per trap group) was assumed to follow a Poisson distribution and was analysed by the GENMOD procedure in SAS/STAT using Likelihood ratio tests (SAS Institute 1990). The variable was analysed in relation to the class variable dosage only and corrections for over-dispersion were made. The replicates (sets of traps) were all from the same plot due to the experimental design, and this lack of "true" replications weakens the reliability of the test results.

The sampling was insufficient for statistical analyses.

No statistical analyses were conducted. Percent infested plants on a given sampling date are presented.

3.1.3 Results (pilot study)

Water traps

The results from the water traps were unsatisfactory, variation in catch was very high and no results will be presented.

It should be stated that pesticide effects on Carabidae and Staphylinidae (rove beetles) populations might be a result of treatments the previous year, during which some of the adult individuals, caught the current year, were at the larval stage. This, of course, weakens the possibility of obtaining significant results in a one-year pilot experiment. Only results of barley and beets are presented (Table 3.1), since no experimental sprayings were carried out in wheat. Significantly different numbers are found within a limited range of families and species of Carabidae and Staphylinidae. The strength of the results, however, in spite of the limitations of the method and statistical analysis, is a consistent picture. The significant differences found, all confirm the trend of higher catches in quarter dosage followed by half and normal dosages.

Table 3.1.
The estimated number of specific Carabidae (ground beetles) and Staphylinidae (rove beetles) per pitfall group on Gjorslev. Numbers are total catches from 9 May – 1 August 1996 in the barley, and 9 May – 10 October in beets. Estimates given are least squares means. Significant differences between dosages (p<0.05, paired t-tests) are indicated by different letters. P-values of the variable dosage are given (*: p<0.0.5, **: p<0.01, ***: p<0.001).

The number of A. gamma larvae per 21 plants per plot was counted 4 times during the season, and results are presented in Fig. 3.1. The counting on 16 July was carried out right before the application of dimethoate (see Appendix A for pesticide treatments). This application was not very effective and another application with Lambda-cyhalothrin was conducted 22 July. The latter application seemingly knocked down the population, and apparently even quarter dosage was enough to strongly diminish the population.

Fig. 3.1.
Estimated population densities of Autographa gamma larvae at the three different dosage levels in sugar beets at Gjorslev 1996. The two relevant insecticide treatments are marked with arrows.

In Table 3.2 results of the aphid countings in beets are given. Only the pirimicarb spraying had an effect, which was obviously dosage related. No insecticide sprayings were conducted in the cereal fields and only slightly differences were found in the Aphididae populations between the different dosage plots.

Tabel 3.2.
Mean percentages of sugar beets plants (Gjorslev, 1996) infested by the aphid Aphis fabae. Dates of relevant insecticide applications are given.

3.1.4 Discussion (pilot study)

Water traps are easy to handle but suffered from the well-known problem of lacking density estimation. Besides insects blown into the traps only animals attracted by the colours and/or humidity may be caught, including species not related to the crop. The results obtained with this method are inconsistent.

Pitfall trapping. Pitfall trapping provided some interesting results. It is, however, very laborious to trap intensively at a large scale, and the subsequent sorting in the laboratory is often very demanding since catches may be high. Especially in beets there may be problems because of repeated weed hoeing. The lack of density estimates can be overcome by fencing the pitfalls thereby sampling a specific area only. This is, however, costly.

Suction sampling. Little experience was obtained due to the insufficient equipment. The method is however, widely used and accepted. The D-vac has proved to be a practical sampling device under a wide range of conditions. It is very effective at sampling most arthropod taxa, diurnally active in the vegetation layer and on the soil surface (Thomas & Marshall 1999). Suction machines may, however, not efficiently sample some life stages, e.g. juvenile stages of beneficials. Examples are larvae of Syrphidae (hooverflies) and Coccinellidae (ladybirds) (Sunderland et al. 1995). Furthermore the efficiency of suction samplers is likely to be influenced by vegetation structures and density. A significant limitation to the use of techniques involving suction samplers is that the habitat to be sampled must be dry (Sunderland et al. 1995), not only to avoid invertebrates getting stuck before reaching the collecting container, but also to ease the following sample treatment. Sampling is also often limited to daylight hours, and may therefore underestimate nocturnal species, which include most of the carabids found in agricultural habitats (Thiele 1977). The method, however, is rather independent of the activity of the sampled organism, and is therefore generally less prone to error (Thomas & Marshall 1999).

Field counting of ears with/without insects is a well-established binomial technique providing useful results for Aphididae and O. melanops. For easily recognisable A. gamma occurring in high numbers, counting the number per plant is a very useful method.

Generally, both pitfall trapping and direct counting proved useful, and it was decided to use both methods in the main phase of the study. The weakness of the pitfall sampling method in relation to obtaining absolute density estimates led, however, to the decision of downgrading the method and to use it in wheat only due to resource limitations.

Instead of pitfalls it was decided to focus on suction sampling as it provides the density estimates, which are relevant in population studies, and also enable comparisons with somehow similar investigations carried out in Denmark earlier. Based on the literature but taking the difficulties during the pilot phase into account, it was decided to develop a more powerful suction sampler than the well-known types. For practical reasons it was also decided to use suction sampling as the "back bone" sampling method throughout the investigations (see 3.2.1.1).

The pitfall-sampled carabids as well as the Pirimor-sprayed Aphididae in beets showed a tendency towards higher densities at reduced pesticide dosages. The extent of the sampling combined with the lack of true replicates, however, obstruct the possibility to make any straightforward conclusions.

Regarding A. gamma there did not seem to be a pronounced dosage effect, due to the common slopes of the curves. It is, however, possible that the higher number of larvae recorded on beets plants at reduced dosages is caused by a higher weed density which may have imposed the pregnant females to lay their eggs there. The actual concentration of larvae especially in the plots half and quarter is higher than illustrated in Fig. 3.1, since larvae found on weed (not counted) contribute significant to the population.

3.2 Main studies 1997-1999

3.2.1 Methods

In order to achieve comparable samples from a total area of at least 270 ha, it was decided to use a 4WD vehicle to transport the suction sampler between the sampling sites. The size of the suction sampler was therefore less important, opening possibilities to construct a more efficient suction sampler than the one used in the pilot studies and other commercially available models. Based on experiences from 1996 a new suction sampler was designed to extract insects from the crops at a higher level of efficiency (Fig. 3.2). Test of sampling efficiency was done with larger carabids known as difficult to extract by suction (Hald & Reddersen 1990). The test showed an 80% recapture efficiency of carabids, which was a higher rate than that of the vac-samplers (incl. the D-va^®) with which it was compared. Further information will be published.

Fig. 3.2.
The suction sampler. A. Suction hose/pipe. B Collection jar. C. Fan providing high speed air flow. The construction was mounted on a trailer with a 4WD as traction.

Many replicates are necessary to obtain useful density estimates. Within each plot, 15 sub-plots (sub-plot: ± 10 m from a marker stick) were selected. Sampling was restricted to the main crop area and plot margins were therefore excluded (minimum 20 m). Thereby interference between plots and effects from field edges was minimised.

The 15 sub-plots were evenly distributed along 3 tramlines in order to minimise crop damage. The length of the suction tubes limited sampling to within 1.5 meters from the tracks. The sampling of one field was always finished within 1 - 1.5 hour. If the sampling of a field had to be cancelled, e.g. because of rain, samples already collected were removed and a new set collected as soon as possible. All samples were collected between 10.00h and 21.00h under dry conditions. Each sample comprised 10 sub-samples of 10 seconds application of the vac-nozzle (total area 0.2 m²) in cereals. In sugar beets it was necessary to reduce the number of sub-samples to 5 (total area 0.1 m²) because large amounts of soil accumulated in the collecting jars due to the lesser plant cover compared to the cereals. Sampling in beets was only done within the rows since hardly any arthropods were observed at the almost bare field between the rows. The samples were labelled and placed directly into cooling bags. Later, the same day, they were frozen until further treatments.

Sampling was carried out 5-7 times during each season from mid May to mid August (Table 3.3). The first samples were taken in wheat at mid May, whereas sampling in barley and beets began by early June. The last samples were collected in sugar beets at mid August whereas sampling ended in early August in the cereals. If possible a sampling was conducted just before insecticide application and no samples were taken until about one week after spraying to make sure that killed arthropods were decomposed. The sampling order of the farms varied, as did the order of the fields sampled within the farms, whereas sampling within the fields always followed the same order.

Table 3.3.
Statistical information about the sampling in the experimental fields. For each combination of crop and year, the number of sample visits is shown (average and range). Also, the period during which the sample collections were performed is shown together with the median date.

Look here!

After defrosting the samples were sieved through 3 grids (2.0, 1.4 and 1.0 mm laboratory test sieves, Endecotts Ltd. London) to extract animals from soil and debris. The arthropods were hand-picked from the grids and transferred to 70% alcohol. All arthropods, except Aphididae and Collembola (spring-tails) found within the 3 grids, were collected but animals passing through all grids were ignored. The sample content was subsequently identified under binocular microscopes at 5 - 40 x magnification.

Araneae (spiders), Chilopoda (centripedes), Dermaptera (earwigs), Diplopoda (millipeds), Ephemeroptera (mayflies), Isopoda (woodlice), Orthoptera (grasshoppers) Opiliones (harvestmen), Plecoptera (stone flies), Psocoptera (booklice) and Thricoptera (caddis flies) were not identified further. Hemiptera were identified to at least family. Lepidoptera were identified to at least superfamily. Diptera were divided into 6 groups: 1. Asilidae (Robber flies) & Empidoidea, 2. Bibionidae (bibionid flies), 3. Syrphidae (Hoover flies), 4. Tipulidae (Crane flies), 5. Other Brachycera/Cyclorrhapha and 6. Other Nematocera. Neuroptera (lacewings) were identified to family whereas for Hymenoptera, parasitic wasps were identified as a group and the rest identified to family. Coleoptera were all identified at least to family but Carabidae imagines were identified to genus or species. All arthropods were separated into developmental stages except Aranae, Chilopoda, Dermaptera, Diplopoda and Opiliones for which life stage is not easily identified.

Collembola, an important detritivorous group of prey was not included because they require a comprehensive soil-sampling program in order to obtain reliable population estimates. Such a program was beyond the frame of this project.

Two variables were used to describe the data: 1. dry mass and 2. number of individuals. The dry mass is mainly important for arthropod – bird relations, and the numbers are mainly important when evaluating populations and predator – prey relations. Each type of data was analysed separately.

Dry mass was used as a measure of the available amount of arthropod food for birds. It is a variable relatively easy to measure making it widely used. Dry mass, however, does not take into account that the actual food quality may vary between the different arthropods. E.g., the food quality of carabids may be smaller than that of butterfly larvae due to their relatively higher cuticle content, which is of minor nutritious value.

Arthropod dry mass was estimated from the formula W = 0.0305 x L^2.62 mg, where L is the length of the arthropod in mm (Rogers et al. 1977). When the arthropods were identified to species the length of adults used for the calculations was obtained from the literature (Fauna Entomologica Scandinavica) using the mean of the length intervals given. For the remaining arthropods, including juvenile stages, the length of the arthropods was obtained by measuring between 50 and 100 individuals of each relevant group (but sometimes less if occurrence was rare). Because the data were reasonable symmetrically distributed, simple arithmetic mean lengths were calculated. The individuals measured were selected randomly from all farms, crops, sampling dates and dosages. A restriction was that a maximum of 5 individuals from each relevant arthropod group was taken from one field at a certain sampling date, to ensure a representative length estimate, especially for juvenile life stages, which change in size over time. Changes in the mass of a given species because of change in life stage during a season were consequently not included in the analyses.

All arthropods are not equally important as food items for all bird species. Some species may be characterised as "important food items" because of the selective food choice of farmland birds. Their preferences are probably based on the abundance, size, availability, nutritious value and (lack of) defence/escape mechanisms of the arthropods. Skylark was the most common bird throughout the study period (Chapter 4) and several studies on its diet have been conducted, e.g. Elmegaard et al. (1994 , 1999). The food choice of this species was therefore selected as reference bird prey. The diet of Skylarks listed by Cramp & Simmons (1977-94) comprise almost all arthropods found, including even very small species like Collembola and small Ichneumonoidea (ichneumonid wasps). Such small species were not found to be relevant in a comprehensive study in Denmark of Skylark food references, because of their limited overall energy contribution (Elmegaard et al. 1999). They found that Carabidae dominated, accounting for 42% of the estimated food dry mass in faecal pellets. Lepidoptera imagines and larvae contributed with 19% and Heteroptera (bugs) 7%. The rest of the groups contributed less than 5% to the total food dry mass including Diptera, which represented < 3% in Elmegaard et al.’s results. No findings of Collembola or smaller Hymenoptera were reported although some may have been included in the < 2% "rest" group.

There is apparently no distinct line between "preferred" and "non-preferred" arthropods, making the "important food items" variable rather arbitrary. Focus, however, was put on Danish research with experiments by Elmegaard et al. (1999) as the main reference. The list of relevant arthropod food items selected in this experiment comprised Araneae and Coleoptera except Coccinellidae and Cantharidae. Others taxa included were Chilopoda, Dermaptera and Diplopoda. Among Diptera only Bibionidae and Tipulidae were relevant. Also Ephemeroptera and Hemiptera were included but among Hymenoptera only Symphyta were relevant. Finally, Lepidoptera, Orthoptera, Opiliones, Plecoptera and Trichoptera, were all on the list. Neuroptera were never identified in faecal pellets by Elmegaard and was therefore not included.

The experimental unit was a dosage plot (see chapter 1), each of which was represented by 15 samples per sampling date. These 15 samples were summarised to form the dependent variables "mg food item dry mass / 3 m²" or "mg total dry mass / 3m^2" in the cereals (3 m² equals 15 samples of 0.2 m²). In beets the sample area per plot was 1.5 m² only. After summation, the data were log_e(x+1) transformed to stabilise the variation. As described in 4.2.1.2., repeated sampling in the same plots may violate the required independence of data. To avoid this, a geometric mean of the data collected during the actual period (entire season, before/after insecticide spraying or other relevant periods within the sampling seasons) was used, leaving only one figure per combination of the variables period, farm, crop and dosage.

General Linear Models (GLM) (SAS Institute 1990) were used for the analyses of dry mass (mg food item dry mass or total dry mass / 3 m²) (1.5 m² in beets), in order to determine a possible general effect of reduced dosages of pesticides on arthropods as a food resource. The dry mass was analysed in relation to the three class variables: year, farm, dosage as well as the interactions dosage´ farm and dosage´ year and the two numerical variables comprising the normal dose treatment intensity indices for insecticides (I-index) and herbicides (H-index). The model was extended with H-index and I-index (see Appendices A.2-A.4) to take into account that normal, half and quarter dosages were not reflecting uniform dosages between years and farms. Contradictory to the birds, arthropods may be directly affected by the actual dosage. By adding the treatment intensity indices the models were generally improved, as reflected by generally lower p-values of the models. Since sampling was restricted to the main crop area, it was not necessary to consider possible edge effects in the analyses.

In the population studies, data were separated into carnivore and non-carnivore groups, which were again separated into more specific taxonomic levels. If the juvenile stage was carnivorous, imagines were also recorded as carnivores e.g. Syrphidae. Overall the data sets were constructed as described above under dry mass analyses. The dependent variables in this case, however, were numbers of individuals / 3 m². Data of many insect populations follow a Poisson distribution. To analyse such data, Generalised Linear Models (GENMOD) (SAS Institute 1990) with Likelihood ratio tests on log_e(x+1) transformed data were used for the population analyses. The procedure can also be used for those arthropod groups not following a Poisson distribution since the difference between a Normal distribution and a Poisson distribution after log_e(x+1) transformation with adjustment for over dispersion is limited. Therefore this procedure was used throughout the population analyses to make them comparable. For each analysis least squares means were estimated. When a significant dosage effect was found, t-tests were used to interpret pairwise dosage-differences.

A non-parametric test was performed to reveal if there were a significantly higher number of the most abundant arthropod groups improving under a reduced pesticide regime. The arthropod groups were divided into the six superior groups: Barley carnivores (n = 18), Barley non-carnivores (n = 17), Wheat carnivores (n = 18), Wheat non-carnivores (n = 18), Beets carnivores (n = 19) and Beets non-carnivores (n = 17)) (see Table 3.8-10). A Friedman test based on ranked data was conducted for each the superior groups using the FREQ procedure in SAS/STAT (SAS Institute 1990).

Carabids contribute significantly to the arthropod fauna. They are important both as beneficials (Lövei & Sunderland 1996) and as food items for birds (Elmegaard et al. 1999) making it important to estimate their density and therefore to sample them efficiently. Especially larger species, however, are not easy to sample by suction since they may be nocturnal and concealed in refugia during the day. Furthermore they may occur at very low densities at the soil surface (Lövei & Sunderland 1996).

Pitfalls where chosen because they are very suitable for catching carabids. Due to resource aspects, sampling was limited to winter wheat on three farms: Gjorslev, Oremandsgård and Lekkende.

In 1998 and 1999 enclosed plots were established by surrounding 10 x 10 m areas with 60 cm high metal plates buried 20 cm into the ground. The barriers made it possible to obtain estimates of the actual density of carabids in the sampling period by catching nearly all carabids within the enclosures, which at the same time secured that no ground beetles from outside of barriers could reach the traps. 4 enclosures per plot were established at least 25 m from the plot edges in normal and quarter dosages treatments. The enclosures were equally spaced along a longitudinal gradient in an attempt to minimise variation. Also the barriers were placed at the same distance from field margins if possible to maximise comparability. The establishment in normal and quarter dosages only was due to resource limitations. The enclosures were established 18 – 26 May 1998 and 12 – 20 May 1999. At this time of year it was assumed that field invasion from the wintering sites was completed.

Nine pitfalls were placed within each enclosure in a way that aimed at ensuring maximum catches. Five were placed in the middle in the same pattern as number 5 on a dice separated by four 100 cm x 15 cm metal plates positioned between each peripheral trap and the central trap, a method that increase trapping efficiency. Furthermore one pitfall was placed in each corner of the enclosures since carabids are known to follow vertical edges. In 1998 the weekly samplings were carried out in the periods 1 July to 7 August at Gjorslev and Oremandsgård and 17 June to 7 August at Lekkende. In 1999 the period was 23 June to 3 August at all 3 farms. Furthermore a pre-insecticide sampling was conducted in both years in mid June. After the five-day pre-insecticide sampling period, the pitfalls were closed with lids to temporary stop further sampling within the enclosures. One week after insecticide spraying the lids were removed again and sampling continued non-stop to the end of the season.

Catches were stored in glass containers with 70% alcohol until further treatment. In the laboratory adult carabids were counted and identified at least to genus.

The pre-insecticide catches were generally low and scattered, which weakened the possibility of making a reliable analysis on the pre-insecticide application data considering the low number of replicates. These data were therefore pooled with the post-insecticide spraying samples. Two variables were used to describe the data: 1. total dry mass and 2. number per genus. Due to competition among species of the carabid-family it seemed of less value to analyse the total number of carabids caught per barrier.

As described for the suction samples (see 3.2.1.1) the dry mass was estimated using the length of the beetles. Most lengths were obtained from the literature (Lindroth 1985 & 1986) but some beetle groups (Bembidion, Trechus and Amara) were identified to genus only. For those groups an arithmetic mean of the length of at least 50 randomly selected individuals per group were used for the dry mass estimations.

The experimental unit was the enclosure in which catches were collected 5-7 times during each season. These 7 samples were summarised to form the dependent variables "mg dry mass / 100 m²" and "number per genus / 100 m²" to avoid dependent data. Summation was followed by a log_e(x+1) transformation to stabilise the variance.

General Linear Models (GLM, SAS/STAT) (SAS Institute 1990) were used for both the dry mass analyses and the population analyses in order to determine a possible general effect of reduced dosages of pesticides since the log-transformed data were assumed following a normal distribution. The dependent variables were analysed in relation to the 3 class variables: year, farm and dosage as well as the interactions dosage´ farm and dosage´ year. Stepwise model reduction was used. For each analysis least squares means for dry mass and population size were estimated.

G. polygoni (knotgrass beetles) is of particular interest because it is important for farmland birds, e.g. partridge (Perdix perdix) and it is a potential control agent for its host plants Polygonum convolvulus and Polygonum aviculare (Sotherton & Moreby 1992).

At Gjorslev a high numbers of G. polygoni - adults and larvae - were caught in the fenced pitfalls in 1999. The number of adults within each enclosure was counted in order to reveal a dosage effect on the beetles. The statistical analysis was conducted as described above, but the dependent variable "no. / 100 m²" was analysed in relation to the class variable dosage only.

The aphid counting was conducted as described in section 3.1.1, with the exception that 100 beets plants were inspected instead of 50 on each assessment day. If possible the first inspection was carried out immediately before insecticide application and the following about one week after application to get the full effect of the application. The overall aim was not to estimate crop damage, but to reveal possible effects on the populations.

Statistical analyses were not conducted on these data. Percent infested cereal ears / beets plants are presented in Appendix D.

In 1997 a severe attack of Miridae (mirid bugs) occurred in sugar beets on Gjorslev. The attack was restricted to the border zone. Two days after insecticide application 40 samples per plot were taken to evaluate the effects of reduced dosages. A sample comprised 10 standardised sweeps with a butterfly net (diam. 36 cm) with one sweep pr. row in the 10 outermost rows.

The basic sampling unit was considered to be one sample. The data followed a Poisson distribution and a log-linear model was fitted, using the GENMOD procedure. Log Likelihood ratio tests were used to estimate the difference in numbers per sample between treatments. Since sampling was carried out in one field and one year only, the replicates (n = 40) were all taken from the same dosage plot and should therefore be considered as pseudo-replicates, weakening the possibility to generalise the result.

3.2.2 Results

The results of the analyses of variance are summarised in Table 3.4. The factors year and farm always had significant effects and were consequently included in all the models. The factors dosage and I-index were significant in barley. In wheat and beets no significant effects of dosage was found, and the treatment intensity indices for insecticides and herbicides only occasionally proved significant which seemed incidental. With none of the tested models significant interactions were found.

The factors year and farm constituted, not surprisingly, an absolutely dominating part of the variation, leaving only a minor part to be explained by the factor dosage. Especially in the cereals year was dominating.

Table 3.4.
Schematic summary of the dry mass analyses based on dry mass means of the entire sampling season. I-index / H-index. (treatment intensity indices for insecticides / herbicides, for definition see section 1.1). Statistical significance is indicated as follows: */+/–: 0.01<p<0.05, **/++/– –: 0.001<p<0.01, ***/+++/– – –: p<0.001. A +/¸ indicates if the correlation is positive/negative.

The estimated mean dry masses are presented in Table 3.5. The estimates are means over the entire season.

Significant differences of the estimated mean dry masses between dosages were found only in barley with higher dry mass at quarter dosage than in normal and half dosages that mutually were not significant different. A 30% higher dry mass of food items was found at quarter dosage than in normal dosage. A corresponding 28% difference was found for the total dry mass.

Table 3.5.
The estimated mean dry masses of relevant bird prey (based on known skylark prey (Elmegaard et al. 1999) and the estimated mean dry masses of all arthropods collected in the three different crops. The numbers given are least squares estimates of the mean dry mass, of 5-7 samples per year in the period late May - mid August. In cereals the estimates are "mg/3 m²" and in beets "mg/1.5 m² plants". The estimates are followed by 95% confidence intervals. P-values for test of significance of the factor "dosage" are given (**: p<0.01). Significant differences (p<0.05) between the different dosages are found by paired t-tests and are indicated by different letters.

Look here!

Even though no significant differences between dosages were found in wheat and beets, there was a tendency towards higher dry masses at reduced dosages. In beets there was a 13% higher estimated food item dry mass at quarter dosage than in normal dosage. In wheat a correspondingly 7% higher dry mass was indicated. Similarly in wheat a 17% higher total biomass was revealed at quarter dosage than in normal dosage and in beets a 16% higher total dry mass. In barley and beets the estimated dry masses at half dosage were closer to the normal dosage estimates than to quarter dosage but in wheat the estimated dry mass at half dosage was closer to the dry mass at quarter dosage.

To reveal possible dosage effects limited to the post insecticide treatment period, statistical tests on the wheat data were conducted on the biomass data limited to the whole period after insecticide applications. The farm and year proved significant and no significant interactions were found. A test for a dosage effect on the bird prey biomass did not show a significant effect (p = 0.4973) (estimated means and 95% confidence intervals: Normal = 410.0 [ 339.4; 495.3] mg/3 m², half = 459.7 [ 380.6; 555.3] mg/3 m², quarter = 475.6 [ 393.7; 574.5] mg/3 m²). A similar test on the total biomass did neither reveal significant differences (p = 0.2354) (normal = 929.6 [ 790.4; 1093.2] mg/3 m², half = 1047.7 [ 890.9; 1232.2] mg/3 m², quarter = 1129.1 [ 960.1; 1327.9] mg/3 m²). Statistical tests for the 14 days period after insecticide applications did not show significant dosage effects on bird prey biomass (p = 0.5649) (normal = 458.7 [ 363.2; 579.3] mg/3 m², half = 514.0 [ 407.0; 649.1] mg/3 m², quarter = 543.8 [ 430.6; 686.6] mg/3 m²) or on total biomass (p = 0.6953) (normal = 1118.0 [ 933.3; 1339.4] mg/3 m², half = 1179.3 [ 984.4; 1412.7] mg/3 m², quarter = 1243.7 [ 1038.2; 1489.9] mg/3 m²).

In beets, tests for the corresponding periods after insecticide spraying were performed but only comprising data from farms which carried out a mid-summer insecticide application. Due to the lower number of replicates the treatment intensity indices for herbicides and insecticides were not included in the model. In all the models tested, the farm and year had a significant impact but no significant interactions were found. For the entire post-insecticide period, a nearly significant dosage effect (p = 0.0755) was found on the bird prey dry mass (estimated means and 95% confidence intervals: Normal = 220.9 [ 181.7; 268.7] mg/3 m², half = 239.0 [ 196.5; 290.7] mg/3 m², quarter = 288.5 [ 237.2; 350.8] mg/3 m²). The test for a dosage effect of the total biomass showed significance (p = 0.0273^*) (normal = 378.5 [ 298.2; 480.4] mg/3 m², half = 406.0 [ 319.9; 515.4] mg/3 m², quarter = 555.5 [ 437.6; 723.5] mg/3 m²). Pairwise t-tests showed significant differences between normal and quarter dosage and between normal and half dosage. Tests for the 14-day period only after insecticide spraying did not reveal a significant dosage effect on bird prey biomass (p = 0.2445) (normal = 158.7 [ 123.5; 203.9] mg/3 m², half = 175.6 [ 136.7; 225.5] mg / 3 m², quarter = 203.2 [ 158.2; 261.0] mg/3 m²) but a significant effect was found on the total biomass (p = 0.0130^*) (normal = 209.1 [ 157.6; 277.5] mg/3 m², half = 237.3 [ 178.8; 314.9] mg/3 m², quarter = 353.1 [ 266.1; 468.5] mg/3 m²). Pairwise t-tests showed significantly higher dry masses at quarter and half than at normal dosage.

The generally significant effect of the dominating factor year (table 3.4) may to a large extent be explained by climatic factors, which are important for arthropod population sizes. Also different spraying intensity between years and a possible accumulating pesticide effect through the experimental years, may have contributed to the significant effect. Estimates of the parameter differences of the factor year are presented in table 3.6.

Table 3.6.
Parameter estimates of the year-differences of the bird prey dry mass analyses.

From table 3.6 it appears clearly that the estimated arthropod dry mass in the cereals in 1997 was overall low. This fact may be the main reason behind the significant effect of year in the analyses of cereals mentioned above the table. In beets the highest estimated arthropod dry mass was in 1998 followed by 1997 and 1999.

To evaluate the development in the amount of available important bird prey and the total arthropod dry mass during the sampling season, the relevant dry masses were estimated in period intervals for each of the three crops. Results are presented in Fig. 3.3 and 3.4. In barley and wheat the amount of important food items remained stable with a tendency to increase during the sampling season and peaking in July. In beets, however, there was a substantial and steady increase in the amount of available prey during the season. In barley there was a tendency towards higher bird prey biomass in quarter pesticide dosage throughout the season. After the period of insecticide treatments there was a higher dry mass at half than at normal dosage. In wheat there was a higher bird prey dry mass at half dosage in the first period, mid May to mid June, before insecticide spraying. During the rest of the season more biomass was found in quarter dosage followed by half and normal dosage. In beets there were only indications of a difference in the last period, 1 - 20 August, with more bird prey biomass found at quarter than at half and normal dosage.

Fig. 3.3.
The development in the amount of important arthropod food for birds in the three different crops. The shaded areas indicate periods of insecticide treatments. Notice that data in barley and wheat are per 3 m², whereas data in beets are per 1.5 m² (75 plants). The error bars indicate 95% confidence intervals.

Fig. 3.4.
The development in the total arthropod dry mass in the three different crops. The shaded areas indicate periods of insecticide treatments. Notice that data in barley and wheat are per 3 m², whereas data in beets are per 1.5 m² (75 plants) The error bars indicate 95% confidence intervals.

Overall the total dry mass of arthropods was roughly about twice the amount of the bird prey dry mass. Generally, the development during the season of the total dry mass was similar to bird prey dry mass with two exceptions: 1. The increase during the season in the total dry mass in wheat was much steeper compared to the bird prey dry mass. 2. Contrary to the bird prey dry mass in beets, there seemed to be a higher total dry mass at reduced dosages for the period 16 – 31 July.

The composition of the biomass, including estimates and statistical tests for dosage effects on the most dominant orders is presented in Table 3.7. In the cereals the dominant order across dosages were Coleoptera (about 50%) followed by Diptera (30-40%) whereas the rest of the abundant orders were more evenly distributed. In beets Diptera dominated with about 50% of the dry mass and Coleoptera with 20%. Lepidoptera contributed here 10%, which was remarkably higher than in the cereals. Significant differences between dosages were found in barley for the predominant taxa Coleoptera and the undefined rest group "others". There was, however, in all crops and for all orders except Araneae a general trend towards higher biomass in quarter dosage compared to normal dosage. The dry mass estimates of half dosage did not reveal a uniform trend.

Table 3.7.
Estimated dry masses of arthropods separated into the most important orders. The numbers given are least squares estimates of the mean dry mass of 5-7 samples per year in the period late May - mid August. In cereals the estimates are "mg/3 m²" and in beets "mg/1.5 m²". The estimates are followed by 95% confidence intervals in square brackets. P-values for test of significance of the factor "dosage" are given (*: p<0.0.5, **: p<0.01, ***: p<0.001). Significant differences (p<0.05) between the different dosages are found by paired t-tests and are indicated by different letters.

Look here!

In Tables 3.8 – 3.10 density estimates of the most common arthropod groups are presented. The results of the analyses of variance of each factor for each arthropod group except dosage are excluded, since it would be too comprehensive to present. Other arthropod groups than those listed were found but their relevance to this study were considered minor, or their abundance was too low to be relevant. Aphididae are considered the most important insect pest and special emphasis will be put on their predators. Polyphagous predators like Carabidae, Staphylinidae and Araneae have often been considered of special importance because they are abundant throughout the season. Emphasis will also be put on the specialised aphid predators Syrphidae, Chrysopidae and Coccinellidae, as well as on species considered of special importance as food items. Regarding the non-carnivores, focus will be on the potential pests as well as on groups considered important food items for birds.

Table 3.8.
Barley. Mean densities of the most common carnivore and non-carnivore groups (no./3 m²). Estimates given are least squares means with 95% confidence limits in square brackets per sampling. Significant differences between dosages (p<0.05, paired t-tests) are indicated by different letters. P-values for test of the factor dosage are given (ns: P>0.05, *: p<0.05, **: p<0.01, ***: p<0.001). Abbreviations: Col.= Coleoptera, Dip.= Diptera, Hem.= Hemiptera, Hym.= Hymenoptera, Neu.= Neuoptera, Img.= Imagines, Lar.= Larvae, Nym.= Nymphs, Pup.= Pupae.

Look here!

Results of the population analyses in barley are presented in Table 3.8. Dosage had a significant effect on nine groups, of which seven were juvenile groups. Dermaptera, which was significantly affected by dosage, was not divided into imagines and juveniles and the only group purely of imagines, which responded significantly to dosage, was "other dipterans". The groups showing significant responses were almost evenly divided between carnivores and non-carnivores. It is noticeable that all aphid-specific juvenile predator groups Coccinellidae, Syrphidae and Chrysopidae were significantly affected by dosage with the highest densities at quarter dosage.

The population analyses of carnivores in wheat (Table 3.9) revealed results quite similar to those found in barley. The two aphid specific predators Coccinellidae and Syrphidae responded significantly to dosage but in contrast to barley the significant effect was for adult Coccinellidae. The estimates for the larvae, however, also indicated an effect with twice as high estimates at quarter than in normal dosage. Also for Dermaptera there was an effect in wheat whereas contra to barley a dosage effect was found for Carabidae instead of Staphylinidae.

Table 3.9.
Wheat. Mean densities of the most common carnivore and non-carnivore groups (no./3 m²). Estimates given are least squares means with 95% confidence limits in square brackets per sampling. Significant differences between dosages (p<0.05, paired t-tests) are indicated by different letters. P-values for test of the factor dosage are given (ns: p>0.05, *: p<0.05, **: p<0.01, ***: p<0.001). Abbreviations: Col.= Coleoptera, Dip.= Diptera, Hem.= Hemiptera, Hym.= Hymenoptera, Neu.= Neuoptera, Img.= Imagines, Lar.= Larvae, Nym.= Nymphs, Pup.= Pupae.

Look here!

An equal number of carnivore and non-carnivore groups revealed a significant dosage effect. The dosage effects on non-carnivores in wheat were on other groups than in barley. In wheat significant dosage effect were found for larvae/pupae of Diptera, Auchenorrhyncha and adults and nymphs of Miridae, the last one considered important bird prey.

Table 3.10.
Beets. Mean densities of the most common carnivore and non-carnivore groups (no./1.5 m²). Estimates given are least squares means with 95% confidence limits in square brackets per sampling. Significant differences between dosages (p<0.05, paired t-tests) are indicated by different letters. P-values for test of the factor dosage are given (ns: p>0.05, *: p<0.05, **: p<0.01, ***: p<0.001). Abbreviations: Col.= Coleoptera, Dip.= Diptera, Hem.= Hemiptera, Hym.= Hymenoptera, Neu.= Neuoptera, Img.= Imagines, Lar.= Larvae, Nym.= Nymphs, Pup.= Pupae.

Look here!

In beets (Table 3.10) the number of groups significantly affected was about evenly distributed between carnivores (5) and non-carnivores (4). The carnivore group comprised the generalists Staphylinidae and Dermaptera as well as the aphid specific specialists, Coccinellidae and parasitoid wasps, which were a broad ranged group comprising a lot of specialists. The non-carnivores affected significantly by dosage were all true herbivores: Adult Curculionidae, Meligethes (pollen beetles), Miridae and Auchenorrhyncha, all considered relevant bird prey.

A non-parametric Friedman test confirmed that a significantly higher number of the most common carnivore groups did improve at quarter dosage in all three crops (Table 3.11). Regarding the non-carnivores, significantly more groups benefited from quarter dosages in barley and beets, but significant differences were not found in wheat. Half dosage did not show any uniform pattern. For the groups "barley non-carnivores", "wheat carnivores" and "beets non-carnivores" the effect of half dosage was in between quarter and full dosage without being significantly different. For "barley carnivores" and "beet carnivores" the effect of half was nearer normal dosage.

Table 3.11.
Results of a non-parametric test performed on the data in Tables 3.8 - 10 to elucidate if a significant number of arthropod groups benefited from reduced dosages of pesticide applications. P-values for test of the factor dosage are given (ns: P>0.05, *: P<0.05, **: P<0.01, ***: P<0.001). Pairwise tests were used to reveal significant differences (p<0.05) between dosages.

The estimated total carabid dry mass differed significantly between quarter and normal dosages (Table 3.12). The difference in dry mass was about 25%. As for the suction samples the factors farm and especially year constituted a dominating par of the variation. A significantly higher carabid dry mass was found in 1999 accordingly to the parameter estimates (not presented).

Table 3.12.
Estimated total dry mass of adult carabids caught in 100 m² enclosures in winter wheat in the periods 1/7- 7/8 1998 and 31/5-3/8 1999. Estimates given are least squares means with 95% confidence limits in square brackets per sampling. P-value for test of the factor dosage is given (*: p<0.05).

The genus Pterostichus dominated across dosages, contributing about 40% to the total dry mass of adult carabids (Fig. 3.5). Also Loricera, Calathus, Agonum and Harpalus contributed significantly to the carabid biomass.

Fig. 3.5.
The composition (by dry mass) of the carabid beetle fauna caught in fenced pitfalls 1998-99.

In table 3.13 the results of statistical analyses of possible dosage effects on the most abundant genera are presented. The results of the analyses of variance of each factor for each group except dosage are excluded, since it would be to too comprehensive to present. The population of the larger carabid Pterostichus increased significantly at reduced pesticide applications but the two genera Bembidion and Synuchus were significantly more abundant at normal dosage. Two other genera, Loricera and Demetrias, also seemed affected by dosage. The populations of both seemed to increase at reduced dosages. The conclusion of these two genera, however, was complicated by the significant interaction farm´ dosage, which revealed that a dosage effect was not found on all farms.

Juvenile stages may be more sensitive than the adult individuals. It was, however, not possible to count all larvae caught in the pit-falls but a pesticide effect on the larva, whether direct (lethal) or indirect (sublethal, changed microclimate or altered food supply), was fund by suction sampling in wheat (Table 3.9).

Table 3.13.
Mean numbers of carabids caught in pit-falls within 100 m² enclosures in wheat in the periods 16/6-5/8 1998 and 31/5-3/8 1999. Estimates given are least squares means with 95% confidence limits in square brackets per sampling. P-values for test of dosage are given (ns: p>0.05, *: p<0.05, **: p<0.01, ***: p<0.001).

A significant dosage effect was found on the number of Gastrophysa polygoni (chrysomelid beetles) caught in pitfalls (Table 3.14). Least squares estimates revealed much higher catches at quarter dosage than at normal dosage. It is unclear whether the effect was due to reduced insecticide spraying or to higher occurrence of its host plants Polygonum convolvus or P. aviculare or to a combination of various factors.

Table 3.14.
Estimated numbers of Gastrophysa polygoni adults (total no./100m²) in wheat on Gjorslev 1999. Estimates given are least squares means with 95% confidence limits in square brackets per sampling. P-value for test of the factor dosage is given (*: p<0.05).

Generally a dosage effect on Aphididae was found in all three crops. The aphid specific insecticide Pirimor (pirimicarb) proved more effective than pyrethoids and Dimethoate. The efficiency of the insecticide applications, however, was highly variable. A table with results is presented in Appendix D.

The results are presented in Table 3.15, which shows a significantly higher occurrence of Miridae in quarter dosage. Since the estimates are not absolute due to the sweep net sampling method, it may be more relevant to look at the ratios between the estimated numbers. The results showed that it could be expected to find between 3 – 17 times more Miridae at quarter dosage compared to normal dosage, and between 2 - 8 times more in half compared to normal dosage.

Table 3.15.
The estimated number of Miridae (Mirid-bugs) per sample (10 standardised sweeps in the outer 10 rows) after insecticide application in beets, Gjorslev 11 July 1997. Estimates given are least squares means with 95% confidence limits in square brackets per sampling. Significant differences between dosages (p<0.05, paired t-tests) are indicated by different letters. The p-value for test of the factor dosage is given (***: p< 0.001).

3.2.3 Discussion

In a tri-trophic context the insect part had a dual aim. One was to research if, and to what extent, reduced dosages of pesticides, insecticides and herbicides, affected the amount of available arthropod food for the farmland birds. The other was to explore if and how much pesticides affected the populations of specific taxa of arthropods. Of special importance were populations of "beneficials", especially predators of crop pests also being important food items for birds. Dosage effects on the most important crop pests, Aphididae (aphids), were roughly estimated by counting tillers/plants with aphids. The overall aim was not to estimate crop damage, but only to reveal effects on the populations.

Possible pesticides effects could be either direct (lethal) or indirect (sublethal, changed microclimate or altered food supply). Due to the complexity of this experiment and the complexity in general it is complicated (if at all possible) to reveal the relative importance of the actual mechanism(s) causing significant findings. However, the most likely causes for effects found will briefly be discussed here and more deeply in chapter 7 and further analyses of correlation between arthropods and weed are presented and discussed in chapter 6 and 7.

Overall there was a general tendency towards more arthropod biomass at reduced dosages of pesticides (Table 3.5). There was considerable difference between the findings in the three experimental crops. In barley, a significantly higher dry mass was revealed at quarter dosage than at half and normal dosages. In wheat and beet no overall significant differences between dosages were found, but in beets a higher total arthropod dry mass was revealed at the reduced dosages after insecticide application. In barley a 30% higher total dry mass and food item dry mass was estimated between quarter and normal dosages. In wheat and beet the corresponding differences indicated were never more than about the half of that in barley. The dry mass estimates for half dosage was mostly in between the two other dosages but sometimes the estimate was nearer quarter and other times it was closer to normal dosage.

Possible reasons for the pronounced dosage effects found in barley were, that the insecticides were applied earlier and barley was a more open crop compared to wheat, allowing pesticides to penetrate deeper into the canopy thereby improving their effects (see also 7.2). Furthermore insecticides were applied more often in barley than in beets and broader ranged products were used in barley. In beets, weed hoeing was always conducted at half and quarter dosage. At Oremandsgård and Gjorslev weed hoeing in normal dosage plots was carried out once per season irrespective of the number of herbicide applications. Nordfeld had done similarly at one instance (1998), while the farms Lekkende and Nøbøllegård have never used weed hoeing in normal dosage plots (see Appendix B). Generally it may be assumed that soil-tilling has a negative impact on arthropods (Holm et al. unpubl.). Reasons for this could be disturbance and altered micro-climate, maybe shading the effects of reduced pesticide dosages.

Wheat had the highest arthropod biomass followed by that of barley and beets. A possible reason could be, that winter wheat is an early established, higher and denser crop probably providing a more favourable environment throughout the season. Furthermore, in wheat no soil tilling was conducted in the spring probably in favour of especially soil-dwelling arthropods. In beets the arthropod dry masses were always lower during the season when comparing with the corresponding periods in the cereals; especially at the beginning of the season. This is most likely due to the canopy development, which affects the microclimate. In beets, the long period of bare soil in early half of the season creates a rather harsh microclimate, which however changes with ongoing crop development towards being shadowy and humid. As mentioned for wheat, in winter cereals, the less extreme conditions already established in early spring creates more favourable conditions for most relevant arthropods. It should be noticed, that the arthropod estimates for beets, which had the lowest dry mass estimates until the end of the season, actually would be lower if sampling had covered not only the crop plants but also the almost bare field between the rows.

When comparing the dry mass fluctuation between years in beets with the climatic data presented in section 1.2.8 it seems likely that the relatively high precipitation in July 1998 had benefited the arthropod populations. In the cereals, in which arthropod populations are established earlier due to the crop phenology, the relatively cold May in 1997 may have suppressed the populations permanently that year. It is also possible that the relatively higher catches in the cereals in 1998 and 1999 are due to an accumulated effect of reduced pesticide dosages. It is, however, not possible to analyse such an effect isolated. Between-year differences of the amount of pesticides applied are limited and do apparently not explain the fluctuations (Table 1.1, Appendix A.2-A.4).

A non-parametric test (Friedman test) confirmed that numbers of the most common arthropod groups did increase under a reduced pesticide regime, but with the group "wheat non-carnivores" as an exception. There was a clear effect of quarter dosage, whereas there was no general effect of half dosage.

There was a tendency towards, that most affected arthropod carnivores in all three crops were aphid specific, often at juvenile stages. The populations of Dermaptera (earwigs) were higher at quarter dosage in all three crops, and they are also known as important aphid predators (Sunderland & Vickerman 1980). It is possible that this was due to prey removal, rather than a direct lethal effect on the predators. On the other hand the specific aphid predators are very exposed to insecticides due to their location high in the canopy. Furthermore the juvenile stages have limited mobility making them good indicators of pesticide effects compared to the often highly mobile adults having the ability to re-colonise quickly. The increase of the aphid specific predators responding significantly to the reduced dosages of pesticides was in the range of 20% - 175%, most pronounced for Coccinellidae larvae in barley. The populations, however, were probably still too low to have a significant impact on the aphid populations. Among the non-carnivores, it was generally not the same non-carnivore groups, which were significantly affected by dosage in the three crops, however the number of groups significantly affected was the same (4).

A dosage effect on Aphididae was found in al three crops. Aphids are considered the most important crop pests and they are the main targets of a majority of the insecticide applications. The aphid specific insecticide Pirimicarb, which is considered less harmful to most arthropod predators, proved more effective than pyrethoids and Dimethoate in all three crops. The absolutely lowest damage threshold in barley and wheat is a 30% ear infestation at the most vulnerable growth stages (Nielsen et al. 2000). Therefore, the insecticide applications in both cereals in 1997 and in wheat in 1998 could probably have been omitted. The other insecticide sprayings in the cereals seemed justified. Despite a high variation in the efficiency of the applications, which blurs the overall picture, quarter dosage seemed close to the required minimum. In beets all the Pirimicarb applications proved efficient, even at quarter dosage, contrary to the other insecticides.

With fenced pitfalls a significantly higher estimated dry mass (25%) of the important carabids was found at quarter dosage in wheat. The results obtained on carabids using fenced pitfalls are not reflected in the suction samples probably because carabids are poorly extracted by suction. It was not possible to conduct corresponding experiments in barley and beets due to resource limitations (and the ongoing weed hoeing in beets). It is, however, most possible that the results obtained in wheat could be found in barley and beets too. This is an interesting hypothesis since Carabidae is a very dominating family within Coleoptera, which already constitute a significant part of the dry mass of the suctions samples. A significant effect for carabids in barley and beets could therefore turn the overall tendency even more towards a clear dosage effect. The most abundant genus Pterostichus responded positively to reduced dosages with 62% higher density at quarter dosage. The most abundant Pterostichus species was P. malanarius which is a medium to large sized species. It is a widely studied species, known as an important predator of many crop pests.

Because the pre-insecticide catches were insufficient for statistical analysis, it was not possible to distinguish between insecticide or herbicide effects on the adult populations. Furthermore the life cycle of carabids, with larval stages in the soil having different emergence periods and consequently population fluctuations difficult to access, complicates the conclusions. However, since larger carabids generally are not very sensitive to insecticides at the applied dosages, weed cover may play a key role in the differences found between dosages. P. melanarius is nocturnal and prefers probably a dense plant cover as found at quarter dosage, whereas e.g. the most abundant Bembidion species are diurnal and may therefore prefer the less dense plant cover found at normal dosage (see 7.2). The pesticide effects found on the adult beetles may also be due to lethal effects on the larvae, especially on those with a pronounced epigaeic activity. A species, which apparently was affected by a differentiated spraying regime, was Loricera pilicornis, which has epigaeic activity during the period of insecticide spraying (Traugott 1998). It is possible that the significantly higher catches in 1999 were due to an accumulated effect of reduced dosages, but it cannot be documented statistically.

The significantly higher numbers of Gastrophysa polygoni (knotgrass beetle) at quarter dosage (estimated number 15 – 16 times higher, but with high variation) on Gjorslev 1999 is in line with the results of Kjær & Jepson (1995) who found increased populations at reduced field rates of dimethoate. G. polygoni is found on the aerial parts of its host plants and is therefore directly exposed to the spraying droplets. This probably makes it very sensitive to the pyrethoid spraying (Tau-fluvalinat) conducted. It may also be because of an increased host plant resource although this was not verified. Also when herbicides do not kill the host plants, but only limit their growth and quality, the abundance of G. polygoni may be severely reduced (Sotherton 1982, Kjær & Elmegaard 1996).

Even isolated it is a very interesting case since earlier research have documented G. polygonum (as well as its host plants) as a key factor for the partridge Perdix perdix (Sotherton 1982).

4 Birds

(Petersen, B.S.)

The ornithological studies were aimed at describing the number of birds of different species utilizing the fields at different pesticide dosages. Although many factors may affect the distribution of birds in the agricultural landscape, it is believed that the number of birds occurring in a certain field, or dosage plot, is to a significant degree related to the amount of available food. The occurrence of birds in relation to the amount of animal and vegetable food resources is analysed and discussed in chapters 6 and 7.

The scale of the study and the resources available did not allow investigations of population size and dynamics – i.e., no attempts of measuring the absolute number of territories, production or survival rate were made. On the other hand, with three different crops, fifteen fields of >18 hectares, and three study years it is believed that the results are of considerable general value.

4.1 Pilot studies

The aim of the pilot studies was twofold: (1) to test and adjust the census methods in the field; (2) to collect data on the occurrence of birds on the study fields. The purpose of this initial collection of data was to check if the planned dimensioning of the study (number of fields, field size, sampling period, sampling frequency) was adequate to make the collection of a sufficiently large amount of data possible.

Two separate pilot studies were carried out. Firstly, a study was performed during the breeding season of 1996 on six fields at Gjorslev. Secondly, the occurrence of birds during the winter months was studied in the winter of 1996-97 on one field at each farm. The results of these investigations are briefly dealt with in the following. They do not contribute significantly to the elucidation of the main problem, but they provide documentation for the expedience of the methods used.

4.1.1 Breeding season 1996

During the breeding season, two investigations were conducted at Gjorslev. (1) On the three fields that were to be used in the main study (1997-1999), censuses were carried out with the purpose of providing data on the distribution of birds on the fields when these were homogeneously sprayed. This information could be important for the laying out of dosage plots as well as for the choice of census technique. (2) On the experimentally sprayed pilot test fields, a series of counts was carried out with the dual purpose of optimizing the census method and providing initial data on the possible differences in bird densities between dosage plots.

The three main study fields were sown with winter wheat, spring barley and maize, respectively, in the harvest year 1995-96. All three fields are surrounded by hedges (cf. Fig. 1.3), and the study was primarily designed to reveal the effect of these hedges on the distribution of birds on the fields. Using marker sticks, each field was divided into four zones: 0-12 m, 12-50 m, 50-100 m and > 100 m from nearest hedge; a 0-12 m zone was also marked around any habitat islands within the fields. Each field was then divided into 12 census plots of approximately equal size (1.5-2 ha). Point counts of 10 minutes duration were carried out from the edge of each census plot. All birds seen or heard within or immediately above the census plot were recorded and assigned to one or more zones. Two censuses of each field (one between 8 and 11 a.m., one between 12 and 15) were carried out on five dates between 22 May and 17 June 1996, yielding a total of 10 (not independent) counts per field. For analysis, the number of individuals of each species recorded within each zone was calculated for each of the 10 counts separately, by adding up the counts from the 12 census plots.

In the pilot test fields, four 1.5 ha subplots (census plots) were delimited within each dosage plot. This subplot size roughly corresponds to the area of a field which can be surveyed from one point. Ten minutes point counts were carried out from the edge of each subplot. All birds seen or heard within or immediately above the subplot were recorded and assigned to one or more of the following four zones: hedge/habitat island; field 0-12 m (from hedge/habitat island); field > 12 m; air. Two censuses of each field (one between 8 and 11, one between 12 and 15) were carried out on 14 dates between 1 May and 11 July 1996. At the latter date, the number of birds in the cereal fields had clearly peaked, and the censuses were stopped. In the beet field, however, numbers were still high, and four additional censuses were carried out between 18 July and 12 August.

The counts revealed that some birds were not recorded when the observer was stationary at the census points, but were flushed or became vocal when the observer moved between the points. Therefore, the point counts were supplemented by transect counts from 21 June onwards. Each route through a subplot (from one census point to the next) formed a transect, and transect time was standardized at 5 minutes. Only birds within the subplot were recorded. In the analyses only registrations from the census points (not from the transects) were included. For each dosage plot, the counts from the four subplots were added up separately for the morning and afternoon counts, and the maximum of the two counts was used.

The distribution of the seven most frequently recorded bird species on the homogeneously sprayed fields is shown in Table 4.1. There were no big differences between fields, so results from the three fields have been pooled in the presentation. It is clear from the table that all of the species are affected by the presence of hedgerows, either negatively (Skylark) or positively (the others). Because the census points were located along the hedges, the detection chance decreased (to an unknown degree) towards the mid-field. Thus, the Skylarks' avoidance of areas close to hedges is more pronounced than indicated by the data.

Table 4.1.
The distribution of selected bird species on homogeneously sprayed fields, classified according to distance to nearest hedge. The mean number of individuals per count is given. Percentages indicate the proportion of total field area belonging to each distance zone or (for the first three bird species) proportion of total number recorded within each of these zones.

Fig. 4.1.
Examples of results0 from the bird counts on the pilot test fields, summer 1996. The mean number of individuals (+/– standard error) per count is shown for plots with normal, half and quarter dosage. A: Beets; swallows and martins. B: Beets; Skylark. C: Beets; seed-eaters (all records). D: Beets; seed-eaters (only birds on the field). E: Barley; Skylark. F: Barley; seed-eaters (only birds on the field).

Some results of the counts on the experimentally sprayed pilot test fields are visualized in Fig. 4.1. The census period was divided into three parts for the beet field (1 May - 5 June; 10 June - 2 July; 6 July - 12 August) and two parts for the barley field (1 May - 10 June; 14 June - 11 July); in the wheat field no experimental sprayings were carried out. Results are shown for swallows, Skylark, and small seed-eaters (sparrows, finches and buntings).

The small seed-eaters occurring in farmland mainly feed in the fields but nest and seek cover in hedges and coverts. They are thus recorded on the field as well as in the surrounding hedges. For these species, analyses were carried out on all records as well as on the smaller sample of records from the field proper.

A formal test of differences between dosage plots is hampered by the lack of true replications; the observations on different dates within the breeding season are not strictly independent. However, to give at least some indication about the significance of the apparent differences, a log-linear model was fitted for each field and species (group), assuming the number of individuals recorded on a certain plot and date to follow a Poisson distribution. Likelihood ratio tests were used to test for differences between plots with respect to mean numbers and development in numbers between census periods. To be conservative, tests were adjusted for over-dispersion (variance greater than would be expected from a Poisson distribution), whereas under-dispersion (variance smaller than expected, indicating non-independence between observations) was not allowed. The analyses were performed using the GENMOD procedure in SAS/STAT software.

The following differences were found:

Beets, swallows: Significant differences between plots (p = 0.045*); pairwise contrasts reveal more birds at quarter dosage than at normal and half dosages (p = 0.025* and 0.047*, respectively). A closer look at the data shows that this difference did only occur in mid- and late June.

Beets, Skylark: Significant differences between plots (p = 0.0045**); more birds at half dosage than at normal (and maybe quarter) dosages (p = 0.0012** and 0.052, respectively).

Beets, seed-eaters (all): Significant differences between plots (p = 0.0092**); fewer birds at half dosage than at quarter (and maybe normal) dosages (p = 0.0024** and 0.054, respectively).

Beets, seed-eaters (field): Significant differences in development in numbers between plots (p = 0.019*); less positive development during the season at normal dosage than at half and quarter dosages (p = 0.0099** and 0.020*, respectively). No differences in overall mean numbers, but significant differences between plots in the last period (p = 0.029*); fewer birds at normal dosage than at quarter (and maybe half) dosages (p = 0.0090** and 0.053, respectively).

Barley, seed-eaters: Significant differences between plots (p = 0.014*); more birds at quarter dosage than at normal dosage (p = 0.0044**).

The effects of hedgerows are clear from Table 4.1 as well as from the results presented in Fig. 4.1. In the beet field, the plots which received normal and quarter dosages were situated along the hedges surrounding the field whereas the half-dosage plot just bordered on hedges at its ends. Thus the Skylarks' avoidance of areas close to hedgerows is also the reason for their preference of the half-dosage plot. Conversely, the apparent avoidance of the half-dosage plot by the small seed-eaters, which include Linnet and Yellowhammer, (Fig. 4.1 C) is caused by their preference for hedgerows. Hedgerows may also affect the distribution of airborne species like swallows: swallows were fairly evenly distributed over the test field except on two observation days in mid-and late June, when quite strong westerly winds dominated and the swallows were almost exclusively foraging sheltered by the hedgerow bordering the quarter dosage plot (Fig. 4.1 A).

If only records of seed-eaters from the field proper are considered (Fig. 4.1 D), the balance between half and normal dosage is shifted in favour of the half-dosage plot. It is notable that there was a pronounced increase in numbers during the season in the plots with half and quarter dosage but not in the normal-dosage plot.

This might well reflect that towards the end of the season, more birds' food items were available in the half- and quarter-dosage plots than in the normal-dosage plot. Despite the name, the small "seed-eaters" during the breeding season to a large extent feed on arthropods, especially when feeding young (Christensen et al. 1996).

In the barley field, which was not bordered by hedgerows, a similar difference was found: significantly more small seed-eaters were recorded in the quarter-dosage plot than in the normal-dosage plot (Fig. 4.1 F). There were no differences in the distribution of Skylarks between plots.

Thus, from the ornithological pilot studies during the breeding season, two important conclusions emerged: (1) On the experimentally sprayed fields, some differences in the distribution of birds between dosage plots occurred. Dosage plots of 6 hectares and a sampling program with counts every fifth day during a three month period seemed sufficient to detect the differences. (2) The occurrence of hedgerows affects the distribution of all common farmland species to such an extent that even sizable treatment-related differences between plots may be masked.

Methodologically, it appeared that the point counts should be supplemented with line transects in order to increase the detection chance for stationary birds in the field. Further, it was clear that proper registration of swallows and martins demanded so much of the observer's attention that the registration of other, more important species suffered. So, as their distribution seemed to be very much affected by the meteorological conditions, it was decided that swallows, martins and swifts should not be recorded.

4.1.2 Winter studies

As part of the pilot studies, censuses were carried out during the winter of 1996-97 on four fields with autumn-sown wheat (the experimental wheat fields at Gjorslev, Lekkende, Oremandsgård and Nordfeld). Three counts per field were conducted between November and March. It was not clear from these pilot censuses whether the density of birds on the fields in winter was sufficient to allow statistical testing. Therefore, a full-scale winter study was carried out after the first project year (1997) on all 15 experimental fields.

The winter counts were performed as line transects. Following a set pattern, the observer scoured each dosage plot, using a constant effort of 5 minutes per ha. All birds seen or heard on the field, above the field, or in hedgerows or habitat islands adjacent to the field were recorded. Before moving into the field, the observer scanned all three plots with a binocular, with the purpose of counting any birds of shy species that might be flushed when the transects were started. The number of birds recorded in each plot was converted to a standard plot size of 6 ha before analysis. Censuses were carried out between 1 November 1997 and 15 March 1998. Each of the 15 fields was visited twice a month, yielding a total of 9 counts per field.

During the censuses 1996-97, a total of 168 birds of 21 species, or 4.7 birds per plot and count, were recorded (excluding a flock of 140 Common Gulls). No attempts to carry out any statistical analyses on this small sample were made.

In the winter of 1997-98, 4506 individuals of 49 species were recorded. Six species (groups) were selected for statistical analysis: Skylark, Crow, tits, Blackbird, Yellowhammer and small seed-eaters. Two analyses of the occurrence of Crow, Yellowhammer and seed-eaters were carried out: one using all records, one with records from hedgerows and habitat islands being excluded. Each of the three types of fields, winter wheat, "after wheat" (to be sown with beets) and "after beets" (to be sown with barley) was dealt with separately, so the total number of analyses performed was 27. Analysis of variance was used to test for differences between dosages treating the different farms as blocks. Because the repeated censuses of each plot were not strictly independent, the geometric mean of the 9 counts, calculated separately for each species (group), was used as the dependent variable. In the calculation of the geometric mean, the densities from the individual counts (which might contain zeros) were x+1 transformed.

Only one of the 27 tests for differences between dosages was significant at the 5% level: Blackbird on fields after wheat (p = 0.011*, more birds at quarter dosage than at half and normal dosages). A further two tests, both relating to the "after beets" fields, were significant at the 10% level, with the highest number of birds being found at normal dosage (Crow) and quarter dosage (Yellowhammer).

The number of significant tests is equal to the number that would be expected by chance, if no true differences between dosage plots exist. The inevitable conclusion is that after one year of experimental treatments, no differences between dosages could be detected with respect to the occurrence of birds on the plots in winter. This is not very surprising, considering that the seed production of the year with experimental dosages was ploughed in during autumn.

Of greater importance is the fact that bird densities during winter proved quite low. Although a total of 4506 birds may sound impressive, the geometric mean number of individuals per count per plot (6 ha) did not exceed one in any of the species analysed. At such low densities, chance events become of considerable importance, and it was considered unlikely that any reliable differences between dosage plots could be detected. Therefore, winter counts were not performed during the remainder of the study period.

4.2 Methods

During the main phase of the study (1997-99), two kinds of ornithological investigations were carried out. Firstly, counts were performed on all 15 fields in all three study years from shortly after the germination of the spring crops until a few weeks before the harvest of barley and wheat. This census period coincides with the main breeding period of most bird species. Secondly, counts were carried out on the stubble fields in early autumn, from immediately after harvest until the fields were ploughed. These counts, coinciding with the dispersal and migration period, were performed in 1998 and 1999, i.e. after the winter counts were dropped.

4.2.1 Breeding season counts

Each observation day, two kinds of censuses were carried out: morning counts (between 8 and 11 a.m.) and afternoon counts (between 12 and 15). Different methods were used for the two kinds of censuses. The morning counts were mainly based on point counts, and birds on the field as well as in the surrounding vegetation were censused. The afternoon counts were performed as line transects, and only birds on the field were censused. In all censuses, all species except swallows/martins and swifts were recorded.

In the morning, the sampling unit was the subplot. Before the counts began in early May, 12 subplots of about equal size were demarcated in each field. The number of subplots within each dosage plot was four or six, depending on the presence of hedgerows at the field borders and thus the comparability of plots (cf. section 4.1.1). If there were no hedges around a field, or if all three plots bordered on an equal amount of hedgerow and thus were directly comparable, four subplots of 1.5 ha were delimited within each plot (Fig. 4.2 A). In 7 of the 15 fields, however, the presence of hedgerows made it impossible to lay out three fully comparable dosage plots (Fig. 4.2 B). In these cases, six subplots of 1.0 ha were delimited within each of the two similar plots (always the normal- and quarter-dosage plots, cf. section 1.2.5), and the half-dosage plot was not censused during the morning counts.

The birds within each subplot were counted by means of a combination of point and transect counts. Firstly, a 10 minutes point count was carried out while the observer was stationary at the edge of the subplot. Then the observer spent 5 minutes walking through the subplot to the border of the next, adding to his list of records any birds within the subplot that had not been recorded from the census point. During the 15 minutes spent in each subplot, all birds seen or heard within or immediately above the subplot were recorded and assigned to one or more of the following four zones: hedge/habitat island; field 0-12 m (from hedge/habitat island); field > 12 m; air. Care was taken not to count an individual more than once. If a bird was recorded in more than one subplot, the observer assigned it to one, and just one of these. The chosen subplot should be the one where the bird did most of its feeding, not necessarily the one in which the nest was assumed to be located.

On the afternoon counts, the sampling unit was the plot, and all three plots were censused in all fields. Line transects were used. Following a set pattern, the observer walked through each plot, using a constant effort of 6 minutes per ha. All birds seen or heard within or immediately above the plot were recorded and assigned to one or more of the following three zones: field 0-12 m (from hedge/habitat island); field > 12 m; air. Because these counts primarily aimed at recording birds foraging in the fields, birds in hedgerows and habitat islands were not censused. Like at the morning counts, a bird recorded in more than one plot was assigned to one of these by the observer.

Fig. 4.2.
Two examples of the delimitation of subplots within plots. A: Four subplots of 1.5 ha within each plot (Nøbøllegård). B: Six subplots of 1.0 ha within the normal- and quarter-dosage plots (Gjorslev). Aerial photos by courtesy of Kampsax.

A total of 17 or 18 counts (each consisting of a morning count and an afternoon count) were performed at each field per year. In 1997, 17 counts per field were carried out between 6 May and 31 July. In 1998, the census period was 11 May to 5 August, and 17 counts per field were performed. Finally, in 1999 18 counts were performed between 6 May and 5 August. To avoid systematic biases from changes in the birds' activity during the day, the starting plot varied from count to count according to a rotating scheme.

The basic experimental unit was the plot, and the standard plot size was 6 ha. Consequently, for each bird species, field and year, the dependent variable was the number of individuals per 6 ha plot per count. The results of the morning and afternoon counts were analysed separately. For the morning counts, the number of individuals recorded in each subplot within a plot were summed up to yield a 6 ha plot total for each species. For the afternoon counts, the number of birds recorded within each plot was converted to the standard plot size of 6 ha by division. In order to improve approximation to a normal distribution, and make the variance independent of the mean, all standard plot totals were log_e(x+1) transformed before further analysis.

The repeated censuses of a certain plot during the season are not statistically independent. Therefore, some kind of repeated-measures analysis is needed to test for differences between dosage plots (e.g., Stryhn 1996). Described graphically, a suitable method is as follows: The dependent variable is depicted as a function of the census date; for each species and type of count, this creates 135 curves (3 plots in 15 fields in 3 years). The 135 curves are then tested against each other for differences between dosages. This can be done by fitting a 16th or 17th degree polynomial (depending on the number of census dates) to each curve, whereupon the parameters (zero degree terms, 1st degree terms etc.) are tested against each other by means of analysis of variance. If a parameterization which yields orthogonal polynomials is used, the different parameters may be tested and (to some extent) interpreted independently of each other. In practice, the high-order coefficients are difficult to interpret, for which reason analyses of this kind are often limited to tests of zero, 1st and 2nd degree terms.

In the present case, the basic hypothesis is as follows: At the beginning of the breeding season, before any notable effects of pesticide sprayings occur, no systematic differences between dosage plots exist. As the season progresses, and the effects of the herbicide and insecticide sprayings manifest themselves, differences in the amounts of insects and other birds' food items develop between plots. Consequently, the quarter-dosage plots during the breeding season become relatively more attractive to the birds as feeding sites and the normal-dosage plots become relatively less attractive, with the half-dosage plots placed somewhere in between. This should be reflected in the seasonal development in the number of birds foraging in each of the three plots within a field.

The crucial check of this hypothesis is to test whether the first-order coefficient of the polynomial, i.e. the slope of the curve, is significantly greater in the quarter- (or half-) dosage plots than in the normal-dosage plots. Differences in mean number (the zero-order term) between plots may be a corollary of the differences in development, or they may reflect more permanent differences between plots. Because the relative dosage applied to a certain plot was the same in all three years, effects may have accumulated during the study. Accordingly, towards the end of the study period, dosage-related differences between plots might be present from the start of the breeding season.

To deal with this approach in practice, the mean and slope of the number-of-birds vs. date curve were calculated for each species, plot and year, using the REG procedure in SAS/STAT. The means and slopes were then analysed for differences between dosages by means of analysis of variance, taking into account the effect of differences between crops, years, farms and fields. It should be noticed that dosage was treated as a class variable, so no assumptions about the effect of half dosage falling in between those of quarter and normal dosage were made. The anova design applied is shown in Table 4.2; starting with the full model, stepwise model reduction was used. If a significant effect of dosage was revealed, least-squares means were calculated, and t-tests were used to test for pairwise differences. The analyses were performed using the GLM procedure in SAS/STAT.

Table 4.2.
The factors included in the anova of the occurrence of birds during the breeding season, the nature of each factor (fixed/random), and the denominator (error term) used in the test of significance of each factor in a balanced design (see text).

The error terms given in Table 4.2 are only fully valid in a balanced experimental design. In the present case, however, only the afternoon counts represent such a design. Because not all of the half-dosage plots were censused at the morning counts (cf. section 4.2.1.1), the experiment in this case is unbalanced and the F-tests must be modified, causing the tests to be only approximate. The modifications applied involve a weighting of the different terms in the denominator of the F-tests and an adjustment of the degrees of freedom, as called by the RANDOM/TEST statement in the GLM procedure in SAS/STAT (SAS Institute 1990).

While the test procedures thus were based on a model with just mean and slope - in order to make the interpretation of the tests as plain as possible - 2nd degree models were calculated for illustrative purposes and to provide more information about the seasonal development in bird numbers.

Two species and one species group were selected for analysis: Skylark, Whitethroat and small seed-eaters (sparrows, finches and buntings). These species were the only ones occurring in sufficient numbers in the fields to make a reliable analysis possible. As previously mentioned, morning and afternoon counts were analysed separately. Furthermore, the analyses of morning counts of Whitethroat and seed-eaters were performed on the total sample as well as on the smaller sample from the field proper. Thus, a total of eight different (but not mutually independent) series of curves were analysed. Table 4.3 summarizes the dependent variables selected and the size of the material.

Table 4.3.
The species selected for analysis and the census variables used. The number of individuals recorded is given for each of the eight variables.

The autumn counts on harvested fields were basically carried out in the same way as the winter counts (section 4.1.2.1), but with a census effort corresponding to that of the afternoon counts during the breeding season. Thus, the sampling unit was the plot, all three plots were censused in all fields, and the census method was line transects. Before the transects were started, the observer scanned all three plots with a binocular, and any birds of shy species (geese, plovers etc.) that might be flushed when he moved into the field were counted. Then the observer walked through each plot following a set pattern, using a constant effort of 6 minutes per ha. All birds (except swallows/martins and swifts) that were seen or heard within or immediately above the plot, or in the adjacent vegetation, were recorded and assigned to one or more of the following four zones: hedge/habitat island, field 0-12 m (from hedge/habitat island); field > 12 m; air. Like on the other censuses, care was taken not to count any individual more than once, and if a bird was recorded in more than one plot, it was assigned to one of these by the observer. The starting plot varied from count to count according to a rotating scheme.

Only the ten cereal fields were censused. The counts were started as soon as possible after harvest, i.e. between 22 August and 8 September. Each field was censused once a week until it was ploughed, although with a maximum of 9 counts per field in 1998 and 7 counts in 1999. In 1998 the farmers were allowed to follow their normal routines and plough the fields as soon as they pleased; this led to great differences in counting periods and number of counts per field. Therefore, in 1999 the farmers were compensated for postponing the ploughing of the study fields until 7 counts had been performed. Unfortunately, one farm did not want to participate, so only 8 fields were censused in autumn 1999. A total of 64 and 56 counts were carried out in 1998 and 1999, respectively. The average number of counts per field, the period during which the counts were performed, and the median counting date are shown in Table 4.4.

Table 4.4.
Statistical information about the censuses on stubble fields in 1998 and 1999. For each combination of crop and year, the number of counts is shown (average and range). Also, the period during which the censuses were performed is shown together with the median counting date.

For each plot and census, the density of each species recorded was expressed as the number of birds per 6 ha, and the densities were log_e(x+1) transformed before further analysis. Because the repeated censuses of each plot within a year are not strictly independent, the mean of the log-transformed densities (equivalent to the geometric mean) were used as the dependent variable in the analyses. Contrary to the situation in the breeding season, any dosage-related differences in food abundance between plots were expected to be manifest from the beginning of the census period, so tests of trends were not performed.

The geometric means of selected species were analysed for differences between dosages by means of analysis of variance. The experimental design is incomplete in this case. Basically, the whole experiment is a well balanced multiple Latin Square design: 3 crops, 3 fields, 3 years and 5 replicates (farms). The autumn counts, however, concern just a subset of the Latin Square: two crops and two years, but still 3 fields per farm (due to the rotation of crops). Therefore, the effects of crop, year and field cannot be estimated simultaneously; this led to a modification of the anova design (Table 4.5).

Table 4.5.
The factors included in the anova of the occurrence of birds on stubble fields, the nature of each factor, and the denominator (error term) used in the test of significance of each factor in a balanced design (see text).

In this anova, the interaction term crop´ year contains variation arising from differences in counting periods (cf. Table 4.4) as well as variation from year-dependent crop effects. The 2nd order interaction term crop´ year´ farm allows these effects to vary between farms, but includes as well some variation stemming from differences between fields.

As previously mentioned, there were 5 replicates in 1998 but just 4 in 1999, so the experimental design is unbalanced and the tests only approximate. The approximate F-tests were constructed in the same way as for the morning counts in the breeding season (section 4.2.1.2). Starting with the full model shown in Table 4.5, stepwise model reduction was used. All analyses were performed using the GLM procedure (with RANDOM/TEST statement) in SAS/STAT.

Two species and one species group occurred in numbers that were sufficient for analysis: Skylark (3,434 individuals recorded), Meadow Pipit (722) and small seed-eaters (3,415). Whereas Skylarks and Meadow Pipits chiefly occur on the fields, Yellowhammers and other small seed-eaters are also frequently recorded in the adjacent hedgerows and habitat islands. The analysis of the occurrence of these species was therefore carried out on the sample of birds from the field proper (1,861 individuals) as well as on the total sample.

4.3 Results

4.3.1 Breeding season counts

The results of the analyses of variance are summarized in Tables 4.6 - 4.8. It is clear from the tables that all main factors have significant effects on the distribution of all three species. The partitioning of the sums of squares (not shown) indicates that the quantitatively most important factors affecting the mean numbers of birds are farm and field (i.e. block factors), while crop plays a minor, but still important role. Crop, farm and field also account for the majority of the variation in slope (i.e. development in numbers), but farm and field are less dominant, and effects of dosage become apparent. So, while there are great differences between farms, and also between single fields, with respect to the mean number of birds present, the variation between blocks is less pronounced with respect to the changes in bird numbers that occur during the season, exposing the effects of the experimental factors crop and dosage. Variation between years, although often significant, never accounts for more that 10% of the total variation.

Table 4.6.
Schematic summary of the analyses of the occurrence of Skylarks on the experimental fields during the breeding season. Statistical significance is indicated as follows: *: 0.01<p<0.05, **: 0.001<p<0.01, ***: p<0.001.

Table 4.7.
Schematic summary of the analyses of the occurrence of Whitethroats on the experimental fields during the breeding season. Statistical significance is indicated as follows: *: 0.01<p<0.05, **: 0.001<p<0.01, ***: p<0.001.

Table 4.8.
Schematic summary of the analyses of the occurrence of small seed-eaters on the experimental fields during the breeding season. Statistical significance is indicated as follows: *: 0.01<p<0.05, **: 0.001<p<0.01, ***: p<0.001.

In all three species analysed, the occurrence differs between crops, although the effect of crop often varies between farms (as evidenced by a significant crop´ farm interaction). Concerning the effects of dosage, an important result is the lack of significant dosage´ crop interactions, implying that the effect of dosage can be analysed independently of any crop effects. In some cases the dosage´ farm interaction is significant (although not highly so), indicating that the effect of dosage varies between farms. Finally, the lack of any significant dosage´ year interaction indicates that the effect of dosage has been the same in all three study years.

In the Skylark, the pattern of occurrence is mainly determined by the crop, while the dosage effects are barely significant (Table 4.6, Fig. 4.3). The concordance between the models based on the morning and afternoon counts is good, considering that they use data from two independent series of counts, with different census methods. In the interpretation of the results, due attention should be paid to the fact that the number of birds recorded is a function of the number of birds present in the field and their activity. The number of Skylarks seems to peak in June, but numbers are surely still high in July when they are increased by the newly fledged young. However, the territorial activity is much higher in May and June than in July.

In the beginning of the season, before the germination of the spring crops, the highest Skylark densities occur in winter cereals. But when the growth of the spring cereals makes the barley fields attractive to Skylarks around 1 May, numbers increase here, reaching a culmination in June. In the beet fields, the number of Skylarks increases throughout the breeding season, and from mid-July onwards the highest densities are found in this crop.

Fig. 4.3.
Models of the development in Skylark numbers on the experimental fields from early May to early August in relation to crop (A, B) and dosage (C, D). A & C are based on the morning counts, B & D on the afternoon counts. All models are based on log-transformed data, so the densities indicated are not comparable with normal arithmetic mean densities.

Compared with the crop effects, the effects of differences in dosage on the distribution of Skylarks are small (Fig. 4.3 C-D). On the afternoon counts, the between-dosages differences in development (slope) are significant (p = 0.021*); pairwise t-tests indicate a significant difference between half and normal dosage (p = 0.0061**), but not between quarter and normal (p = 0.085). On the morning counts, the dosage effects are not significant (p = 0.053), but the picture is the same as on the afternoon counts: the decrease in Skylark numbers begins earlier in the season in normal-dosage plots than in plots treated with half or quarter dosage. Actually, pairwise t-tests reveal a significantly "better" (i.e. less negative) development in quarter-dosage plots than in plots treated with normal dosage (p = 0.047*, Tukey-Kramer adjustment for multiple comparisons).

The first Whitethroats occur in the first week of May, and four weeks later almost all of the population has arrived from its African winter quarters. Territorial activity is high in the first half of June, and from around 20 June onwards the number of Whitethroats is increased by the first fledglings. During July, territorial activity decreases and the birds become less visible (Fig. 4.4 A).

Fig. 4.4.
Models of the development in Whitethroat numbers on the experimental fields from early May to early August in relation to dosage (A, C, D) and crop (B). A-C are based on the morning counts, D on the afternoon counts. A concerns all records (including the border vegetation), B-D only records inside the field. All models are based on log-transformed data, so the densities indicated are not comparable with normal arithmetic mean densities.

Although they frequently feed in the fields, Whitethroats need trees and bushes, or at least tall herbs, at the field borders for song-posts and nesting. Because the border vegetation varies strongly between the study fields, block factors farm and field account for the majority of the variation in Whitethroat numbers. Nonetheless, dosage effects are prominent, even when all records (including those in the border vegetation) are considered (Table 4.7). There are clearly significant differences in slope between dosages (p = 0.0021**); pairwise t-tests indicate that the overall increase in numbers during the season is significantly stronger in quarter-dosage plots than in normal-dosage plots (p = 0.0005***), with the development in half-dosage plots falling somewhere in between (Fig. 4.4 A).

About 17% of the morning records of Whitethroats are from the fields proper. The Whitethroats' use of the fields is crop-dependent, but generally the proportion of birds recorded in the fields increases during the season, accompanying the growth of the crops. Barley fields are used very little while wheat fields are more frequently used. The beet fields are initially not used at all, but become important feeding sites from around 10 July onwards (Fig. 4.4 B).

The dosage has a pronounced effect on the Whitethroats' distribution within the fields (Fig. 4.4 C-D). On the morning counts, the differences in development (slope) between dosages are highly significant (p = 0.0002***); pairwise t-tests indicate significant differences between quarter and normal dosage (p < 0.0001***) as well as between half and normal (p = 0.018*). The differences in mean numbers follow the same pattern. On the afternoon counts, the between-dosages differences in slope are also highly significant (p < 0.0001***), and pairwise t-tests indicate a highly significant difference between quarter and normal dosage (p = 0.0002***). However, the development in Whitethroat numbers in half-dosage plots is significantly different from that in quarter-dosage plots (p < 0.0001***) but does not differ from the development in plots treated with normal dosage. A probable reason for this difference between the morning and afternoon counts is that on the morning counts, only half-dosage plots surrounded by the same amount of hedgerows as the normal- and quarter-dosage plots were censused (cf. section 4.2.1.1), whereas all plots were censused on the afternoon counts, including those half-dosage plots that abut on a smaller amount of hedgerow than the normal- and quarter-dosage plots and thus are less likely to be visited by Whitethroats.

The small seed-eaters (sparrows, finches and buntings) nest in hedgerows, coverts and around farmsteads, whereas a major part of their foraging takes place in the fields. The number of individuals occurring in and along the fields rises steadily during the breeding season, partly because the population sizes are increased by fledglings, partly because the birds (especially sparrows) move from the farmstead surroundings to the fields when the young have fledged. Farm and field differences account for the majority of the variation in numbers because of great between-fields variation in border vegetation and distance to farm buildings.

Even when all records are considered, effects of dosage can be distinguished (Table 4.8, Fig. 4.5 A). There are significant differences between dosages in the development in seed-eater numbers during the breeding season (p = 0.020*); pairwise t-tests indicate slope differences between quarter and normal dosage (p = 0.037*) and between half and normal (p = 0.011*). There are similar differences in mean numbers although the difference between half and normal dosage is not significant (p = 0.096).

About 21% of the morning records of small seed-eaters concern birds inside the fields. The birds' use of the fields for feeding differs between crops (Fig. 4.5 B). In winter wheat there is a slow but steady increase in the number of birds during the season. The spring crops are used during a short period in May, but later on barley is the least preferred crop. Conversely, beets are used throughout the season, and beet fields are very important feeding sites from the beginning of July onwards.

Besides the crop effects, the distribution of small seed-eaters in the fields is affected by dosage effects. On the morning counts (Fig. 4.5 C), there are significant differences in seasonal development between dosages (p = 0.0081**); pairwise t-tests indicate significant slope differences between quarter and normal dosage (p = 0.013*) as well as between half and normal (p = 0.0069**). The differences in mean numbers follow the same pattern. On the afternoon counts, the differences in development (slope) between dosages are not significant (p = 0.17), and the effects of dosage on mean numbers vary between farms. However, the models for quarter, half and normal dosage (Fig. 4.5 D) follow a pattern which is quite similar to that seen on the morning counts.

Fig. 4.5.
Models of the development in the numbers of small seed-eaters on the experimental fields from early May to early August in relation to dosage (A, C, D) and crop (B). A-C are based on the morning counts, D on the afternoon counts. A concerns all records (including the border vegetation), B-D only records inside the field. All models are based on log-transformed data, so the densities indicated are not comparable with normal arithmetic mean densities.

4.3.2 Autumn counts

The results of the analyses of variance are summarized in Table 4.9. There are great differences between the analysed species and very few, if any, general conclusions can be drawn. The partitioning of the sums of squares (not shown) indicates that the dominating sources of variation are farm differences and/or the crop´ year´ farm interaction (which includes between-fields differences, cf. section 4.2.2.2). Effects of differences between years and between crops are less pronounced. No general effect of dosage is indicated, but in the Skylark and the seed-eaters, the dosage´ farm interaction accounts for 17% and 11%, respectively, of the total variation.

Table 4.9.
Schematic summary of the analyses of the occurrence of selected bird species on the experimental fields in autumn. Statistical significance is indicated as follows: *: 0.01<p<0.05, **: 0.001<p<0.01, ***: p<0.001.

In the Skylark, the major sources of variation are farm and field differences. There is a clearly significant dosage´ farm interaction, indicating that some differences between dosages exist, but vary between farms. Testing each farm separately reveals significant (or almost significant) dosage effects in two farms only: Gjorslev (p = 0.0056**) and Nøbøllegård (p = 0.058). At Gjorslev, Skylark numbers are clearly higher in half-dosage plots than in normal- and quarter-dosage plots (pairwise t-tests: p = 0.0048** and 0.0031**, respectively). At Nøbøllegård, numbers tend to be lower in half-dosage plots than in normal- and quarter-dosage plots (p = 0.056 and 0.028*, respectively) (Fig. 4.6).

Fig. 4.6.
The occurrence of Skylarks on the experimental stubble fields at Gjorslev and Nøbøllegård in autumn. The geometric mean numbers (+/– 2*standard error) are shown for plots sprayed with normal, half and quarter dosage.

In both cases, the differences in Skylark numbers are probably due to structural differences, rather than to differences in pesticide use. At Gjorslev, all fields are surrounded by hedges, making the outer plots (always normal and quarter dosage) less attractive to Skylarks than the central (half dosage) plots (cf. section 4.1.1). At Nøbøllegård the reverse is true: the field that was censused in both years (and thus contributed 50% of the data) is bordered by a hedgerow along the half-dosage plot, making the quarter- and normal-dosage plots the most attractive to Skylarks.

In the Meadow Pipit, the crop´ year´ farm interaction is the dominant source of variation. In addition, the factors farm and crop´ year each account for 15-20% of the total variation. There are no indications of dosage effects. The Meadow Pipits occurring on the stubble fields are almost exclusively resting migrants, with numbers peaking from late September to mid-October. In 1998, when the number and timing of the counts varied between fields (Table 4.4), the censuses on almost all of the barley fields stopped before the peak of the migration, whereas the censuses on wheat fields continued until late October and thus included the main migration period. This is probably the main reason for the dominance of the crop´ year and crop´ year´ farm interactions.

The occurrence of small seed-eaters, contrary to the other species, is to a large extent dependent on the existence of hedgerows or other suitable vegetation along the field borders. The farms and fields vary in this respect, which is probably the reason why farm differences is the dominant source of variation in the numbers of small seed-eaters. There are significant differences between years; for unknown reasons more birds were recorded in 1999 than in 1998 (no such difference was apparent during the breeding season). A significant difference between crops is evident when all records are included, but disappears when only records from the field are considered and is therefore of little interest. There are no significant dosage effects.

4.4 Discussion

The bird species analysed represent different phenologies and strategies in their exploitation of the farmland environment and may thus be considered representative for the bird community associated with arable fields. The Skylark is the only species occurring in large numbers that both breeds and forages in agricultural fields, where it is dominant and numerous from March until November. During the breeding season, its diet mainly consists of arthropods, but in autumn cereal grains and other seeds become important, and in winter and spring green plant material is frequently eaten as well (Christensen et al. 1996). Probably as part of their anti-predator strategy, Skylarks avoid areas close to hedgerows and other vertical structures (cf. Petersen 1996, Chamberlain & Gregory 1999).

Contrary to the Skylark, the small seed-eaters (Yellowhammer, Linnet, Chaffinch, Greenfinch, Tree Sparrow etc.) take advantage of the existence of suitable vegetation bordering the fields, or habitat islands within the fields, for cover and breeding. However, most of their foraging takes place inside the fields. The species occur in Danish farmland year round, often forming flocks outside the breeding season. Over the year, plant seeds (including cereal grains) make up the bulk of their diet, but during the breeding season, insects are an important part of the diet in most species.

Whitethroats are strongly associated with hedgerows and tall herbaceous vegetation along the field borders but make frequent foraging trips into the fields. Being tropical migrants, they arrive in Danish farmland in May and leave in August-September. Whitethroats are chiefly insectivorous, but fruits and berries are of some importance in late summer and autumn.

Like the Whitethroat, the Meadow Pipit is mainly insectivorous, but in early autumn seeds may constitute 15% of the diet and even more later in the year (Cramp & Simmons 1977-94). The species breeds on permanent grassland and is rarely found on arable fields during the breeding season, but during autumn and spring migration (September-October and April) large numbers rest on suitable fields. Meadow Pipits show neither preference nor avoidance of hedgerows.

Put together, the selected species utilize a broad spectrum of the animal and vegetable resources available all over the fields. The major limitations follow from their relatively small size (max. Skylark 40 g) and the fact that none of the species use their bills to probe the soil (like Starling). No larger species occurred in numbers allowing reliable statistical analyses of their distribution.

Optimal use of the available resources by the birds involves changes of crop preferences during the breeding season. The Skylark's shift in preferences is visualized in Fig. 4.3. Skylarks seemingly prefer not too dense crops of 15-45 cm height, which is probably the reason for their gradual shift from winter cereals to spring cereals in May and general leaving of the cereal fields during July (cf. Schläpfer 1988, Jenny 1990b, Wilson et al. 1997, Chamberlain et al. 1999, Toepfer & Stubbe 2001). In dense crops Skylarks often concentrate their foraging activities around tramlines and other unsown areas where locomotion is unhindered and prey items are more easily seen (Odderskær et al. 1997b). Row crops like beets, by contrast, allow the birds to walk about unrestricted all over the field, and beet fields become useful for Skylarks when the crop biomass (and hence the structural diversity) starts to increase rapidly from mid-June onwards.

The numbers of Whitethroats and seed-eaters (and several other bird species breeding in the border vegetation) that occur in the fields generally increase as the season progresses, accompanying the growth of the crops. This is especially pronounced in beet fields where the number of birds rises tremendously during July (Figs. 4.4 B & 4.5 B). A strong increase in Yellowhammers' utilization of beet fields after 10 July was also found by Biber (1993). Beet fields offer good cover and easy access and also hold good numbers of preferred birds' food items from mid-July onwards (Fig. 3.4). In cereal fields there is a slow but steady increase in the number of foraging birds of these species throughout the breeding season, most pronounced in wheat; two British studies show a quite similar picture for both Yellowhammer (Stoate et al. 1998) and Whitethroat (Cracknell 1986). Yellowhammers and other small seed-eaters also show some preference for spring crops during a short period after sowing, when the loose soil and low vegetation suit their feeding behaviour.

Contrary to the situation in the breeding season, no crop preferences were found on the autumn counts. None of the analysed bird species show any signs of preferring wheat stubble to barley stubble or vice versa. In Britain, Donald et al. (2001) found that wintering Skylarks preferred barley stubbles to wheat stubbles. Generally, stubble fields are preferred habitats during autumn and winter, compared to fields with autumn-sown crops, grass ley or bare till (Petersen & Nøhr 1992, Wilson et al. 1996, Robinson & Sutherland 1999, Donald et al. 2001).

Despite clear differences in the birds' use of the crops during the breeding season, the effects of dosage do not in any case interact significantly with the effects of crop. In other words, the effect of dosage on the occurrence of the analysed bird species is principally the same in all three crops studied.

In some cases, the dosage´ farm interaction term is significant, indicating that the effect of dosage varies between farms. However, a significant interaction is almost exclusively found on the afternoon counts, where all plots were censused, including those half-dosage plots that are not fully comparable with the normal- and quarter-dosage plots (cf. section 4.2.1.1). Therefore, a significant dosage´ farm interaction is probably chiefly due to the lack of true comparability of plots, rather than to real between-farms differences in the effects of dosage. The implications of the lack of full comparability of plots for the estimations of dosage effects are further discussed in section 4.4.1.

As explained in section 4.2.1.2, the basic hypothesis is that no systematic differences between dosage plots exist at the beginning of the breeding season, but that differences develop as the season progresses and the effects of the pesticide sprayings come to the fore. If dosage effects accumulate over the years, differences between dosage plots might exist from the outset in years 2 or 3. This would show in the analyses as a significant dosage´ year interaction term, as would also be the case if the effect of dosage varies between years. In all analyses, however, the dosage´ year interaction is far from significant, indicating that there have been no between-years differences in the effect of dosage and no traceable accumulation of effects over the three year study period.

Clear dosage-related differences are found in the breeding season counts of Whitethroats and small seed-eaters; the most pronounced differences occur in the former. In all tests, the number of Whitethroats increases significantly more rapidly in quarter-dosage plots than in normal-dosage plots. Similar significant differences are found in the morning counts of small seed-eaters, while the afternoon counts follow the same pattern without the differences being significant.

It is probably safe to conclude that the development in the numbers of Whitethroats and small seed-eaters occurring in the fields from May to August is in full accordance with the basic hypothesis: towards the end of the season, the birds clearly prefer plots treated with quarter dosage to plots treated with normal dosage of herbicides and insecticides. All these species make foraging trips into the fields from the surrounding vegetation, but do not establish territories inside the fields. Consequently, each individual is free to choose an optimal foraging site within a certain field.

In the Skylark, the differences between dosages are much less clear than in the species discussed above. However, numbers seem to decline earlier in the season in normal-dosage plots than in half- and quarter-dosage plots, both on the morning and on the afternoon counts (Fig. 4.3 C-D). On the morning counts, there is a significant difference in development between quarter- and normal-dosage plots, whereas on the afternoon counts it is the difference between half- and normal-dosage plots that is significant at the 5% level.

Although slight, the shift from plots treated with normal dosage to half- and quarter-dosage plots is in accordance with the basic hypothesis. The shift may reflect that, as the season progresses, the birds prefer to forage in plots with reduced dosages. Another explanation may be that fewer pairs carry through a second breeding attempt in normal-dosage plots than in half- and quarter-dosage plots (cf. Odderskær et al. 1997a).

The Skylarks perform almost all of their feeding inside the fields. So, on the face of it, it may be surprising that the most numerous and specialised field species among those studied is the one that shows the weakest response to differences in pesticide treatments. One obvious explanation may be that the graded dosages do not lead to appreciable changes in the distribution of the Skylarks' food resources (but see chapter 3). Another reason may be that vegetation structure (which may or may not be dosage-related), rather than food abundance, is the most important distributing factor (Odderskær et al. 1997b, Chamberlain et al. 1999). Also, the territorial system of Skylarks may seriously restrain the individual birds' choice of feeding sites within a field, so that only a limited response to changes in the availability of food is possible during the breeding season. Finally, the methodological uncertainties are notable: the number of Skylarks within a plot is rather difficult to establish accurately, due to their numerousness and their habit of flying around above the field, chasing each other.

Territoriality is not assumed to affect the distribution of Skylarks on the stubble fields in autumn. However, the autumn counts did not reveal any differences in the distribution of Skylarks on the experimental fields that may be ascribed to differences in pesticide treatments. Likewise, no dosage-related differences in abundance were found in Meadow Pipits or small seed-eaters. There are two possible explanations: (1) The incomplete, unbalanced design does not allow sufficiently accurate estimations and powerful tests. (2) The differences in pesticide dosage do not result in differences in the amount of birds' food items on the stubble fields that significantly affect the distribution of birds (cf. section 2.3).

4.4.1 How many more birds at reduced dosages?

From an administrative, as well as from a scientific point of view, this is an essential question. However, it is impossible to give a definite answer to it, among others because no measurements of breeding population densities or production of fledglings were made in this study. Even if this had been the case, though, it would not have been possible to assess the effect on the following year's population size, because this depends on the post-fledging survival, return rate, density-dependent population regulation etc.

A simple answer to the question may be given by simply comparing the mean numbers occurring in the plots over the whole breeding season. However, because it was shown that the differences between dosages do not exist from the outset, but develop during the season, it may be more appropriate just to use data from the second half of the census period, i.e. from 21 June onwards. Such a reduced data set was used to calculate the estimates in Table 4.10.

Table 4.10.
Least-squares estimates of the increase in bird densities on the experimental fields achieved by reducing the amounts of insecticides and herbicides to one-quarter or one-half of normal dosage. Point estimates and 95% confidence intervals are given. Only data from counts performed between 21 June and 5 August (incl.) were used in the calculations.

The choice of cutoff date is quite arbitrary; other dates might equally well have been chosen and would have resulted in a different set of estimates. The effect of changes in cutoff date may be judged from Figs. 4.3 - 4.5.

The figures presented in Table 4.10 should not be misinterpreted. They are estimates of differences in (or rather, ratios between) the number of birds occurring in different dosage plots, i.e. they show to which extent birds of different species prefer to stay (and probably feed) in plots with reduced dosages of pesticides compared to plots with normal dosage. They are not estimates of differences in population size.

It appears from a consideration of the 95% confidence limits that the increases in bird densities are estimated with a sizable degree of uncertainty. This is an inherent problem in the estimation of differences and ratios (because the variance of a difference between two variables is the sum of the two variances), but the interval widths also reflect the great variation among the farms and fields included in the study.

As to the quarter vs. normal dosage comparison, there is nonetheless a remarkable concordance between the estimates based on the morning counts and those based on the afternoon counts. In other words, the point estimates may in this case not be as unreliable estimators of the true values as might be expected from a consideration of the confidence intervals. A quick conclusion goes that a 75% reduction of the dosages of herbicides and insecticides results in a 20-25% increase in the number of Skylarks, a 50% increase in the numbers of small seed-eaters and a doubling of the number of Whitethroats visiting the field.

The effect of a halving of pesticide dosages is more difficult to evaluate. This is largely a result of the experimental design, as it was decided to give priority to a comparison of quarter and normal dosage at the expense of half dosage in those cases where it was impossible to lay out more than two fully comparable plots within a field (cf. section 1.2.5). This was the case in 7 of the 15 experimental fields. Only truly comparable plots were censused on the morning counts (cf. section 4.2.1.1), so these counts should give the more reliable estimates. However, because just 8 half-dosage plots were censused at the morning counts (as opposed to 15 plots with quarter dosage), the half vs. normal comparison is subject to greater uncertainty than the quarter vs. normal comparison.

On the afternoon counts all plots were censused, so the sample sizes used for the half vs. normal and quarter vs. normal comparisons are identical. However, even if only birds occurring in the field proper were recorded, the results are biased due to the lack of full comparability between plots. In the vast majority of "incomparable" cases, the normal- and quarter-dosage plots were bordered by hedgerows but the half-dosage plot was not (Fig. 4.2 B), implying that the half-dosage plot was more favourable for Skylarks but less favourable for small seed-eaters and (especially) for Whitethroats than the two other plots (cf. Table 4.1). Consequently, the afternoon counts overestimates the positive effect of a halving of the pesticide dosages on Skylark numbers but underestimates the gain for Whitethroats and (to a lesser extent) for small seed-eaters.

Despite these uncertainties, and even if the averaged confidence intervals for the half vs. normal comparisons in Table 4.10 in all cases include zero, it must be emphasized that a significant, positive effect of a 50% reduction of herbicide and insecticide dosages has been established for all three species (groups) (cf. section 4.3.1). The true value of the gain in numbers is probably closer to the point estimates derived from the morning counts than to those from the afternoon counts.

5 Yield and economy

(Jensen, A.-M.M., Rasmussen, C., Rasmussen, S. & Esbjerg, P.)

This chapter is dealing with the yield and economy aspects of reduced pesticide use. Firstly, the relation between the weed vegetation and the yield is described for cereals (section 5.1). Secondly, the yield in sugar beets is described (section 5.2). Thirdly, weed problems and clean-up decisions after three years with use of reduced pesticide dosages (section 5.3) are reviewed. Finally, the profitability of use of reduced dosages is considered (section 5.4).

The main aim of the yield trials (sections 5.1 and 5.2) was to provide a measure for compensating the involved farmers for a possible yield reduction each year, not to measure an accumulated decrease in yields. The aim of the clean-up study (section 5.3) was to reveal weed problems arising after three years with reduced use of herbicides and insecticides seen from the agronomical point of view. However, three years are not enough to reveal long-term consequences for the growing practices.

5.1 Yield and vegetation in cereal fields

(Jensen, A.-M.M.)

5.1.1 Purpose

The aim of this study was to analyse the yield responses of spring barley and winter wheat to reduced use of herbicides and insecticides. The reductions performed were 50 % (half dosage), 75 % (quarter dosage) and in some years a 100 % (non-sprayed) reduction of the normal dosage used in each field and crop. Weed plants compete with crop plants for water, nutrients and light in particularity. Therefore, it is expected that a high density of weeds at reduced dosages is correlated to a decrease in yield. In this study the weed vegetation measured as weed density and species richness was related to the yield and yield quality at reduced pesticide use. A higher density of insects at reduced dosages might also affect the yield, but the occurrence of insects was not measured in this study, so yield changes could not be correlated to the presence of insects, unfortunately.

5.1.2 Methods

5.1.2.1 Field design

Thirty field trials were performed, one in each field of spring barley and winter wheat in each of the three study years in the main study. One field trial in winter wheat was omitted because no differentiation in application of herbicide dosage was made. Each field trial (size: 10m x 40m) was situated in a corner of the field within the plot sprayed with normal dosage at least 30 meters from other plots, hedges, habitat islands etc. and surrounded by a 2.5 m wide buffer zone.

All field trial areas were treated exactly as the surrounding field - sown the same day, sprayed with same dosages and products on the same day etc. (Appendices A.2-A.4). Each field trial consisted of four blocks (replicates) and each block consisted of 3 plots in 1997 and 4 plots in 1998 and 1999. Each plot measured at least 10 m x 2.5 m but they were sometimes longer. The plots within a block were sprayed with different dosages of herbicides and insecticides (Fig. 5.1). The dosages were 1/1: Sprayed with normal dosage of herbicides and insecticides, 1/2: Sprayed with half dosage, 1/4: Sprayed with quarter dosage, 0: Not sprayed. Non-sprayed plots were not included in 1997.

Fig. 5.1.
Design of one field trial. In 1998 and 1999 each field trial contained 16 plots. In 1997, the design consisted of 12 plots, because non-sprayed plots were not present. I to IV are the blocks, 0, 1/1, 1/2 and 1/4 the dosages. The squares in the middle of the plots illustrate the weed investigation areas.

Products and dosages chosen by the farmer as normal dosage were very different from field trial to field trial and not equal recommended dosages (chapter 1 and Appendices A.2-A.4). The farmers chose products and dosages after their experiences and individual needs. The location of a field trial within a field was not totally fixed, so the same plot did not necessarily receive the same level of herbicides and insecticides throughout the three years. Therefore, cumulative effects of reduced dosages could not be expected.

5.1.2.2 Weed observations

In each plot, the weed flora was registered in an area of 0.6 m x 0.4 m (Fig. 5.1). All plants were identified as described in section 2.2.1.6 and counted, giving number of plants per square meter (density) and number of species per 0.24 square meter (species richness). The areas were fixed within a season and were visited in spring (early May) and in summer (late July) (before and after spring sprayings). Care was taken to avoid damaging the weed and crop plants under the investigations.

5.1.2.3 Yield and yield quality measurements

Each plot in each field trial was harvested separately and the yields were measured by the local advisory services. Yield was adjusted to 15 % moisture content and given as tons per hectare. Moreover, the following yield components and quality parameters were measured by the local advisory services: thousand kernel weight, moisture content, content of protein, content of starch, pureness and sorting. Samples from plots sprayed with same dosage within a field trial were pooled before measuring the quality parameters.

5.1.2.4 Treatment intensity index and efficiency

In order to compare dosages of different pesticides, two treatment intensity indexes were calculated: an index for herbicides alone and an accumulated index for herbicides and insecticides together. Herbicides applied to control broad-leaved species were included in the treatment intensity index for herbicides, so products mainly applied to control Elymus repens and Avena fatua were not included. Treatment intensity indexes were calculated for normal dosage in each field trial (Table 1.1) according to Danmarks Jordbrugsforskning & Landbrugets Rådgivningscenter (1997, 1998 and 1999).

Efficiency of a treatment is in this chapter defined as the herbicide treatment's ability to reduce the weed density and is calculated as (n₀-n₁)/n₀, where n₀ is the total density of weed plants in spring (before spring spraying) and n₁ is the total weed density 2-3 months after spring spraying.

5.1.2.5 Statistical tests

The following response variables were tested for dosage effects, separately: weed density, species richness, efficiency, yield and the yield quality parameters. Data from spring barley and winter wheat were tested separately, because the crop has a profound effect on weed density and species richness (see sections 2.2.2.1 and 2.2.2.2). The basic experimental unit in the analyses of these variables was the plot (n=220 for spring barley and n=208 for winter wheat).

Before running the analyses, weed density was transformed with the natural logarithm (log_e (y+1)) to achieve a normal distribution and homogeneity of variances. Similarly, the species richness, efficiency, yield and quality parameters were square root, exponential or arc sinus transformed to approach a normal distribution, if necessary.

The statistical test used was analysis of variance with farm, year, dosage and interactions between these as explanatory factors in addition to the block within a field trial. Moreover, in the analysis of weed density after spraying, the weed density before spring spraying was used as a covariate. Similarly, the species richness before spring spraying and the weed density after spraying were used as covariates in the analyses of species richness after spraying. In the analysis of yield, weed density, species richness and efficiency were used as covariates. In addition, the six yield quality parameters were tested for effects of pesticide dosage using mean weed density after spraying, mean species richness after spraying and mean yield per dosage per field trial as covariates and farm, year, dosage and first order interactions between those as explanatory factors. At last, the yield in each field trial was analysed separately. In these analyses the explanatory factors were: dosage, block, weed density after spraying and species richness after spraying.

The number of explanatory factors was reduced during full-scale model calculations using an iterative procedure to remove the variables with p > 0.10 until the model consisted only of variables with p £ 0.10.

Each analyse was run three times; first with the interaction between farm and year and afterwards with the two treatment intensity indexes replacing the farm´ year interaction in the reduced model. This was done to clarify if the variations between field trials could be explained by variations in the treatment intensity indexes.

All analyses were performed using the general linear model (GLM) procedure in SAS (SAS Institute 1999). If a significant effect of dosage was revealed, the differences among dosages were tested by a Tukey-Kramer test for a-posteriori comparisons of least-squares means. Because non-sprayed plots were not present in 1997, the design was unbalanced and thus required modifications in the F-tests by using the Random/Test statement in the GLM procedure.

The proportion of variation explained by a certain factor was calculated as the sum of squares for that particular factor divided with the total sum of squares.

5.1.3 Results

5.1.3.1 Dosage effects on overall weed density, species richness, efficiency and yield

Table 5.1.
Results of tests of effects of reduced dosages on the weed density, species richness, efficiency and yield in spring barley and winter wheat. Grey areas indicate factors not included in the full model. Statistical significance is indicated as follows: +: 0.05£ p<0.10, *: 0.01£ p<0.05, **: 0.001£ p<0.01 and ***: p<0.001.

Look here!

Weed density was strongly affected by dosage, where a decrease in dosage resulted in an increase in density of weed plants in both spring barley and winter wheat (Table 5.1 and Fig. 5.2A). The estimated mean weed density in non-sprayed spring barley plots was 177 plants/m², which was 33 % higher than in plots sprayed with quarter dosage (p=0.0001). In winter wheat, the estimated mean weed density in non-sprayed plots was 96 plants/m², which was significantly higher than in plots sprayed with quarter dosage (p=0.020).

In both crops, species richness was influenced by dosage, but the effect varied between farms (Table 5.1 and Fig. 5.2B). The species richness was highest in non-sprayed plots; 6.2 species per 0.24 m² in spring barley and 3.3 species per 0.24 m² in winter wheat.

As expected, the efficiency of the treatments was highly affected by dosage (Table 5.1 and Fig. 5.2C). In spring barley the efficiency was 25 % at quarter dosage, which was significantly lower than the 35 % at half dosage and the 40 % at normal dosage (p=0.043 and p=0.0003, respectively). No significant differences were found between normal and half dosages in spring barley (p=0.46). In winter wheat normal dosage was almost significantly more effective in weed control (46 %) than quarter dosage (28 %) (p=0.053), no significant differences in efficiencies were found between normal and half dosage (31 %) (p=0.16).

Fig. 5.2.
Weed density, species richness, efficiency and yield of spring barley and winter wheat at normal and reduced dosages of pesticides. Each bar represents a mean of 40-60 values. Error bars indicate the 95 % confidence limits.

As usual in Denmark, the yield per hectare was much higher in winter wheat than in spring barley (Fig. 5.2D). There was a significant effect of dosage on yield of spring barley (Table 5.1). Non-sprayed spring barley plots had on average 8 % lower yields compared to sprayed plots (p<0.0001 in all pairwise comparisons). The mean yield at quarter dosage was significantly lower than at half dosage (p=0.043) and normal dosage (p=0.015). No differences in yields were found between plots sprayed with normal and half dosage (p=0.92). The yield in winter wheat was significantly affected by the dosage, but the effect varied between years and farms (Table 5.1). Yields in non-sprayed plots were on average 7 % lower than in plots sprayed with normal dosage (p=0.0002), whereas yield did not vary significantly between normal, half and quarter dosage (p>0.31 in all pairwise comparisons).

Table 5.2.
Percentage increases (+) and decreases (-) in weed density, species richness, efficiency and yield at reduced or zero dosages compared to mean values at normal dosage. Significant difference from normal dosage is indicated as follows: +: 0.05£ p<0.10, *: 0.01£ p<0.05, **: 0.001£ p<0.01 and ***: p<0.001.

Look here!

The effects of reduced dosage on weed density, species richness, efficiency and yield are summarised in Table 5.2.

All response variables showed a highly significant effect of the interaction between farm and year (Table 5.1), which corresponds to the variation between individual field trials (included among other things variations in location of the field trial and the treatment intensity index). This interaction explained most of the total variation in most response variables. Furthermore, the dosage effect varied considerably between field trials (Table 5.1).

When the treatment intensity index for herbicides replaced the interaction farm´ year in the reduced models in spring barley, the models were not improved (they showed a higher p-value). Despite this, the treatment intensity index for herbicides had a significant negative effect on weed density and species richness and a positive effect on the efficiency. The yield in spring barley was not significantly affected by the treatment intensity index. In winter wheat models the p-values decreased, when the treatment intensity index for herbicides replaced the farm´ year interaction. Thus, the differences in treatment intensity index for herbicides could explain the variations between field trials. The treatment intensity index for herbicides had a significant negative effect on weed density, species richness and yield and a positive effect on the efficiency. In the case of winter wheat, the result thereby indicates that the higher the treatment intensity index for herbicides the lower the yield at harvest, which is in contrast to the expected results. The yields at normal dosage in both spring barley and winter wheat are illustrated in Fig. 5.3 as a function of the treatment intensity index for herbicides. The figure indicates that no meaningful relationship exists between yield and treatment intensity index. For both crops, all models showed a higher p-value if the accumulated treatment intensity index for herbicides and insecticides replaced the farm´ year interaction.

Fig. 5.3.
Relationship between treatment intensity index for herbicides and mean yield at normal dosage. Each dot represents data from one field trial.

5.1.3.3 Yield quality parameters

Of the six yield quality parameters only moisture content in spring barley grains was significantly affected by dosage (p=0.042): The moisture content was significantly lower at quarter dosage, than at half and normal dosages (Tukey-Kramer tests, p=0.0006 and p=0.003, respectively), and no significant differences in moisture content were found between non-sprayed and sprayed plots (p=0.98, p=0.10 and p=0.20 for pairwise comparisons with quarter, half and normal dosage, respectively). No significant effect of dosage was found in any of the other five quality parameters in neither spring barley nor winter wheat.

Look here!

Fig. 5.4.
Yield changes at reduced dosages (0.5, 0.25 and 0) in percent compared to normal dosage (1) shown for all field trials. 0 on the y-axes corresponds to yield at normal dosage.

The total weed densities before and after spring spraying and the dominant weed species varied considerably from field trial to field trial (Appendix E). Yield changes at reduced dosages of pesticides compared to yield at normal dosages are shown in Fig. 5.4 for all field trials in winter wheat and spring barley.

When each field trial was analysed separately, yield was significantly affected by the dosage in only 8 of 29 field trials (Appendix E.2). In six cases, the yield was significantly lower in non-sprayed plots than in sprayed plots. In three field trials, the yield at quarter dosage was significantly lower than at normal dosage. No field trial showed a significantly lower yield at half dosage compared to normal dosage. The simple model explained at least 75 % of the variation in yield in 86 % of the field trials (Appendix E.2).

5.1.4 Discussion

5.1.4.1 Dosage effects on yield

These field trials cover a wide range of weed species and densities (Appendices E.1A and B) and in the majority of cases a yield decrease (Fig. 5.4) at reduced dosages was found. However, the decreases were often small and only statistically significant in 28 % of the field trials. These results are in accordance with Davies et al. (1989) and Salonen (1993b), who found it difficult to show yield response to reduced broad-leaved weed control in winter wheat and spring barley in Scotland and spring barley in Finland, respectively. Also Pallutt (1999) showed that application of quarter of recommended herbicide dosage was enough to prevent yield losses in strong competitive cereal stands in Germany.

It was generally not possible to show significant differences in the yield between plots sprayed with normal, half or quarter dosages, whereas the yield at unsprayed plots decreased with 8 % in spring barley and 7 % in winter wheat. Davies et al. (1989) showed that 1/8 of recommended dosage in some cases gave a yield reduction although it was not significant when a mean of five experiments was considered (Davies & Whiting 1990). In some experiments on clay soil Salonen (1992b) reported a higher yield at reduced herbicide dosage than at recommended dosage. This is also found at both half and quarter dosage in some field trials in this study (Fig. 5.4). Jensen (1986) suggested that recommended dosages of some herbicides could harm the crop, however other herbicides are applied today than in the early eighties, so this might not be the case today, although also modern herbicides in some field trials result in a higher yield at reduced dosages than at recommended dosages (Landbrugets Rådgivningscenter 1999b). Yield decrease in response to no chemical control is lower on clay soils than on sandy and organic soils (Jensen 1985). This could be one of the reasons why only small decreases in yield are found in these field trials on clay soils. From this study, it is clear that dosage can be halved without notable yield decreases within one season.

The yield decrease recorded in non-sprayed plots of spring barley (8 %) and winter wheat (7 %) was almost the same size although winter wheat has a higher competitiveness than spring barley (Rasmussen et al. 1997). Therefore, it was expected that winter wheat would have shown a smaller decrease in yield at reduced dosages. On the other hand experiments performed in winter wheat have shown a yield decrease at 12 % (Davies et al. 1989) or 15 % (Wilson 1986) whereas experiments in spring barley have shown a decrease of 0 %-7 % (Courtney & Johnston 1986, Davies et al. 1989, Salonen 1992b, Salonen 1993a). Therefore, no consistent relation between crop competitiveness and yield response seems to exist. In addition, the competitiveness of crop cultivars is very variable (Valenti & Wicks 1992, Grundy et al. 1993) and confuses the picture even more.

Yield responses to reduced dosages varied from farm to farm and from year to year (Table 5.1 and Fig. 5.4), indicating that there may be differences in the effectiveness of weed or insect control between farms and between years within a farm.

From this study, it must be concluded that no meaningful relationship exists between treatment intensity index and the yield (Fig. 5.3).

5.1.4.2 Weed density and yield

Large yield reductions might be associated with high densities of weeds. However, these field trials showed no significant effect of weed density after spraying on the yields of spring barley and winter wheat.

In the separate analyses, only 14 % of the field trials showed a significant effect of weed density and richness on yield (Appendix E.2), half showing a negative relationship and half a positive relationship. These field trials (showing yield decrease above 4.5 %) were not the ones with the highest weed densities after spraying (Appendix E). One field trial (spring barley at Oremandsgård 1997) had a mean density over all treatments of less than 10 weed plants after spraying, and still showed a significant yield decrease at reduced dosages. In that field trial, the yield reduction might reflect a higher level of herbivorous insects at reduced dosages, although the occurrence of aphids was rather low in the surrounding field (Appendix D).

Many studies, like this, have made it clear that there is a poor correlation between total weed density and yield (Jensen 1991, Salonen 1992b), weed biomass and yield (Salonen 1993a) and weed cover and yield (Fischer et al. 1993). This might be the reason why it has been difficult to show yield response to reduced broad-leaved weed control in winter wheat and spring barley (Davies et al. 1989, Salonen 1993a).

Not all weed species are strong competitors with cereals. Some species have a higher competitive ability than others, measured as the ability to suppress the crop biomass (Wilson 1986, Wilson & Wright 1990, Jensen 1991, Jensen 1996). As suggested by Wilson (1986), yield responses could be influenced more by weed species than by weed densities. Today’s weed control strategy is therefore based on both species and their density (Rasmussen et al. 1997).

In spring barley but not in winter wheat, weed biomass was better correlated to yield loss than the weed density (Jensen 1996). Therefore, a possible difference in responses of density and biomass might not alone explain the reason why no clear correlation between yield and weed density was found in this study (Table 5.1). It can be seen from Appendices E.1A and B that barley field trials with high densities of Brassica napus/Sinapis arvensis, Polygonum aviculare and Trifolium sp. had a large decrease in yield at low dosages. Especially Sinapis arvensis is known as a strong competitor (Scragg 1980). No such species, accounting for major decreases in yield, can easily be found in winter wheat. Among the weed species found in this study, Galium aparine is the most competitive (Wilson & Wright 1987) but the highest yield decreases did not coincide with the presence of Galium aparine. In conclusion, it must be said that in field trials with high densities of weed species with a high competitiveness reduced dosages might have the greatest potential for affecting the yield negatively. In none of these field trials populations of strong competitive grasses as Apera spica-venti, Avena fatua, Bromus sp. or Elymus repens were observed. These species might even in low densities be able to reduce the yield.

In both spring barley and winter wheat, significantly higher weed densities were found not only in non-sprayed plots but also in plots sprayed with quarter dosage than in plots sprayed with norma1 dosage (Fig. 5.2A). Thus, strongly reduced herbicide dosages did not keep the weed densities at the same level as the normal dosages did as reported from Finland (Salonen 1993a).

After spraying with reduced dosages the surviving weed plants might be weakened and are therefore not able to compete with the crop. This might explain that the yield did not decrease remarkably even when the weed density increased highly at reduced dosage of pesticides.

5.1.4.3 Dosage effects on efficiency

In many short-term experiments with cereals, reduced herbicide dosages have provided adequate control of broad-leaved weeds (e.g. Davies et al. 1989, Davies et al. 1993, Salonen 1993a) without notable yield decreases. The efficiency was as expected increasing with increasing dosage. At normal dosage, the efficiency was on average 49 % in winter wheat and 41 % in spring barley. This is a low efficiency compared to efficiency of recommended dosages, which reach 70 % in 78-91 % of the field trials performed by Salonen (1993a). The yearly variation in herbicide efficiency may be very large. One year Derksen et al. (1995) found a 90 % reduction in weed number, next year the reduction was only 39 % despite the field, the product and the dosage being equal.

The biomass reduction is even higher than the density reduction mentioned here (Salonen 1993a). This means that weed plants have a lower average biomass in sprayed plots than in unsprayed plots. Furthermore, figure 2.12 indicates that the average biomass per plant was higher at reduced dosages than at normal dosages because a plant weighs more in the generative than in the vegetative phase. The long time (2-3 months) between spring sprayings and registration of the weed density after spraying may have influenced the efficiency both negatively and positively. If new plants have germinated within that period, it would be reflected in a lower efficiency, whereas competition through most of the growing season from the cereals might have resulted in dead weed plants reflected in a higher efficiency. Not many seedlings were observed in cereals in July, so the latter might be more likely than the first.

Many other factors might affect the efficiency of the herbicides (e.g. weed composition, the size of the weed, technique used for spraying, the climate around the time of spraying, the competitiveness of the crop (Kudsk & Mathiassen 1991) and water stress (De Ruiter & Meinen 1998)).

5.1.4.4 Dosage effects on yield quality parameters

Statistically significant differences in yield parameters between dosages were only detected in one case. It must be concluded that dosage has little, if any, effect on yield parameters as content of moisture, starch, protein, thousand kernel weight, pureness and sorting. However, Davies and Whiting (1990) have shown that the higher the cover of Stellaria media, the higher the moisture content in the grains at harvest. The effects of weeds on yield quality are proportionately much smaller and can not occur independent of the effects on yield (Whiting et al. 1991), so it might not be very important to focus on yield quality. Other factors varying between farms and years had a much greater influence on the yield quality parameters than dosage.

5.2 Yield in sugar beets

(Esbjerg, P.)

In contrast to the preceding section concerning yields in cereals which also comprise botanical elements, which are a trade off of having mini-plots, this section solely comprises establishment of a background for the landowners yield situations. This is in essence the basis for the payment in a few cases of compensation for losses causes by the lowered dosages.

After the problems with harvesting within mini-plots of sugar beet at Gjorslev estate in the pilot phase (cf. section 1.2.9) the work on yield estimates was changed to a sampling procedure following principles for sampling aiming at estimating effects of variety, fertilization etc. The procedure was selected on the basis of recommendation from "The Foundation for Sugar Beet Research". In the present case the aim was to reveal if the effect of reduced dosages was reflected in the yield levels.

5.2.1 Sampling

In each of the three dosage plots per farm a total of 20 random selected samples were dug up manually.

Each sample consisted of 4 metres of the row at a particular position determined by the random selection procedure. The beet tops were removed in the field by standardized cutting in accordance with instructions from "The Foundation for Sugar Beet Research". This research body also took care of washing, weighing and labelling of samples for subsequent analysis of sugar content. The last part was carried out at the sugar factory in Maribo. The sugar content was only determined as safety measure but it proved to be much less meaningful than the weight due to the in-field type of variation. In addition the prices were estimated according to the type of contract with the sugar factory and the amount and quality of the harvest of the whole field. Therefore the estimated weight yields formed the best background for evaluating whether or not the landowners had losses to be compensated.

5.2.2 Yields estimated

The results are presented in Table 5.3. It appears that significant yield losses are very few and with one exception only occur at quarter dosage. In summary in 1997 Nordfeld lost 12 % of the yields at half and 14 % at quarter dosage but part of this loss may have been caused by troubles with the new equipment for mechanical hoeing under slippery conditions at sloping terrain. (During the two next seasons this problem was counteracted by computerised steering). In 1998 Lekkende lost 25 % at quarter dosage and in 1999 no significant losses were found, except in quarter dosage at Nøbøllegård where spots of weeds, mainly common couch Elymus repens, were known as problematic at that particular area. However, it would have been disturbing for the project to permit the solving of the problem by use of full dosage of Glyphosate in stubble before growing beets.

Table 5.3.
Yields estimated in sugar beets. It is important to notice that during the pilot phase at Gjorslev 1996 the dosage plots were all treated with the particular dosage applied as broad swath. In addition the yields at different dosages were estimated by producing mini-plots: During the subsequent years the dosage reductions were obtained by treating only bands over the rows and for yield estimations random samples were used. Every yield marked * is significantly lower than the corresponding yield of the normal dosage plot shown in the same line (the same farm and the same year).

5.3 Clean-up decisions

(Jensen, A.-M.M.)

5.3.1 Purpose

In this study, effects on the weed vegetation by use of reduced dosages of herbicides and insecticides during three years were emphasised.

The objective was threefold: 1) To spot major weed problems after three years with reduced pesticide use. 2) To estimate economic compensations, so the farmers could bring the areas sprayed with reduced dosages in the same state as areas sprayed with normal dosage with regard to weed populations. 3) To estimate compensations for future yield decrease as a result of accumulated weed populations.

5.3.2 Method

In July 1999, observations during field walks were made by an agricultural adviser from the local advisory service, a botanical investigator and the farmer in each of the three fields used in the main study on each of the five farms. Each of the 15 fields was observed for about an hour. Weed species with a higher density or abundance in plots sprayed with reduced dosages compared to plots sprayed with normal dosage were spotted and the area in which they were present was estimated. Most species occurred in spots within the estimated area of the plot. The possibilities for reduction of these populations as well as a price estimated for the future control strategy were discussed. In addition, economic compensations for presumable yield decrease in 1999 and 2000 were estimated.

5.3.3 Results

Different weed species were observed in the different fields (Table 5.4). In total, 19 weed species increased in densities in plots sprayed with reduced dosages, 12 of these species were spotted in more than one of the 15 fields. Of these species, eleven were broad-leaved and one was a grass: quackgrass (Elymus repens). The perennial Cirsium arvense was the species observed in most fields, although it was present in small areas compared to annual species as Galium aparine and Sinapis arvensis.

No clear differences in species with increased density and the area they covered were observed between half dosage and quarter dosage.

Table 5.4.
Weed species with increased abundance or density within plots sprayed with reduced dosages. The estimated area of the 6 ha-plot(s), where the species was present, is given in percentage. Prices estimated for cleaning the fields from these weed species are given in Table 5.8.

Elymus repens was the only species that caused heavy yield decreases after three years with reduced dosages (Table 5.5).

Table 5.5.
Estimated yield decrease in 1999 and 2000.

5.3.4 Discussion

Many weed species appeared in increased densities, but only the high density of Elymus repens caused heavy yield reductions in some areas. Therefore, in this study, reduced dosages of glyphosate (compound used to control Elymus repens) have economic consequences already after few years. Most broad-leaved weed species increased in densities without causing remarkable yield decrease after three years with reduced use of pesticides. The reason why broad-leaved species did not cause major yield reductions might be that the weed infestation at the beginning of the main study was very low, which are also found in other experiments running more than two years (Erviö et al. 1991). Thus, the accumulation of the broad-leaved weed population during three years has probably not yet reached a level, where it reduces the yield remarkably. Further increased densities of the broad-leaved species mentioned in Table 5.4 may result in yield decreases at reduced dosages over a period longer than three years.

Three other studies controlling broad-leaved weed species with reduced dosages of herbicides during some years have shown that: 1) Recommended dosage applied one year out of three was as effective in controlling broad-leaved species as half dosage in three years (Skorda et al. 1995). 2) Half of recommended dosage each year or recommended dosage two of three years maintain the weed density at a stable level (Jensen 1991). 3) No difference in the broad-leaved weed density or the yield was found between the following three treatments with reduced pesticide application over three years in three places: spraying in year three, spraying in years two and three or spraying in all three years (Courtney & Johnston 1986). Thus in regard to the yield and the control of the weed species, use of reduced dosages of herbicides might be performed as reduced dosages every year as well as no spraying one year followed by spraying with normal dosage the next year. However, this might not be the case, where weed plants with high competitiveness are present in high densities, then half dosage might be preferred each year to keep the yield decrease low over the years (Landbrugets Rådgivningscenter 2000b). No experiments have studied the effects on the weed and animal diversity of the latter senario, which is very important before an evaluation of the different strategies can be made.

The yield decrease at reduced dosages is in a short term of minor economic and quantitative importance whereas the demand for a harvest without problems and the long-term aspects are very important. Especially, a build-up of seed reserves in the soil is a strong concern to farmers. No significant differences between pesticide dosages were found in the seed rain after three years (section 2.3) or in the germinating weed density after two years (Fig. 2.8). Also Salonen (1992a) could not detect statistically increases in the number of weed seeds in the soil seed bank after continuous use of on third of recommended herbicide dosages for three years. This indicates that dosage differences in the soil seed bank and the seed rain are small in comparison with the differences caused by farms, crops, land-use history, the great weed community variations within fields etc. However this qualitative clean-up study seems to indicate that an increased density of weeds may cause more seeds in the seed bank. Three years study is not enough to detect severe consequences of use of reduced dosages from the agronomical point of view - more years are needed. The extent of changes in weed population dynamics as a result of low pesticide dosages usually shows only after three to ten years, and can therefore be safely determined only in long-term trials (Pallutt 1999). Few long-term trials with reduced herbicide dosages have been performed yet (Jones et al. 1997, Pallutt 1999) both showing increased weed densities at reduced dosages either as weed seeds in the topsoil (Jones et al. 1997) or as infestations of weed species with strong competitiveness (Pallutt 1999).

The results are not based on a scientific documentation of increase in density but reflect species that mainly were spotted due to their height or the farmers' worries. These species are, anyway, weed species with a high competitiveness and therefore species that may cause economic problems over a longer period with reduced dosages of herbicides. Because this study was mainly qualitative, it could not detect the quantitative differences in weed density between half and quarter dosages as found in the main study (chapter 2).

It is a problem that the field walks did not take place during the late summers of 1996, 1997 and 1998. It is thus not possible to conclude whether or not the weed densities have accumulated over the three years. The differences in weed vegetation between normal dosage and reduced dosages might have been noticeable after one year only, as the main study indicates (Fig. 2.8), even though the total weed density has increased in all dosages during the three years.

5.4 Profitability analysis

(Rasmussen, C. & Rasmussen, S.)

5.4.1 Purpose

This section presents and discusses the results of the profitability analysis of reducing dosages of herbicides and insecticides. The aim of the study was to estimate the profitability consequences in crop production when dosages of herbicides and insecticides are reduced. The objective was: 1) to estimate the impact on short-term profitability and 2) to give an example of long-term profitability effects of a dosage reduction.

5.4.2 Method

A spreadsheet budgeting model was developed for the three crops: winter wheat, spring barley and sugar beet, calculating profit for each of the three crops in three different pesticide scenarios. It is assumed that farmers are interested in maximizing net returns or profitability per hectare. Profit is the total value of the product less the total factor cost, given fixed product and factor prices. Profit is calculated by aggregating the net returns for each season, and one season is defined as one year.

Total value of the product includes for wheat and barley both grain and straw value, but for sugar beet the total value includes only beet value, which depends on the sugar content (%). It is assumed that reduced pesticide dosages do not affect the straw production. Sugar beet prices depend on sugar content meaning that a varying content of sugar may contribute to differences in the profits calculated in the three scenarios (see section 5.4.3 for results).

Total factor costs are divided into two separate measures categorised as costs I and costs II. The costs I category includes costs of seeds, NPK fertilisers, additives and all pesticides. The costs II category includes the costs of labour and machinery, both of which are calculated using machine pool standard prices. All product and factor prices are standard prices stated by Landbrugets Rådgivningscenter (1998, 1999a and 2000a) and Danmarks Jordbrugsforskning & Landbrugets Rådgivningscenter (1998, 1999 and 2000). Developments in product and factor prices are important causes of differences in profit from one year to another. Price developments are stated in the literature above.

The model accounts for the following input factors: NPK fertilisers, herbicides, insecticides, fungicides, additives and growth regulating additives. Theoretically the output produced varies with the dosage of these input factors. Please note that only the dosages of herbicides and insecticides vary between the three scenarios. Furthermore it is important to note that the other input factors are not necessarily added in the same dosages on each farm. Apart from the variation in input factors, the treatments of the crops also varies between the five farms. Because of the variation in input factors, dosages and treatments the reference scenarios (labelled scenario A) are not identical for the five farms. Even though the reference scenarios are not identical, it seems probable that the change in profit is related to the reduction in the dosages of herbicides and insecticides. All other input factors are kept constant on each farm with the exception of sugar beet production where mechanical weed control also varies between the scenarios.

5.4.3 Results

Three scenarios differentiated by the dosage of insecticides and herbicides are considered in the model. Scenario A corresponds to normal dosage. Scenario A is not necessarily the optimal dosage that maximizes profit but serves as a reference scenario.

Compared to the dosage used in the reference scenario (A), scenario ½ is characterised by a 50 % and scenario ¼ by a 75 % percent reduction in the dosage of herbicides and insecticides.

Results of the economic analysis are presented in Table 5.6. In scenario A, the average profit for the five farms is calculated for each of the crops winter wheat, spring barley and sugar beet. Note that results for scenario ½ and ¼ are shown as both the absolute and the relative change in profit compared to the reference scenario.

More detailed results are presented in Table 5.7 showing the figures from Table 5.6 broken into product value, costs I and costs II. Detailed results for each farm are presented in Appendices F.1-F.5.

Table 5.6.
Calculated profit in current prices (DKK per Hectare) and changes in relation to A at scenarios ½ and ¼.

Table 5.7.
Change in product value, costs I, costs II and profit with a ½ and ¼ dosage of insecticides and herbicides in the period 1997-1999. Results are presented in DKK per hectare in current prices.