1

RABBIT MANAGEMENT FOR GROWERS OF SHORT ROTATIONWILLOW COPPICE

Julie Dendy, Gordon McKillop and Richard Watkins.E-mail: j.dendy@csl.gov.uk

Central Science Laboratory (CSL)

INTRODUCTION

Short rotation coppice (SRC) is vulnerable to damage by a range of pest species.However, damage by rabbits (Oryctolagus cuniculus) has been identified as one of themain factors which can affect the economic success of the crop, in some casesrequiring replanting of the coppice (Beale & Heywood 1997). Rabbit numbers areincreasing throughout the country and, given the planned increase in areas of SRC tobe grown in the future, the threat rabbits pose to these new plantations is considerable.

This paper provides a synopsis of a recently published advisory leaflet ‘Advice onrabbit management for growers of short rotation willow coppice’ commissioned bythe DTI’s New & Renewable Energy Programme (McKillop & Dendy 2000). Its aimis to provide growers and advisors with information on the potential impact of rabbitson SRC plantations and advice on how to manage problem rabbit populations.

RABBIT NUMBERS

The most recent surveys suggest that there may be as many as 30 million rabbits inthe UK and numbers are still increasing at a rate of 2% per year (Smith & Trout1994). Rabbits are estimated to cause around £100 million damage to the UK’sagricultural interests each year, a figure that does not include their impact on forestryor amenity land. As the impact of the lethal virus, myxomatosis, on the populationwanes, so numbers will continue to increase. Rabbit numbers are already at pre-myxomatosis levels in some areas, but are, overall, at about 30% of these levels.Accordingly, the potential for damage to SRC plantations is now very high in someareas and increasing in others.

RABBIT BIOLOGY

Adult rabbits weigh between about 1.2-2.0 kg and can eat up to 30% of their bodyweight in food in a single night (McKillop et al. 1993). Their breeding season extendsfrom January to August, when a succession of litters, usually of 3 to 7 young can beproduced by the does at 30 day intervals. Does reach sexual maturity from as early as4 months old. Despite this prodigious productivity, only 10% of the young willsurvive longer than twelve months (Cowan & Garson 1985).

The size of a local population, and therefore the risk to a plantation, will be dependenton soil type, climate, the number of predators and the availability of cover. Rabbitsare often more numerous on sandy than clay soils and are more abundant in thewarmer and drier south-east of the UK. This may be the result of high mortalityamongst the young in areas prone to water-logging; the young are particularlyvulnerable to the effects of hypothermia (Trout & Smith 1995). Numbers also tend tobe higher in areas where predator control is carried out, such as on shooting estates.

2

Warrens are a precious resource for the rabbit as both a refuge and breeding site.Consequently adult rabbits rarely move more than 200m from their burrows and oftenuse well-defined paths or runs within their home range (Cowan et al. 1987). However,juvenile rabbits can disperse, usually at the end of the breeding season, over distancesof up to 4km. These individuals do present a threat to SRC, even if the plantation issited in what was originally a relatively rabbit-free area. Growers should therefore bevigilant to this threat.

RABBIT DAMAGE

Rabbits often damage trees, particularly newly planted ones, by browsing or by ringbarking. Browsing is the most common form of feeding damage on young trees andcan occur up to a height of 0.5m (higher in snow). It occurs most often in winter andspring and can be identified by a characteristic sharp-angled cut on the end of thestem, with the removed portion often being eaten (Pepper 1998). Ring barking andbrowsing damage can either kill the young trees or reduce their vigour, making themmore susceptible to other pests and diseases or environmental stress (e.g. drought). Ifthe damage is extensive then replanting will be necessary with obvious implicationsfor the profitability of the crop.

ASSESSING THE SIZE OF RABBIT POPULATIONS

Estimates of the distribution and abundance of rabbits in areas to be planted can beobtained by counts carried out at night using a spotlight, or at dawn and dusk undernatural light. These techniques only give an index of population size rather than actualnumbers present. Nevertheless they provide a useful measure and enablescomparisons to be made between sites regarding the potential for rabbit infestation.The best time of year to do this type of monitoring is during the winter months priorto planting, as over-wintering populations at any site are subject to less seasonalvariation between years than summer populations. Monitoring of rabbit abundanceshould be undertaken before and after a control programme to assess its effectivenessand to determine whether further treatments are necessary.

However, growers may not always have time to conduct a series of night-time countsand in newly planted coppices it may be difficult to spot the rabbits amongst thewillow shoots. Further work is therefore needed to develop simple census methodsthat can be conducted quickly and reliably based upon, for example, signs of rabbitoccupation such as faecal pellets, burrows or scrapes.

RABBIT MANAGEMENT

If significant numbers of rabbits are present then management programmes need to becarried out to remove rabbits from the proposed SRC site and adjacent land. Rabbitpopulations, because of the species prodigious reproductive capacity, can withstandhigh mortalities. Recent work completed in Australia suggests that over 80% of thepopulation should be removed if the grower is to achieve a sustained reduction inrabbit abundance (Twigg et al. 2000).

The most effective time to remove animals is between November and March whennatural mortality will have reduced rabbit numbers to their lowest point in thepopulation cycle (Tittensor 1981). Action at this time will reduce the adult breeding

3

population before the next breeding season begins.

A range of lethal and non-lethal methods are currently available to manage rabbitpopulations (Table 1). Choice of method will vary depending on cost-effectivenessand circumstance. Often growers use several methods in combination, but littleinformation is currently available on the best order in which they should be applied. Anumber of computer-based pest management support systems are already underdevelopment. Landowners or wildlife advisors will be able to feed in information ontheir pest problem and these systems provide them with a site-specific managementprogramme. The programmes detail the most cost-effective combination ofmanagement methods to use, the order in which they should be applied and theoptimal time for application. Should information on the costs of rabbit damage towillow plantations become available, similar systems could be developed for SRCplantations; systems that would enable growers to plan and cost rabbit managementprogrammes for their sites.

CURRENT RECOMMENDATION FOR GROWERS OF SRC

Choice of site for SRC plantation

Where possible establish plantations

• On heavier soils.• At least 200m away from existing infestations.• Away from woodland.Assessing size of rabbit populations

• Estimate rabbit numbers using dawn/dusk or night counts.• Estimate numbers before and after management programmes to determine

whether they have been effective.

Rabbit management programmesIf significant numbers of rabbits are present• Remove rabbits from plantation site and adjacent land (see Table 1)• Take steps to prevent re-invasion (e.g. protect plantation using wire netting or

electric fencing).

DISCUSSIONDiscovering signs of damage or rabbit infestation does not necessarily mean thatprotective measures need to be taken. The decision should be objectively based uponeconomic and ecological costs and benefits. However, at present, little research hasbeen conducted to quantify the impact rabbit feeding damage has on yield. Researchrelating rabbit numbers to extent and costs of damage to SRC is needed along thelines of similar work that has already been conducted for arable farmers. This researchshowed that rabbit grazing on winter wheat, for example, has a profound effect oncrop yields, resulting in losses of about 1%/rabbit/ha which is equivalent to a loss of£6 per rabbit at current yield and grain prices (£70/tonnne).

If this information should become available growers will be able to make informedjudgements on: (a) what the likely costs of rabbit damage will be for any one site andconsequently whether a control program is required; (b) if control is required, what

4

the most cost-effective management option will be; (c) how long the crop isvulnerable to damage and therefore how long it needs to be protected. For example, itis currently recommended that once rabbits have been removed, new plantations areprotected by an effective, but expensive, rabbit-proof fence (MAFF-specification;Pepper 1992). Such a fence provides effective protection for 10 years, but accountsfor up to 40% of the plantation’s establishment costs. If the rabbits only have aneconomically significant impact on the crop during the early stages of itsestablishment (e.g. the first 18 months), then more temporary and therefore cheaperforms of fencing could be used. We would therefore recommend that damageassessment models, similar to those being developed for arable crops, are also appliedto SRC. These models will firstly predict what impact a local population of rabbitswill have on yield and, secondly, select the most appropriate and cost-effective optionfor preventing rabbit damage.

In the longer-term we recommend that future plant breeding programmes shouldselect for rabbit resistance alongside traits for rust and beetle resistance. There isabundant evidence of variation in the susceptibility of different willow species tomammal damage and many of these differences can be associated with the chemicalcharacteristics of the bark. For example, studies in Finland have shown that mountainhares prefer certain willow species over others and mature shoots over juvenile ones(Tahvanainen et al. 1985). There is also evidence to suggest that browsing resistanceis greatest in willow species growing in regions subject to severe browsing pressure(Bryant et al. 1989). This evidence indicates that selection for resistance amongstwillow species is possible and that trees from regions with high historical browsingpressure could form the focus of future searches for resistant varieties.

The use of rabbit resistant varieties in SRC plantations would reduce the need forcostly damage prevention measures by the grower. It would also answer theincreasing concerns from the public over the use of lethal control techniques inwildlife management, an important consideration for a developing crop that is relianton its ‘green’ credentials for public support.

REFERENCES

Beale, C. V. & Heywood, M. J. (1997). Productivity of commercial crops of shortrotation coppice at six sites in southern England. Aspects of Applied Biology49, 181-188.

Cowan, D. P. & Garson, P. J. (1985). Variations in the social structure of rabbitpopulations: causes and demographic consequences. In: Behavioural ecology:the ecological consequences of adaptive behaviour (Eds. R. M. Sibly & R. H.Smith), Blackwell Scientific Publications, Oxford. pp 537-555.

Cowan, D. P., Vaughan, J. A. & Christer, W. G. (1987). Bait consumption by theEuropean wild rabbit in southern England. Journal of Wildlife Management51, 386-392.

McKillop I. G., Ginella S. V., Wilson C. J., Hanlon, A. J. & Pugh B. D. (1993). Theeffects of power failure on European wild rabbits at electric fences. Journal ofApplied Animal Behaviour Sciences 35, 277-290.

McKillop, I.G. & Dendy, J.A. (2000). Advice on rabbit management for growers ofshort rotation willow coppice. CSL Advisory Leaflet. 13pp.

5

Pepper, H. W. (1992). Bulletin 102. Forest fencing, Forestry Commission,Edinburgh.

Pepper, H. W. (1998). Forestry Commission Practice Note 13. Prevention of mammaldamage to trees in woodland. Forestry Commission, Edinburgh. 12pp.

Rees, W. A., Ross, J., Cowan, D. P., Tittensor, A. M. & Trout, R. C. (1985) Humanecontrol of rabbits In: Humane Control of Land Mammals and Birds, (Ed. D. P.Britt), UFAW, Potters Bar, London, pp. 96-102.

Ross, J. & Sanders, M. F. (1987). Changes in the virulence of myxoma virus strains inBritain. Epidemiology & Infections 98, 113-117.

Smith, G. C. & Trout, R. C. (1994). Using Leslie matrices to determine wild rabbitpop growth and the potential for control. Journal of Applied Ecology 31, 223-230.

Tahvanainen, J., Helle, E., Julkunen-Tiitto, R. & Lavola, A. (1985) Phenoliccompounds of willow bark as deterrents against feeding by mountain hare.Oecologia 65: 319-323

Tittensor, A. M. (1981). Rabbit population trends in southern England. In:Proceedings of the World Lagomorph Conference (Eds. K. Myers and C. D.McInnes). University of Guelpf, Ontario pp 629-632.

Trout, R. C. & Smith, G. C. (1995). The reproductive productivity of the wild rabbitin southern England on sites with different soils. Journal of Zoology 237, 411-422.

Twigg, L.E., Lowe, T.J., Martin, G.R., Wheeler, A.G., Gray, G.S., Griffin, S.L.,O’Reilly, C.M. Robinson, D.J. & Hubach, P.H. (2000). Effects of surgicallyimposed sterility on free-ranging rabbit populations. Journal of AppliedEcology, 37, 16-39.

6

TAB

LE 1. Cost-effectiveness of m

ain rabbit managem

ent techniques.M

ETHO

DEFFEC

TIVEN

ESS(%

)M

ATER

IAL

CO

STS (£, excl.

VA

T)

LAB

OU

RC

OST

Sa (£, excl.

VA

T)

CO

ST -EFFECT

IVEN

ESS O

F METH

OD

Wire nettingC

SL-specification

95+£1.60/m

£1.75/m(costs of erectiononly; m

aintenancecosts are

additional)

An extrem

ely cost-effective method of m

anagement; designed to give protection at the

one site for up to 10 years; fences need to be maintained on m

onthly basis to repairdam

age by farm m

achinery or falling branches and to block any burrows dug under

them; in the short term

, less cost-effective than electric fencing but little difference inthe long term

(7+ years)Electricfencing

95+£1.40/m

£0.10/m(costs of erectiononly; m

aintenancecosts are

additional)

An extrem

ely cost-effective method of m

anagement; designed to be m

oved from site to

site; life expectancy of up to 10 years; fences need to be maintained every few

daysduring first m

onth after erection and thereafter at 2-3 week intervals; in the short term

,m

ore cost-effective than wire netting but little difference in the long term

(7+ years)

Fumigation

90£0.60/burrow

£0.20/burrowThe m

ost cost-effective lethal method of m

anagement, despite requiring tw

o people toapply it

Tree guards90+

£0.50/tree£0.50/tree

More cost-effective than fencing at norm

al planting spacings when dealing w

ith areasof up to about 2/ha, assum

ing net guards usedR

epellents95 (A

aprotect only)£0.25/tree/year

£0.50/tree/yearN

eeds to be reapplied annually; many repellents can give unsatisfactory results; tree

guards more cost-effective for sm

all plantations (<2ha), fencing for larger plantationsC

age trapping65

£15/trap£0.50/trap to install

and £0.50/visitThe best m

ethod for removing rabbits w

ithin fenced woodlands; requires visits tw

ice aday w

hen setD

rop-boxtrapping

80 (provisionalfigure)

£55/trap£5.00/trap to install

and £0.50/visitC

an be used to trap rabbits moving in and out of fenced plantations, although blocking

burrows under fences and using cage traps is probably m

ore cost-effective; effortneeded to site or m

ove traps and is therefore relatively inflexibleFerreting

35N

/AN

/AM

ainly used because perceived as a sport; time given freely by those undertaking it;

not a cost-effective method

Shooting30

£0.10/cartridgeN

/AM

ainly used because perceived as a sport; time given freely by those undertaking it;

not a cost-effective method

SnaresN

ot recomm

ended-

--

Spring trapsN

ot recomm

ended-

--

aLabour is costed at £5/hour, the average hourly rate of UK

agricultural workers (source, N

ational Farmers’ U

nion).

7

Integrated Crop Protection in SRC Willow Production

Ming Pei, Tom Hunter and Lori PeacockIACR-Long Ashton Rresearch Station, Department of Agricultural Sciences, University of Bristol,

Long Ashton, Bristol BS41 9AFE-mail: ming.pei@bbsrc.ac.uk

Summary

Melampsora rusts and chrysomelid beetles are the most serious disease and pest,respectively, of SRC plantations. Two rust species, M. epitea and M. capraearum, arecommon in SRC plantations in the UK. They have a complex life-cycle, producing fivespore stages and alternating on European larch. In addition to M. epitea and M.capraearum , an asexual form of rust infects young shoots and stems of S. viminalisclones. In M. epitea, there is a large variation in pathogenicity to different willows andthree mating populations have been identified. Beetle infestation is often severe locally.Four beetle species occur in willow SRC plantations, the major ones being the blue(Phratora vulgatissima.) and brassy (Phratora vitellinae) willow beetles. Beetledamage is caused firstly by overwintered adults early in the season, followed by larvaeand then by newly developed adults late in the season.

Resistance of different willow species/genotypes to rusts and beetles vary greatly.

Many willows have consistently shown a high degree of resistance against rust

infection and beetle feeding. In some willows, resistance to rust was broken down

due to changes in pathogen population. Studies suggest that, compared with

monoclonal plantings, mixtures can, in general, reduce rust levels. Against rust,

the beneficial effects become particularly apparent with willow clones that show

moderate to severe rust infections in monoclonal plantings. Against beetles, blue

willow beetle density, damage and, to an extent, number of eggs, tend to be lower

in mixtures. Such effects seem to be more obvious as the number of willow clones

increases and the distribution of different clones becomes more random. As

controlled experiments have shown that the blue willow beetles prefer to feed on

S. viminalis, mixtures composed of only S. viminalis clones may not be as effective

as mixtures containing different species/species hybrids against beetle damage.

Under laboratory conditions, the rust hyperparasite, Sphaerellopsis filum, is

capable of reducing willow rust spore production by up to 98%. Recent work has

revealed that S. filum varies greatly in pathogenicity to willow rust and its sexual

stage occurs in the UK. Studies are now being undertaken to determine how S.

filum spread in plantations and how it can be utilised to give best control of the

rust disease.

8

Routine use of pesticide in SRC plantations is not practicable for economic, technicaland environmental reasons. The desirable control strategies are integrated utilisationof host resistance and natural processes that limit the damage caused by diseases andpests. These include the selection and breeding for resistance, planting genotypemixtures and biological control of rust using S. filum. Willows are hugely diverseand, accordingly, have an immense pool of genes that could be used for disease andpest resistance. In planting mixtures, choosing a range of genetically diverse types isconsidered to enhance the property of mixtures.

Introduction

As the European Union is legally bound to an 8% reduction of greenhouse gas

emission by the year 2010, a large increase of energy from renewable sources is

expected. Biomass is one of the major sources of renewable energy and it is likely

that, in the near future, increasingly large-scale energy crop plantations will be

established. (MAFF, 2000).

Willows (Salix spp.) are the main crop in short rotation coppice (SRC)

plantations for renewable energy in the UK and Western Europe. They are

grown as the major SRC crop due to their coppicing ability and yield potential.

Most SRC plantings in the UK are established with S. viminalis, S. viminalis

hybrids with S. caprea and S. cinerea, and S. burjatica..

Like other crops, willows suffer from attacks by diseases and pests. Melampsora rustsand chrysomelid beetles are the most serious disease and pest in SRC plantations.Rust defoliates susceptible plantings prematurely and, when severe may reduce yieldsby as much as 40% (Parker et al., 1995). Severe rust predisposes plants to infectionsby secondary pathogens, which often lead to death of the plants. Beetle infestation canbe severe, especially at the local scale. The damage by beetles is caused by loss ofgreen leaf area due to foliar feeding, firstly by adults early in the season, followed bylarvae and, later in the season, by newly emerged adults. As more and more large-scale plantings become established, disease and pest problem will intensify if nocontrol measures are taken.

Melampsora rustsTwo rust species, M. epitea and M. capraearum, are common in willow plantations,with M. epitea being most predominant. Melampsora epitea infects only leaves whileM. capraearum infects both leaves and young stems. The two rust species have acomplex life-cycle (Fig. 1), overwinter on fallen willow leaves and have an alternatehost, European larch (Larix decidua) (Pei et al., 1993).

9

Fig. 1. Life-cycle of Melampsora epitea and M. capraearum

During late spring – autumn, the rusts are seen as yellow/orange pustules containingurediniospores. The uredinospores are capable of producing the next generation of thesame type of spores in 6-7 days. These urediniospores are responsible for many cyclesof disease in a growing season resulting in damage on willow. In the autumn, the rustsproduce teliospores and overwinter on fallen willow leaves. In spring, the teliosporesgerminate to produce basidiospores that infect larch. Genetic materials are reshuffledwhen teliospores germinate and genetic exchange takes place when spermatia fertilisereceptive hyphae. Fertilization of receptive hyphae by spermatia triggers formation ofaeciospores which infect only willow, not larch.

In addition to M. epitea and M. capraearum, a stem-infecting form (SIF) occurs on S.viminalis genotypes. It has only a single spore stage, overwinters on infected willowstems or buds and causes new infections early in the season (Pei et al., 1995).

Rust variationWithin M. epitea, there is large variation in pathogenicity to different biomass willowclones. Many pathotypes (defined according to their ability to infect a certain range ofwillow genotypes) occur in SRC plantations (Pei et al., 1996). So far 15 pathotypeshave been identified from UK sites. Data from pathogenicity tests with rust from UKsites in 1992-1997 suggest that pathotype composition has become increasinglydiverse. More pathotypes have been found in South West England and N Irelandwhere willows have long been grown.

Three distinct mating populations have been found in M. epitea. Each matingpopulation is specific in its pathogenicity to a particular range (section or species) ofwillows. Until very recently, the likelihood of gene exchange in rust had beendetermined by difficult and time-consuming crossing experiments. Recent studiessuggest that use of molecular tools may have promising prospects in determininggenetic relationships between pathotypes.

Rust resistance in willowIn the past 10 years, leaf rust (M. epitea) has, in general, sustained similar levels ofinfection on majority of the genotypes (some 20 – 30) selected in 1980s. However, rust

teliospore

aeciospore

urediniospore

basidiospores

spermatia

LarchWillow

10

resistance in several high-yielding genotypes, which had been regarded as mostpromising biomass willows, has been broken down by M. epitea. The main cause of thebreakdown is the rise of new virulent pathotypes. An important biomass willow, S. %mollissima ‘Q83’, was highly resistant to rust until 1992 when a new pathotype LET4arose which rendered this genotype susceptible. DNA fingerprinting of rust samplescollected in 1992 revealed that the new pathotype spread from different sources. Salixburjatica ‘Germany’, which was resistant to rust until early 1990’s, has now becomeone of the most susceptible genotypes. It appears that the present rust population on‘Germany’ is much more aggressive compared with the rust previously found on thesame genotype. The new, highly promising biomass genotype S. viminalis x burjatica‘Stott 10’ can yield twice as much as most biomass genotypes selected in 1980s. It hasbeen highly resistant to rust in most UK sites since it was bred in late 1980s. However,increased levels of rust on this genotype have been found at a site in Northern Irelandand a site in Southern England since 1997. The willow genotypes introduced ordeveloped in early 1990s as a result of breeding (about a dozen) were mostly lesssusceptible to leaf rust (slight increase of leaf rust infection has been observed on someof these genotypes in recent years).

The stem-infecting form (SIF) occurs specifically on S. viminalis clones. Most of thecommercial biomass willows currently available in UK and European markets are S.viminalis clones and its hybrids with S. schwerinii, produced by Svalof Breedingprogramme in Sweden and introduced into the UK in early 1990s. While these cloneshave generally been resistant to leaf rust, several are infected by SIF in the UK.

Results from inoculation tests and field disease assessments indicate that majority ofwillows introduced from the Far East and North America are resistant to UK rustpopulations.

Chrysomelid beetlesThe main pests of willow plantations are members of Chrysomelidae. Phratoravulgatissima (blue willow beetle) and P. vitellinae (brassy willow beetle) are the mostprevalent while, occasionally, damage is caused by Galerucella lineola andPlagiodera versicolora. The life-cycle of P. vulgatissima is shown in Figure 2. InApril, adults emerge from overwintering sites (usually outside the plantations) to findwillow plantings on which to feed. Female beetles lay eggs on willow leaves and theresulting larvae are responsible for further damage in the plantation during the summermonths. In late summer and early autumn, new adult beetles emerge from the pupaedeveloped from the larvae. After approximately 6-8 weeks, these beetles will leave andhibernate in suitable overwintering sites such as bark crevices and old fence posts. Ithas been suggested that in some years the other willow beetles may produce a secondgeneration of adults.

Summer

Spring Autumn

Adults

Adultsnew generation of adults emerging

Adults leave willows and disperse(fly) to overwintering sites

Adults leave overwinteringsites and fly to willowswhere they feed, mate andbegin egg laying

Eggson willow

Larvaeon willow

Pupaein soil

Larval feeding damage

Adult feeding damage

Adult feeding damage

11

Fig. 2. Life-cycle of blue willow beetle Phratora vulgatissima

Feeding damage in willowTo assess the effect of beetle infestation on biomass yields, beetle feeding damage wassimulated by removing leaves from growing plants of S. viminalis ‘Bowles Hybrid’(Fig. 3., from Kendall & Wiltshire, 1998). In the most extreme case, where leaves wereremoved in both June and August, a defoliation of 90% resulted in approximately 40%yield loss. Previous field surveys have shown that, in general, total leaf area loss bybeetle feeding has been in the region of 10 – 30% (Hunter and Peacock, personalobservations). A loss of 10 - 30% of the total leaf area resulted in approximately 8 -22% yield loss in the simulated defoliation experiments. However, if damage issubstantial over a number of years, the consequence will be more severe.

Fig. 3. Effects of simulated defoliation on biomass yields (means of ANOVA) of S.viminalis ‘Bowles hybrid’. From Kendal & Wiltshire. 1998.

Beetle feeding preferencesSurveys of willow plantations indicate that the severity of chrysomelid beetleinfestation varies between years and between locations (Hunter et al., 1996).Laboratory and field studies have shown that the willow beetles exhibit feedingpreference differences between willow genotypes (Wiltshire et al., 1997). Various

55

60

6570

75

8085

90

95

100

0 10 20 30 40 50 60 70 80 90 100% leaves removed

dry

wei

ght (

% o

f con

trol

)

early (June)late (August)early + late

12

chemicals, including phenolic glucosides, have been found in willow leaves. In general,S. viminalis and its hybrids are favoured by both P. vulgatissima and G. lineola. Itappears that P. vulgatissima prefers willows with low levels of phenolic glucosideswhile P. vitellinae prefers willows with high levels of these chemicals.

Disease and pest control

SRC willow is a new crop and there are no long-established protocols for disease

and pest control. Unlike most food and horticultural crops, routine use of

fungicides and pesticides in SRC plantations is not considered to be a viable

option because (a) it is too costly for the low-input crop, (b) it is difficult to

obtain effective coverage of chemical over the large and dense canopy and (c)

application of large quantities of agrochemicals may have harmful effects on the

environment. A desirable strategy of disease and pest control in SRC plantations

is the integrated utilisation of host resistance and natural processes that limit the

damage by diseases and pests.

HOST RESISTANCEWillows are hugely diverse (300-500 species, each species having numerousgenotypes), hybridise with relative ease and only need 1-3 years to reproduce. Studieshave shown that a wide range of willows, particularly those introduced from the FarEast and North America, are resistant to rust and beetles. These attributes make the useof natural resistance one of the most attractive options to control diseases and pests.Currently, two willow breeding programmes, one based at Long Ashton, UK, and oneat Uppsala, Sweden, are underway. These programmes aim at producing disease andpest resistant, high yielding genotypes and have already produced hundreds of crossesfor selection and evaluation.

As rusts are specialised fungi, changes or shifts of pathogenicity in the population mayresult in severe damage to previously resistant crops. In the past 10 years, rustresistance in several high-yielding genotypes, which had been regarded as amongst themost promising biomass willows in the UK, has been broken down by M. epitea.Therefore, breeding for durable resistance is particularly important in deploying geneticresistance. An understanding of the behaviour of pathogen populations and breedingbased on broad genetic backgrounds will help to produce novel genotypes having long-lasting resistance against rust.

Willow mixturesLarge-scale monoculture plantings are known to be vulnerable to attacks by diseasesand pests. A strategy to reduce such risks is to plant genotype mixtures to increasegenetic diversity. Because genotype mixtures are readily accepted for biomassproduction, willows have a unique advantage over food or horticultural crops, inwhich flavour, colour or shape may have prime importance. The experience in the UKover the last 10 years suggests that, compared with monoclonal plantings, mixtures

13

can, in general, reduce disease and pest levels. In particular, the beneficial effects ofmixtures against rust become apparent with the genotypes which show moderate tosevere rust infections in monoclonal plantings (Parker et al., 1995; McCracken &Dawson, 1998). Studies have shown that the planting of willow mixtures may alsoreduce the extent of damage by beetles. Compared with monocultures, beetle density,damage and, to an extent, number of eggs, tend to be lower in mixtures containingwillows that differ in their susceptibility to beetles (Peacock & Herrick, in press).Such effects seem to be more obvious as the number of willow clones increase and thedistribution of clones becomes more random. Currently, many SRC plantings aremixtures containing only S. viminalis clones. As experiments have shown that beetlesprefer to feed on S. viminalis, S. viminalis single species mixtures may not be aseffective as mixtures containing different species/species hybrids against beetledamage.

One of the major concerns over planting genotype mixtures is that, because virulencepatterns in pathogens can change quickly through gene exchange during the sexuallife-cycle, the mixtures may favour the emergence of ‘super races’ that are capable ofattacking all components of mixtures. If that is so, the effect of the mixtures would begreatly reduced (Wolfe, 1985). Our studies suggest that M. epitea consists of severaldifferent mating populations (Pei et al., 1999a, b). For example, the pathotypesoccurring on S. viminalis genotypes do not exchange genes freely with those on S.burjatica. By choosing genotypes that harbour genetically separate pathogenpopulations, the property of mixtures to suppress rust would be enhanced. At present,because of availability, most new SRC plantings are mixtures containing only S.viminalis genotypes plus one or two S. viminalis hybrids. Whether such mixturesencourage the pathoptypes having wide spectra of virulence is an important questionto be answered in the near future.

In planting willow mixtures, certain characteristics, such as growth habits, soil andwater conditions, and adaptability of the willows to the environment, need furtherconsideration.

Biological control of rustCompared with annual crops, SRC is better suited for biocontrol because of the carry-over effect on biological control agents, assuming a 3-5 year harvest interval.Sphaerellopsis filum is a fungal hyperparasite. It attacks a wide range of rust fungi,including willow Melampsora. Under experimental conditions, the hyperparasite canreduce willow rust spore production by up to 98%. Although S. filum occurs only onrusts in nature, it can be cultured on artificial media. This provides the opportunity ofgrowing the fungus in vitro and then reintroducing it into the field.

Previously, only the asexual stage of S. filum was known in the British Isles. Asexualconidiospores are waterborne and dispersed mainly by rain splash. Recently, it wasfound that sexual ascospores occur in S. filum in the UK (Yuan et al., 1998).Ascospores are airborne and, hence, are more efficiently dispersed. Until veryrecently, it was not clear whether S. filum is specialised in its pathogenicity. Recentstudies revealed that S. filum varies in pathogenicity and some strains are significantlymore virulent than others to willow Melampsora (Yuan et al., 1999). Answers to thefollowing will help to deploy S. filum more effectively for biocontrol of rust in SRC

14

willow: (a) to what extend S. filum varies in pathogenicity, (b) whether S. filumproduces the sexual stage on willow rust and (c) how it spreads.

Other measuresAs European larch serves as an alternate host of willow rust, planting SRC willowsnear larch will cause early disease onset and more severe infection. Therefore, willowplantings should be sited as far away from larch plantations as practicable to delay theonset and reduce rust severity.

When fungicides were used for studies of effects of rust disease on growth and yields,it was found that myclobutanil (Systhane) and benodanil (Calirus) are effectiveagainst rust (Royle, 1991, McCracken & Dawson, 1998). Field experiments showedthat, on susceptible genotypes, at least 5 applications were needed to maintain lowlevels of rust throughout the season (Royle, 1991).

Conclusion

The potential threat of diseases and pests on SRC willow is likely to increase as moreand more willow plantings are established. Integrated control of SRC willow diseasesand pests involves the breeding for resistant genotypes, planting host genotypemixtures and deployment of biological control agents. The encouraging prospect isthat these integrated approaches can benefit from the huge diversity in willows,acceptance of mixtures for energy production and the perennial nature of the crop.

Acknowledgement

We thank the Ministry of Agriculture, Food and Fisheries (MAFF) and the

DTI’s New & Renewable Energy Programme (managed by ETSU), UK, and the

European Commission for financing the work. IACR receives grant-aided

support from the Biotechnology and Biological Sciences Research Council of the

United Kingdom.

References

Hunter, T.; Royle, D.J. ; Arnold G.M. (1996). Variation, in the occurrence of rust(Melampsora spp.) and other diseases and pests, in short rotation coppiceplantations of Salix in the British Isles. Annals of Applied Biology 129; 001-012.

Kendall, D. A. ; Wiltshire. C. W. (1998) Life-cycle and ecology of willow beetles onSalix viminalis in England. European Journal of Forest Pathology 28, 281-288.

MAFF (2000) England Rural Development Plan: Energy Crop Scheme ConsultationDocument. Available from URL:www.maff.gov.uk/farm/acu/acu.htm

15

McCracken, A. R.; Dawson, W. M. (1998) Short rotation coppice willow in NorthernIreland since 1973: development and use of mixtures in the control of foliar rust(Melampsora spp.). European Journal of Forest Pathology 28, 241-250.

Parker, S. R.; Pei, M. H.; Royle, D.J., Hunter, T.; Whelan, M.J. (1995)Epidemiology, population dynamics and management of rust diseases inwillow energy plantations. Final Report of Project ETSU B/W6/00214/REP.Energy Technology Support Group, Department of Trade and Industry, UK.

Peacock, L.; Herrick, S. Responses of the willow beetle Phratora vulgatissima togenetically and spatially diverse Salix spp. plantations, Journal of AppliedBiology (in press).

Pei, M.H.; Royle, D. J.; Hunter, T. (1993) Identity and host alternation of somewillow rusts (Melampsora spp.) in England. Mycological Research 97,845-851.

Pei, M.H.; Royle, D. J.; Hunter, T. (1995) A comparative study of stem-infecting andleaf-infecting forms of Melampsora rust on Salix viminalis in the UK.Mycological Research 99, 357-363.

Pei, M. H.; Royle, D. J.; Hunter, T. (1996) Pathogenic specialization in Melampsoraepitea var. epitea on Salix. Plant Pathology 45, 679-690.

Pei, M. H.; Hunter, T.; Ruiz, C. (1999a) Occurrence of Melampsora rusts in biomasswillow plantations for renewable energy in the United Kingdom. Biomass andBioenergy 17, 153-163.

Pei, M. H.; Hunter, T.; Royle, D. J. (1999b) Hybridisation in larch-alternatingMelampsora epitea (M. larici-epitea). Mycological Research 103, 1440-1446.

Royle, D. J. (1991) Potential Pest and disease problems associated with arable energycrops. In Wood Fuel for Thought, Richards, G. E. (Ed), Harwell Laboratories,Oxford, pp 283-294.

Wiltshire C. W.; Kendall D. A.; Hunter T.; Arnold G.M. (1997) Host-plantpreferences of two willow-feeding leaf beetles (Coleoptera, Chrysomelidae)Aspects of Applied Biology 49, 113-120.

Wolfe, M. S. (1985) The current status and prospects of multiline cultivars and varietymixtures for disease resistance. Annual Review of Phytopathology 23, 251-273.

Yuan, Z. W.; Pei, M. H.; Hunter, T.; Royle, D. J. (1998) Eudarluca caricis, theteleomorph of the mycoparasite Sphaerellopsis filum, on blackberry rustPhragmidium violaceum. Mycological Research 102, 866-868.

Yuan, Z. W.; Pei, M. H.; Hunter, T.; Ruiz, C.; Royle, D.J. (1999) Pathogenicity towillow rust, Melampsora epitea, of the mycoparasite Sphaerellopsis filumfrom different sources. Mycological Research 103, 509-512.

16

Willow Aphids on SRC WillowEffects of aphids on host plants: A project summary

C. Matilda Collins,Department of Biology, Imperial College of Science Technology and Medicine,

Silwood Park, SL5 7PY.Tel: 020 7594 2420 Fax: 020 7594 2308 e-mail: t.collins@ic.ac.uk

ABSTRACTTuberolachnus salignus and Pterocomma salicis are both aggregative aphids that have been notedwithin SRC crops. Experiments with willows grown in soil and in hydroponic culture reveal that T.salignus can reduce the above-ground yield of biomass willows, have severe negative effects on theroots and reduce the survival of both newly planted and established trees. Pterocomma salicis can alsoreduce yield and negatively affect the roots of host trees, but its effects are less marked. Both speciescan lead to plant stress and increase the vulnerability of host trees to other factors. Field observationsindicate that T. salignus may be increasing in abundance in SRC crops.

INTRODUCTIONInsect pests can have adverse impacts on SRC yield and thus on the economics of production. Previousentomological studies have concentrated on the more obvious damage to SRC caused by defoliatinginsects. The damage caused by both leaf and stem-feeding aphids is largely invisible; other than intheir roles as viral vectors and gall inducers they rarely cause obvious alterations to leaf structure andfunction. However, aphid presence is an energy and nutrient drain due to their phloem-feeding habit.The energy and nutrient loss caused to sycamore trees (Acer pseudoplatanus) by the sycamore aphid(Drepanosiphum platanoidis) has been shown to lead to earlier leaf fall, to reduce the size of leaves by40% and the production of stem wood by 62% (Dixon, 1971). The sycamore aphid is much smallerthan the giant willow aphid, Tuberolachnus salignus, which has been shown to ingest thephotosynthetic product of 5-20 cm2 of leaf per day (Mittler, 1958). The magnitude of this drain has thepotential to affect significantly the yield of willows by altering the source-sink relationship within theplant. There may also be potential impacts on the long-term development and fitness of the rootstockleft in situ after coppice harvesting as a result of changes in root health, density and architecture.

THE APHIDS

This study has focussed on two aphid species: Tuberolachnus salignus, the giant willow aphid andPterocomma salicis, the black willow aphid. Both species aggregate and occur in dense colonies on thestems of infested plants, but have contrasting ecologies.

Tuberolachnus salignus (Gmelin), Lachninae, LachniniThe Giant Willow-aphid is one of the largest aphids ever recorded (BL 5.0-5.8 mm) (Blackman &Eastop, 1994). It feeds almost exclusively on willow, but has very occasionally been recorded onpoplar (Populus sp.). Its distribution reflects that of willows; it is virtually cosmopolitan, only beingabsent from Australasia. The species is apparently parthenogenetic; no sexual morphs have ever beenfound and recent DNA studies support this (Blackman & Spence, 1996).The species is strongly aggregative, forming vast colonies on infested trees. These colonies can covermuch of the 1-3 year old stem surface of a tree. Other than its large size the most distinctive features ofthis aphid are a large dorsal tubercule (function unknown), the reddish stain it makes when squashedand synchronised waving of the hind legs over large areas of infestation. Many entomological textsnote this insect as being capable of much damage, even of killing willows up to 40 feet tall (Buckton,1881; Das, 1918; Swirski, 1963), but until now little attention has been paid to its ecology and hostrelationships.

Pterocomma salicis (Linaeus), Aphidinae, PterocommatiniThe Black willow-aphid is another large aggregative aphid (BL 2.7-4.5 mm). It also feeds exclusivelyon willow (Blackman & Eastop, 1994), and is common in the northern hemisphere. Although thespecies is parthenogenetic for much of the year, a sexual generation is produced in autumn and the

17

overwintering egg is found in the buds and on the stems of willows. The species is often attended byants, which may protect it from predation.

In this report I summarise studies on the impact of these two species on the establishment and yield ofSRC willows.

MATERIALS AND METHODSFor full details please refer to my PhD Thesis (Collins, 2001).

Field monitoring.Aphid populations on four willow varieties have now been monitored for three years at Roves Farm,Sevenhampton, Wiltshire, U.K. (Grid Reference: SU 210 888). 20 coppice stools of ‘Bowles Hybrid’,and 21 stools each of ‘Jorr’, ‘Orm’ and ‘Dasyclados’ were assessed for aphids weekly in 1998 andfortnightly in 1999 and 2000.

Effects of aphids on host plantsWillows grown in soil:Fifty-two potted cuttings of the hybrid willow clone ‘Jorr’ were randomly assigned to one of fourtreatments:1 No aphids2 Inoculation with 20 Pterocomma salicis on 12th April 19993 Inoculation with 5 Tuberolachnus salignus on 25th June 19994 Inoculation with both the above.The aphid infestations were monitored weekly and assigned to levels in a six point scale:0.<5 aphids, 1. 5 – 20 aphids, 2. 20 – 50 aphids, 3. 50 – 100 aphids, 4. 100 – 300 aphids, 5.>300 aphidsThe levels recorded on a weekly basis were summed at the end of the experiment to estimate the totallevel of infestation to which each tree was subject.Harvesting. The first harvest of above ground biomass took place nineteen weeks after infestationwith P. salicis and nine weeks after infestation with T. salignus. The trees were subsequentlymaintained ‘aphid free’ and allowed to recover. The second harvest took place ten weeks later.Post harvest recovery. After the first harvest, trees were assessed weekly against a phenology scale inorder to provide an assessment of the speed of their recovery from coppicing.Willows grown in water cultures to give greater detail:Tuberolachnus salignus. These experiments were carried out in a controlled environment chamber setat 15(±1)°C and 16 hours of light to 8 of darkness. In the first experiment the willows were infested atplanting, in the second the cuttings were allowed 60 days to establish prior to infestation. Aphidinoculation was with a single young adult alatae. Harvesting took place 25 days after infestation in bothcases.Pterocomma salicis. This species is less amenable than T. salignus and insufficient numbersestablished under controlled environment conditions. A subsequent repetition of the experiment in aglasshouse (min temp 15±1°C, ambient June daylight regime) has yielded data for infestation atplanting, but not for infestation after 60 days establishment.Measurements: For each plant, the number of leaves was counted weekly, this served as a non-invasive growth measure. In the first experiment the number of primary roots emerging were alsocounted; the longer duration of the second experiment made this impossible and roots were measureddestructively at harvest. Wet and dry mass of the roots, shoots and twig were determined after harvest.Photosynthetic rate: 12 pot grown ‘Dasyclados’ individuals were randomly allocated to one of twotreatments: control and with aphids. Trees were infested by inoculation with 5 4th instar T. salignusapterae and populations were left to develop for three weeks prior to sampling. Photosynthetic rate wasdetermined using an IRGA. 12 measurements were made per tree, four at each of three sites: thebranch tip, the mid-stem and the base.Nitrogen concentration: For analysis of leaf nitrogen content, five leaves were randomly selectedfrom each tree used in the photosynthesis experiment. The leaves were oven dried for 48 hours at80°C, ground in a ball mill to homogenise and a sub-sample of this digested in sulphuric acid andselenium at 400°C for 1.5 hours (Allen et al., 1989). Total nitrogen content was determined using aKjeltec nitrogen analyser (Persop Analytical, Bristol) and the % dry mass calculated.

18

RESULTSAphids in the fieldOver the last three years the giant willow aphid has increased in abundance in Jump Field. In October1998 there was a peak number of 14,000 aphids on the sample of 83 trees, in October 1999 this figurewas over 190,000 (Figure 1). Not only have the numbers of aphids increased, but also the duration ofinfestation. The increase in duration of infestation, from 10.24 ±0.74 weeks in 1998 to 17.17 ±0.54weeks in 1999 is highly significant (T=7.52, d.f.=164, p<0.001). When the survey work for 2000/2001is completed we will be able to see if this trend of increased abundance is continued.This species has an unusual temporal pattern of occurrence for a tree-feeding aphid. It increases insummer, normally considered a time of poor host quality and remains at high levels until the earlyspring, when it vanishes from the sample for several months – a portion of the lifecycle that remainsunexplained. Its continued presence on trees that have shed their leaves, and are thus notphotosynthetically active, indicates unusual feeding abilities. Additionally a high degree of coldtolerance is needed to endure winter frosts and there are indications that survival is high at temperaturesas low as –5°C (L. Peacock, Pers. Comm.). No predators have been observed regularly feeding on thisspecies and there has also been no evidence of parasitoid activity. In the winters of 1998 and 1999colonies of giant willow aphid were decimated by an entomopathogen; the aphids swell and die, butremain hanging by their stylets for several days before exploding. It has not yet been determinedwhether this pathogen is of viral or fungal origin, but after three years of observation thisentomopathogen remains the only prospect for biological control of the giant willow aphid.The giant willow aphid excretes a substantial amount of sugar rich honeydew, which attracts largenumbers of other insects into the crop, possibly influencing the species diversity of willow stands. Thisaccessible carbon source may also have a profound influence on soil chemistry and further studies arein progress to examine this area (Collins & Crawley, Submitted)The black willow aphid population in this crop has declined since the peak numbers of this projectwere recorded in May 1998, and relatively few were found in 1999. The duration of infestation alsodeclined significantly between 1998 and 1999 (T=5.47, d.f.=164, p<0.001) from 2.61 ±0.31 weeks in1998 to 0.63 ±0.18 weeks in 1999. There is anecdotal evidence from the landowner, Mr Rupert Burr,that P. salicis reached a peak of abundance in 1997, prior to this sampling programme. In both 1999and 2000 this species has largely been found in mixed-species colonies with T. salignus, but theirnumbers now appear to be rising again. This species is consumed by many of the common aphidpredators; hoverfly, lacewing and ladybird larvae have all been observed feeding on the black willowaphid. A high level of parasitioid activity was also observed coincident to the decline in numbers in thesummer of 1998, with hyperparasitoids and mummy parasitoids also being very abundant (F.van Veen,pers. comm.).Effects of aphid infestation on yield.In all aphid treatments the biomass at first harvest was reduced compared to the control treatment(F=42.87, d.f.=3,48, p<0.001)(Table 1). Tuberolachnus salignus, which reduced host biomass to lessthan half that of the controls in nine weeks of infestation, has a greater impact on yield than doesPterocomma salicis. The presence of both species on the same tree does not cause an additive declinein tree performance, indeed, the black willow aphid appears both to reduce the numbers of the giantwillow aphid and to mitigate its impact.Infestation with T. salignus, whether alone or with P. salicis, delayed host recovery after harvest(Figure 2). Infestation with P. salicis alone had no impact speed of recovery.The presence of T. salignus prior to first harvest continues to have a negative impact on the host tree atsecond harvest (F=6.15, d.f.=3,38, p<0.002)(Table 1). As aphids were not present during the inter-harvest period, this implies a long-term impact on the growth of the host plant.The proportion of dry matter in the shoots at second harvest is higher in the T. salignus treatment thanin the others. In spite of the absence of direct aphid influence, this growth is not only quantitativelydifferent from the control but also qualitatively. A long-term effect such as this is likely to be mediatedby the roots. Intriguingly, the trees that had been host to the black willow aphid were larger than thecontrols at second harvest. This could however be an artefact of the restricted rooting area available inthe pots; the large control trees may have become pot-bound prior to first harvest, thus restricting theirbiomass increase between the two harvests. Trees infested by P. salicis grew less in the first stage andmay have had more opportunity for root growth, and thus shoot growth, in the second stage.Aquaculture – the detailsTuberolachnus salignus growth rates: Although only a single alate was used for inoculation, underthese experimental conditions an aphid clone can rapidly increase in biomass. Alates have a mean (±1SE) mass of 5.28 (± 0.17)mg, give birth to 34.3 (± 1.4) nymphs which in turn can reach reproduction

19

within 28.11 (± 0.25) days. Thus, within the 25 days of each experiment each alate can lead to an aphidbiomass per twig of 346.7 (± 17.8) mg (Collins & Leather, In Press).Infested at planting. The performance of the willows in each treatment diverges immediately. Thepattern of above ground growth in this experiment is similar to the results from soil grown trees: bothspecies had a severe effect on host plant growth. There is also a qualitative difference in the aboveground growth. Aphid infested shoots have a significantly lower water content than the control shoots,although the effect of P. salicis is minor. By the first root and leaf count on the 11th day of theexperiment aphid infested trees have a significantly lower number of primary roots and leaves (roots:t=3.32, df=26 p<0.01, leaves: t=3.47, df=26, p<0.001). In the infested treatment some of the plantswere observed to reabsorb previously initiated roots.The dry mass of roots is significantly lower in both aphid treatments than in the controls, with T.salignus having the greater impact. Additionally roots in the T. salignus infested treatment had a higherproportion of water than control roots (Table 2). The aphids also have an impact on the woody tissue.Significantly less of the original twig mass remains in the aphid infested treatment.Infested at 60 days. There was no visual difference in size or significant difference in number ofleaves between treatments at infestation (t=0.80, df=31, p=0.42). Ten days after infestation the meannumber of leaves per plant in each treatment differs (t=4.46, df=31, p<0.001). Giant willow aphidinfested trees shed their lower leaves and apical growth was vastly reduced. In spite of the reducednumber of leaves, the water consumption per plant rises markedly after infestation, then falls as theplant starts to weaken and wilt (Figure 3).As above the biomass of shoots and roots are significantly reduced in the aphid treatment compared tothe control. Again, there is also a qualitative difference in both the above and below ground growth.Aphid infested shoots were dehydrated compared to control shoots and roots of infested trees had amuch higher water content than control roots.Photosynthesis and leaf nitrogen content. Both of these measures were significantly raised by thepresence of Tuberolachnus salignus. The photosynthetic rate was raised in all three positions on thetree, i.e. above and below the level of aphid infestation (Table 3). Photosynthesis is a water-demandingplant metabolic process and with reduced root capacity, this may be a key mechanism in the plant stressinduced by aphid presence.

CONCLUSIONSBoth these aphid species do adversely impact the yield of SRC willows. The black willow aphid’seffect is less marked than that of the giant willow aphid, which can reduce not only the growth theplants make subsequent to infestation but also alter previously developed woody tissue. The negativeeffects of the giant willow aphid on the shoots and roots of established and establishing trees are drasticand are both quantitative and qualitative. Impoverished roots can reduce the survival of infested treesas well as increasing the vulnerability of the host to other potentially detrimental factors. It is likelythat mortality previously attributed to drought or frost may have been exacerbated or increased by thepresence of these stem-feeding aphids.As yet we understand little of the population dynamics of these aphids and long-term monitoring withinextensive SRC crops is required to determine the frequency and abundance of their occurrence. Agreater understanding of the ecology of these species is also required to enable the investigation ofnatural enemies and entomopathogens as potential control measures.

ACKNOWLEDGEMENTSI wish to thank my supervisors Simon Leather and Rufus Sage, the Game Conservancy Trust and theDTI’s New & Renewable Energy Programme (managed by ETSU) for their financial support andRupert and Joanna Burr for allowing me unrestricted access to their willow crops.

REFERENCESAllen, S.E., Grimshaw, H.M., Parkinson, J.A., & Quarmb, C. (1989) Chemical Analysis of EcologicalMaterials Blackwells, Oxford.Blackman, R.L. & Eastop, V.F. (1994) Aphids on the world's trees: an identification and informationguide CAB International in association with The Natural History Museum.

20

Blackman, R.L. & Spence, J.M. (1996) Ribosomal DNA is frequently concentrated on only one Xchromosome in permanently apomictic aphids, but this does not inhibit male determination.Chromosome Research, 4, 314-320.Buckton, G.B. (1881) Monograph of the British Aphides The Ray Society, London.Collins, C.M. (2001) Aspects of the ecology of two stem-feeding willow aphid species. PhD, ImperialCollege of Science Technology and Medecine.Collins, C.M. & Crawley, M.J. (Submitted) Indirect effects of aphid feeding: an experiment withartificial honeydew. Functional Ecology.Collins, C.M. & Leather, S.R. (In Press) The effect of temperature on the fecundity and development ofthe Giant Willow Aphid, Tuberolachnus salignus (Gmelin) (Lachninae, Lachnini). European Journal ofEntomology.Das, B. (1918) The Aphididae of Lahore. Memoirs of the Indian Museum, 6, 135-274.Dixon, A.F.G. (1971) The role of aphids in wood formation. I. The effect of the sycamore aphidDrapanosiphum platanoidis (Schr.) (Aphididae), on the growth of sycamore, Acer pseudoplatanus (L.).Journal of Applied Ecology, 8, 165-179.Mittler, T.E. (1958) Studies on the feeding and nutrition of Tuberolachnus salignus . II.-The Nitrogenand sugar composition of ingested phloem sap and excreted honeydew. Journal of ExperimentalBiology, 35, 74-84.Swirski, E. (1963) Notes on Plant Lice (Aphidoidea) of Israel. Israel Journal of Agricultural Research,13, 9-23.

21

Figure 1. The total number of Tuberolachnus salignus (solid line) and Pterocomma salicis (dottedline) present on 83 willow coppice stools at Jump Field, Roves Farm (G.R. SU 210 888) from February1998 to September 2000. Note the logarithmic response axis.

Figure 2: The individual and combined effect of previous infestation with the aphids Tuberolachnussalignus and Pterocomma salicis on the speed of recovery of coppiced shoots of the willow clone‘Jorr’. The response axis represents a phenological scale against which recovery was assessed, 1represents swelling of the buds, 6 full leaf development.

Figure 3: The effect of Tuberolachnus salignus on water use (ml/day/tree) by saplings of the SRCwillow clone ‘Jorr’ grown in water cultures

22

Table 1: The effect of the aphids Pterocomma salicis and Tuberolachnus salignus compared tocontrols on the above-ground biomass, losses and water content of saplings of the SRC willow clone‘Jorr’ grown in soil. All values given are significant at 5%. n.s. denotes non-significant differences.

Aphid Species: Pterocommasalicis

Tuberolachnussalignus

Both Species

Shoot dry mass at 1st harvest - 23% - 52% - 47%losses to 2nd harvest 8 % 33 % 47 %Shoot dry mass at 2nd harvest + 95% - 67% - 54%water content of shoot mass n. s. - 36% n. s.

Table 2: The effect of the aphids Pterocomma salicis and Tuberolachnus salignus compared tocontrols on the biomass and water content of shoots, roots and woody tissue of saplings of the SRCwillow clone ‘Jorr’ grown hydroponically. All values given are significant at 5%. n.s. denotes non-significant differences, - that data was not available.

Aphid Species: Pterocommasalicis

Tuberolachnussalignus

Infested at planting Shoot dry mass - 53% - 58%Shoot water content - 2% - 66%

Root dry mass - 53% - 84%Root water content n. s. + 68%

woody tissue dry mass n. s. - 5%Infested at 60 days Shoot dry mass - - 30%

Shoot water content - - 19%

Root dry mass - - 50%Root water content - + 44%

woody tissue dry mass - - 10%

Table 3: The effect of Tuberolachnus salignus on the net rate of photosynthesis (µmol/m2/s) andnitrogen content (% dry weight) of leaves of the SRC willow clone ‘Dasyclados’. All values given aresignificant at 1%.

Net photosynthetic rate % NBase Mid Tip

+ 119% + 80% + 76% +22%

23

1

1 0

1 0 0

1 0 0 0

1 0 0 0 0

1 0 0 0 0 0

1 0 0 0 0 0 0

F e b -98 M a y -98 A u g -98 N o v -98 F e b -99 M a y -99 A u g -99 N o v -99 F e b -00 M a y -00 A u g -00

T ub ero lac hnus s alignusP tero c o m m a s alic is

24

0

5

10

15

20

25

30

35

40

0 10 20 30 40 50 60 70 80 90

Time in Days

ml/p

lant

/day

Control

Tuberolachnus salignus

25

IMPROVING BREEDING EFFICIENCY OF BIOMASS WILLOWS USINGMOLECULAR MARKER TECHNOLOGY

S J Hanley, J H A Barker and A Karp

IACR- Long Ashton Research Station, Department of Agricultural Sciences,University of Bristol, Long Ashton, Bristol BS41 9AF, UKE-mail: angela.karp@bbsrc.ac.uk

Abstract

Increased energy production from renewable sources is currently a high priority inEurope. At present, biomass accounts for 3% of European inland energyconsumption. It has been proposed that by 2010 this figure should be increased to 8%or 90 million tonnes of oil equivalent (mtoe). To meet this target, efforts areunderway to breed and produce higher yielding biomass cultivars. While significantadvances have been made in this area, the efficient coupling of conventional breedingmethods and molecular marker technologies are required to maximise yield increasesand provide sustainable disease and pest resistance for the future. Molecular markertechnologies can accelerate the efficiency of breeding programs by helping to locategenes controlling agronomic traits and provide markers for marker-assisted selection.In order to achieve this, genetic maps with very large numbers of markers have to beproduced to increase the chance of finding a marker closely linked to a trait ofinterest. The technique of Amplified Fragment Length Polymorphism (AFLP)generates large numbers of markers in a single assay but these markers cannot easilybe transferred from one cross to another. Microsatellites provide fewer markers butcan be used to anchor maps from different crosses. Towards marker-assisted selectionin willows, a genetic map of a full-sib cross of Salix viminalis is being constructed. Inaddition, two reference families, each comprising 950 progeny, have been establishedand initial trait assessments, including laboratory-based beetle feeding assays and rustpathogenicity tests, have been performed. To identify markers linked to traits ofinterest, two approaches will be used:

• QTL analysis - associations between markers and traits will be identified usingsoftware packages that identify quantitative trait loci (QTL)

• Bulked Segregant Analysis (BSA) – DNA from individuals of contrastingphenotype (e.g. resistant and susceptible) will be bulked and markers whichcorrespond to only one phenotype will be identified

Furthermore, in an attempt to merge the mapping efforts of willow with poplar, the

efficacy of using poplar microsatellite markers for mapping studies in willow has

been demonstrated.

Introduction

26

A current limitation facing the breeding of improved willow varieties is our lack ofknowledge on the genetic basis of important agronomic traits. In this project we areapplying state-of-the-art molecular marker technologies to gain an insight into thesefundamental issues. In many crops, it has been demonstrated that conventionalbreeding efforts can be accelerated by using markers to locate genes controllingimportant agronomic traits and providing markers tightly linked to such genes formarker-assisted selection (MAS) at the seedling stage (see Figure 1). Marker-assistedselection is based on the principle that if a gene conferring a trait of interest is linkedto an easily identifiable molecular marker, it may be more efficient to select for themarker than for the trait itself (Mazur and Tingey, 1995).

Figure 1. The Principle of Marker-Assisted Selection

Two main strategies are commonly used to identify such markers for use in marker-assisted selection – genome mapping and Bulked Segregant Analysis.

Genome mapping strategies involve the use of molecular markers to construct geneticlinkage maps that, in turn, provide a framework for the subsequent identification ofgenomic regions controlling important agronomic traits. Identification of these locican be achieved by subjecting combined genotype and phenotype data to quantitativetrait loci (QTL) analyses. Markers shown to display tight linkage with such trait locican be then used for marker-assisted selection and for monitoring trait introgression.

A - Seedlings plantedB - Presence of marker known to be near trait on map is detected in the laboratory, identifying disease resistant plant at seedling stage.C - After 1 year in the field, plant number 2 is shown to be disease resistant.

2

1 3 4

3 4

1

1 2 3 4

DNA markerknown to belinked to rust

resistance

2Production of New Varieties

A

B

2

Marker-Assisted Selection Conventional Breeding

1 Week

1 Year

Identification of Rust Resistant Individual

C

2

27

In poplar, three types of molecular markers (Random Amplified Polymorphic DNAs(RAPDs), Restriction Fragment Length Polymorphism (RFLPs) and Sequence TaggedSites (STSs)) have been used to construct maps of P. nigra and P. deltoides(Bradshaw et al., 1994). These markers have also been used to map monogenicresistance to Melampsora larici-populina and to roughly map a QTL for resistancetowards Xanthamonas populi (Cervera et al., 1996a; 1996b; 1997). Whilst mappingefforts are underway in willow, no linkage maps have yet been published for any Salixspecies.

Although genetic mapping methods are extremely powerful for marker identification,construction of genetic maps is both labour-intensive and time-consuming, asproduction of large numbers of markers, spanning the entire genome at even intervals,is required. Consequently, short-cut approaches, such as Bulked Segregant Analysis,provide exciting opportunities. This technique for identification of markers linked totraits of interest, first described by Michelmore et al., (1991), does not require largenumbers of previously mapped molecular markers. In brief, markers linked to traitsare identified by comparative Amplified Fragment Length Polymorphism (AFLP)analysis of pooled DNA samples (bulks) from individuals showing extremephenotypes, e.g. DNA from the most disease-resistant individuals is pooled, as is thatof the most disease-susceptible individuals, and the AFLP profiles of both pools arecompared. If a marker is present in the profile of one pool, but not the other, then themarker could potentially be linked to the phenotype. Confirmation of any potentiallyinteresting markers is achieved by screening individuals within the bulks to see if therelationship between marker and phenotype still holds.

Types of Molecular Marker

Microsatellite Markers

Microsatellites or Simple Sequence Repeats (SSRs) are highly mutable loci that areabundant throughout the genome (Morgante and Olivieri, 1993) and consist of shortsequence, tandemly repeated mono-, di-, tri-, tetra-, or penta-nucleotide repeatingunits. The hypervariablity of these units is thought to arise from strand slippageduring DNA replication, leading to addition or deletion of repeat units. This, in turn,leads to a high degree of polymorphism at microsatellite loci, which can becharacterised by PCR followed by electrophoresis to determine product length. Since,in most cases, the flanking regions of each microsatellite locus are unique, primerscan be designed to these regions for use in amplification of particular loci (Rongwenet al., 1995).

Apart from their highly polymorphic nature, microsatellite markers possess numerousother favourable characteristics. They present a co-dominant marker system and thuscan be used to detect heterozygosity, an advantage when working with an outcrossingspecies such as willow (Karp et al., 1996). They are also found to be abundant andshow extensive coverage of the genome (Brown et al., 1996; Liu et al., 1995). Thefact that polymorphisms at microsatellite loci are detected by PCR is alsoadvantageous as relatively small amounts of starting material are required whencompared to marker systems such as RFLPs. Also, microsatellite markers are robust

28

and can easily be exchanged between laboratories which is not the case with othertypes of molecular markers, such as RAPDs.

However, while microsatellite markers are extremely informative for each locus, andhave many advantages associated with them, they are costly to develop both in termsof labour and resources. In general, microsatellite markers are non-transferable acrossgenera, i.e. primer pairs developed in a particular genus may not amplifycorresponding loci in another. Also, for applications such as genome mapping, wherelarge numbers of markers are required, they do not produce as many data points perassay as dominant, multilocus marker systems (e.g. AFLPs, RAPDs).

AFLP Markers

The AFLP fingerprinting technique was first described by Zabeau and Vos (1993).Generation of AFLP markers involves digestion of genomic DNA with two restrictionenzymes (a rare and a frequent cutting enzyme). Adapters are ligated to ends of thesefragments, such that the adapter designed to the rare cutting site is biotinylated.Biotinylated fragments are then isolated using streptavidin beads and amplified byPCR using 33P–labelled primers designed to the frequent cutting adapter sequences.These primers allow selective amplification of a subset of the fragments as theyextend, by 1 – 3 bp, into the restriction site and, therefore, only amplify thosefragments in which the primer extensions match the nucleotides present in theflanking region of the restriction site. PCR products are visualised by polyacrylamidegel electrophoresis and autoradiography. In a later modification of the technique (Voset al., 1995), the biotinylated step is removed and two successive rounds of PCR arecarried out using primers with first one (in the first round) and then (in the secondround) one to three selective nucleotides.

Although more technically complicated than some other marker systems, AFLPsprovide a large number of data points per experiment making them especially usefulfor projects where large numbers of markers are required, such as genome mappingand genetic diversity studies (Barker et al., 1999). A further advantage associatedwith AFLP technology is the high level of reproducibility provided by a specific PCR-based system. However, the dominant nature of AFLP markers can be regarded as adisadvantage for mapping studies, as information about heterozygosity is notprovided.

Independently, microsatellites and AFLPs have limitations as markers for mappingbut when used in combination they complement one another to meet the requirementsof this project.

PROJECT OBJECTIVES

The main objectives of this project are as follows:

• Construct a highly saturated genetic linkage map of willow population K3

29

• Locate genomic regions controlling important agronomic traits

• Provide markers linked to traits for marker-assisted selection

Traits of Interest

Important agronomic traits to be examined include resistance to Melampsora rust,resistance to Chrysomelid beetle herbivory and yield. The genetic basis of resistanceto pest and pathogen pressures is poorly understood at present, and may bequantitative (under the control of a large number of genes) or qualitative (controlledby a small number of genes each with large effect). It is assumed that yield is aquantitative trait.

CONSTRUCTION OF THE K3 GENETIC MAP

The K3 mapping population, comprising 66 individuals, was produced by SvalöfWeibull AB from a full-sib cross of Salix viminalis. In a previous project at LongAshton Research Station (LARS), funded under AIR (AIR2-CT92-1617), Dr. J.H.A.Barker initiated construction of a first AFLP/SSR map of Salix using this population.In this project, additional marker information has been added to the original data setand a preliminary linkage map constructed. Development and mapping of furthermarkers has continued to ensure all map regions are sufficiently saturated. This willincrease the likelihood that, when identified, genomic regions controlling traits ofinterest will be positioned sufficiently close to a marker for marker-assisted selectionto be viable.

While this population is suitable for genetic linkage mapping, the relatively smallpopulation size means that it does not lend itself to trait mapping where large progenynumbers are favoured. Furthermore, for mapping, genes of interest must besegregating in the mapping population. It is unknown whether this is case in K3. Tocircumvent these problems, two additional mapping populations, K1 and K8, havebeen established.

MAPPING POPULATIONS K1 AND K8

From 18 potential crosses growing as part of the European Willow BreedingProgramme, K1 and K8 were selected on the basis of seed availability and observedparental segregation for rust and beetle resistance in the field. The pedigrees of thesecrosses are shown in Figure 2.

Figure 2. Pedigrees of K1 and K8 Crosses (R=resistant to rusts (field assessed) andS=susceptible).

K1: (Cross 591)

30

Great grandparents: Astrid (SW880435) x Bjorn(SW910006)

(S. viminalis) (S. viminalis x S.schwerinii)

|Grandparents: Astrid (SW880435) x SW 930984

(S. viminalis) (S. viminalis x S. schwerinii)

Parents (F1): R6 x R11 |

Progeny (F2): 946 planted in trial

K8: (Cross 598)

Great grandparents: Astrid (SW880435) x Bjorn(SW910006)

(S. viminalis) (S. viminalis x S.schwerinii)

|Grandparents: Astrid (SW880435) x SW 930984

(S. viminalis) (S. viminalis x S. schwerinii)

Parents (F1): S1 x R11|

PROGENY (F2): 947 PLANTED IN TRIAL

The large progeny sizes of these populations make them more amenable to traitmapping. Initially, two families were established for this purpose as there was somerisk in ensuring that good segregation would be achieved for all of the traits to beexamined. Following initial assessments one family will be selected for detailedanalyses.

Phenotypic Assessments

In order to identify regions of the genome controlling a particular trait, phenotypicdata is required. This information can be used to identify individuals with extremephenotypes for use in bulked segregant analysis or, alternatively, can be combinedwith genotypic data for QTL analysis. Field assessments for resistance to rust,resistance to beetle feeding will be carried out and a number of yield measurementstaken, including number of shoots per stool, stem diameter, stem height and fresh anddry weights.

In addition to the planned field-based assessments, laboratory-based phenotypicassays have been performed as mixed populations of rust pathotypes and beetle

31

species may be present in the field at time of assessment, making interpretation offield assessment data complex.

Rust resistance/pathogen virulence interactions, initially for two prevalent UKpathotypes, LET1 and LET5, were characterised for a random sample of 100individuals from both K1 and K8 crosses, using a leaf disc inoculation techniquepreviously developed at LARS (Pei et al., 1996). To determine the level of resistancepresented by each individual, the number and size of pustules present and the numberof spores produced are being recorded. Phenotypic data will be used for BulkedSegregant Analysis and QTL analyses.

Beetle feeding assays were performed according to a procedure previously developedat LARS. In brief, willow beetles, of a single species, were allowed to feed on acollection of leaf discs prepared from individual progeny of the population underexamination. Feeding preferences were then determined by analysis of leaf disc arearemaining following the feeding period, a measure of the level ofsusceptibility/resistance displayed by each population. Again, phenotypic data fromthese assays will be used for Bulked Segregant Analysis and QTL analyses.

Preliminary data from these experiments suggest that both K1 and K8 populations aresegregating for rust and beetle resistance.

Overall Marker Identification Strategy

The marker identification strategy used in this project is summarised in Figure3. In brief, the K3 mapping population has been used to construct a highlysaturated linkage map using AFLP and microsatellite markers. A K1 or K8framework linkage map will be constructed and anchored to the K3 linkage mapusing microsatellite markers common to both populations. Using QTL analysis,the K1 or K8 framework map will be used to map traits of interest, with the K3map providing more detailed marker position data required for identification ofclosely linked markers. Furthermore, in an attempt to merge the mappingefforts of willow with poplar, the willow maps will be anchored to an existingpoplar linkage map, again using microsatellite markers. It is thought that willowand poplar genomes possess a high level of synteny (the order of genes may besimilar), and as markers linked to rust resistance and a variety of yieldparameters have already been mapped in poplar, this existing information mayassist in the identification of corresponding markers in willow. An integratedmap will result, incorporating dense marker information from K3, mapped traitsfrom K1/K8 and existing marker and trait data from poplar. Bulked segregantanalysis will also be used as a potential shortcut for marker identification.

32

Figure 3. Overall Marker Identification Strategy.

Identifying Leaf and Stem Traits Relevant to Biomass

Currently, there is a lack of information regarding which ptraits may be indicative of yield for willow. As part of thibeing addressed by Kathryn Robinson and Rachel Ferris o

K3 MapAFLPs &

SSRs

CrossK1 or K8

K3 mappingpopulation

Anchor K3 and Trait mapsusing SSRs

Integrated mapAFLPs, SSRs, Traits

BSA

T ap

Poplarmapping

Poplar Map

Anchor K3 and Poplar mapsusing SSRs

Identification

of Markers for

MAS

SSRsrait M

Yield in Willow

hysiological leaf and stems project, this question isf Gail Taylor’s Group at the

33

University of Southampton. This group has a high level of expertise in this areahaving undertaken similar, detailed studies in poplar.

To determine whether leaf and stem traits that correlate with high biomass yield inpoplar are applicable to willow, physiological comparisons of these traits are beingperformed on consistently high and low yielding willow genotypes from theComparative Yield Trials at LARS. Traits under examination include stem height,stem diameter, number of shoots per stool, leaf extension rate, leaf production rate,leaf area and a number of photosynthetic characteristics. Initial results suggest thatleaf extension rate, a good indicator of yield in poplar, may be applicable to willow.

Once identified, informative traits for yield will be assessed in the K1 or K8 mappingpopulation and the data used in QTL analysis to identify markers linked to these traitsfor use in future selections.

The Future

Marker production will continue in order to saturate the K3 map and to permitconstruction of K1 or K8 framework maps. Further phenotypic assessments, bothlaboratory- and field-based, will be carried out to obtain data for use in markeridentification by QTL analysis and Bulked Segregant Analysis. Any markerspotentially linked to traits of interest will then be further tested in the laboratory, andultimately validated for use in simple assays for marker-assisted selection by thewillow breeder.

Acknowledgements

This project is supported by the DTI’s New & Renewable Energy Programme (managedby ETSU) [Agreement No. B/W6/00599/00/00]. The authors thank Dr KevinLindegaard (European Willow Breeding Programme) for the crosses and Drs. Ming Pei,Tom Hunter and Lori Peacock at Long Ashton Research Station for help with the rustand beetle assays. IACR receives grant-aided support from the Biotechnology andBiological Sciences Research Council of the United Kingdom.

References

34

Barker, J.H.A., Matthes, M., Arnold, G.M., Edwards, K.J., Åhman, I., Larsson, S. andKarp, A. (1999). Characterisation of genetic diversity in potential biomass willows(Salix spp.) by RAPD and AFLP analyses. Genome 42, 173-183.

Bradshaw, H.D., Villar, M. and Watson, B.D. (1994). Molecular genetics of growthand development in Populus. 3. A genetic linkage map of a hybrid poplarcomposed of RFLP, STS and RAPD markers. Theoretical and Applied Genetics98, 167-178.

Brown, S.M., Hopkins, M.S., Mitchell, S.E., Senior, M.L., Wang, T.Y., Duncan, R.R.,Gonzalez-Candelas, F. and Kresovich, S. (1996). Multiple methods for theidentification of polymorphic simple sequence repeats (SSRs) in sorghum[Sorghum bicolor (L.) Moench]. Theoretical and Applied Genetics 93, 190-198.

Cervera, M.T., Storme, V., Liu, B., Gusmao, J., Steenackers, M., Ivens, B., Michiels,B., Van Montagu, M. and Boerjan, W. (1997). AFLP™ genome mapping ofpoplar. Med. Fac. Landbouww. Univ. Gent 62/4a, 1435-1441.

Cervera, M.T., Gusmao, J., Steenakers, M., Van Gysel, A., Van Montagu, M. andBoerjan, W.T. (1996a). Application of AFLP-based molecular markers to breedingof Populus spp. Plant Growth Regululation, 20, 47-52.

Cervera, M.T., Gusmao, J., Steenackers, M., Peleman, J., Storme, V., Vanden Broeck,A., Van Montagu, M. and Boerjan, W. (1996b). Identification of AFLP markers forresistance against Melampsora larici-populina in Populus. Theoretical and AppliedGenetics, 93, 733-737.

Karp, A., Seberg, O. and Buiatti, M. (1996). Molecular techniques in the assessmentof botanical diversity. Annals of Botany 78, 143-149.

Liu, Z-W, Jarret, R.L., Kresovich, S. and Duncan, R.R. (1995). Characterization andanalysis of simple sequence repeat (SSR) loci in seashore paspalum (Paspalumvaginatum Swartz). Theoretical and Applied Genetics 91, 47-52.

Mazur, B.J and Tingey, S.V. (1995). Genetic mapping and introgression of genes ofagronomic importance. Current Opinion in Biotechnology 6, 175-182.

Michelmore, R.W., Paran., I. and Kesseli, R.V. (1991). Identification of markerslinked to disease-resistance genes by bulked segregant analysis: A rapid method to

35

detect markers in specific genomic regions by using segregating populations.Proceedings of the National Academy of Sciences USA 88, 9928-9832.

Morgante, M. and Olivieri, A.M. (1993). PCR-amplified microsatellites as markers inplant genetics. The Plant Journal 3, (1),175-182.

Pei, M.H., Royle, D.J. and Hunter, T. (1996). Pathogenic specialization inMelampsora epitea var epitea on Salix. Plant Pathology 45, (4), 679-690.

Rongwen, J., Akkaya, M.S., Bhagwat. A.A., Lavi, U. and Cregan, P.B. (1995). Theuse of microsatellite DNA markers for soybean genotype identification.Theoretical and Applied Genetics 90, 43-48.

Vos, P. Hogers, R., Bleeker, M., Rijans, M., Van de Lee, T., Hornes, M., Frijters, A.,Pot, J., Peleman, J., Kuiper, M. and Zabeau, M. (1995) AFLP: a new technique forDNA fingerprinting. Nucleic Acids Res 23:4407-4414

Zabeau, M. and Vos, P. (1993). Selective restriction fragment amplification : ageneral method for DNA fingerprinting. European Patent Application, publicationno : EP 0534858-A1, No. 92402629.7.

36

The use of willow mixtures as a disease control strategy

A. R. McCracken1,3, A. C. Bell1,3, W. M. Dawson2,3 & R. J. Marks1,3

E-mail: james.marks@dardni.gov.uk1Applied Plant Science Division, Dept. of Agriculture and Rural Development Northern Ireland,

Newforge Lane, Belfast BT9 5PX, N. Ireland UK2Northern Ireland Horticulture and Plant Breeding Station, Dept. of Agriculture and Rural

Development, Loughgall, Co. Armagh, BT61 8JB, N. Ireland, UK3Dept. of Applied Plant Science, School of Agriculture and Food Science, Queen’s University Belfast,

Newforge Lane, Belfast BT9 5PX, N. Ireland, UK

INTRODUCTION

Rust, caused by Melampsora epitea var. epitea, is probably the greatest limitation to the growth ofwillow (Salix spp.) in short rotation coppice for biomass. The disease can affect a wide range of Salixspecies and varieties. In the most susceptible varieties rust causes heavy premature leaf fall followedby infection by a whole spectrum of secondary pathogens which can result in death of shoots and evenstools (McCracken & Dawson 1992). In N. Ireland rust was first observed on S. burjatica ‘Korso’ in1986. Within a few years this variety was so badly affected that often 60% of stools were killed by thetime of flushing the spring following harvest (Dawson & McCracken, 1993). While there is a range ofsusceptibilities to rust a number of varieties have become so susceptible that they can no longer begrown commercially in many parts of the country.

Short rotation coppice willow is a long-term crop and therefore is it essential to have adisease control strategy which reduces the impact of the rust and at the same timeensures the sustainability of plantations over the life of the crop, which may be up to25 years. Although rust can be controlled by the intensive use of fungicides(McCracken & Dawson, 1990; 1997) it was clear that this was not a viable approachto disease control. Early work reported by Wolfe (1985) indicated that mixtures ofmultilines or cultivars of cereals was an effective method of reducing the impact ofdiseases, especially those caused by obligate pathogens such as rusts and mildews.Therefore in the mid 1980s trials were planted in N. Ireland which comprisedmixtures of willow species and varieties, in order to test their effectiveness inreducing rust. These early trials indicated that when a susceptible willow variety wasincluded in a mixture the first appearance of rust was delayed, the build up of thedisease slowed and the final levels of rust significantly reduced compared to the samevariety growing in a mono-culture (McCracken & Dawson 1997). It was alsoreported that dry matter yields from mixtures were consistently higher than the meanof the constituent varieties grown in mono-culture (McCracken & Dawson 1994).The subsequent trials, reported in this paper, were planted with a view to determiningthe effect of numbers of varieties within a mixture, the impact of planting density andthe performance of some of the newer planting material within mixtures.

In most seasons many of the willow varieties planted carried a heavy infestation ofwillow beetles (Sage & Tucker, 1998) which could cause significant feeding damage.Observations and measurements were carried out to assess the level and importance ofthis damage.

TRIALS

Plantations

37

The plantations were established at Castlearchdale, Co. Fermanagh, N. Ireland(54028’N 7043’W), 60-70m above sea level and on a site with heavily gleyed soilswith impeded drainage. The trial was planted in two replicate blocks. Block 1 wasplaned in spring 1994 and block 2 in spring 1995. Both blocks were cut back at theend of their establishment year and block 1 again in 1995/96 meaning that in thesummer 1996 both blocks were at the same stage, i.e. regrowth from freshly coppicestools. Twenty Salix spp varieties (Table 1) were planted in large mono-culture (32 x14m in Block 1 and 32 x 11m in Block 2) in double rows in a split plot design withthree densities - 10,000ha-1, 15,000ha-1 and 20,000ha-1. Four mixture combinationswere employed comprising 5, 10, 15 and 20 varieties (Table 1).

Table 1: List of Salix spp. varieties planted in 5, 10, 15 or 20 way mixtures

VarietyNo.

Variety Mixture

3 S. burjatica Germany 5 10 15 204 S. mollissima-undulata SQ83 5 10 15 20

13 S. dasyclados x aquatica V7511 5 10 15 2017 S. viminalis 77082 5 10 15 2026 S. dasyclados x caprea V794 5 10 15 20

14 S. viminalis x aquatica V7503 - 10 15 2023 S. viminalis 78118 - 10 15 2028 S. viminalis 78183 - 10 15 2032 S. schwerinii x viminalis x dasyclados V7531 - 10 15 2048 S. viminalis 870146 ULV - 10 15 20

12 S. viminalis x caprea V789 - - 15 2018 S. viminalis 77683 - - 15 2021 S. viminalis 78101 - - 15 2029 S. viminalis 78195 - - 15 2034 S. schwerinii x aquatica V7534 - - 15 20

1 S. viminalis 77699 - - - 206 S. viminalis Gigantea - - - 209 S. viminalis Gustav - - - 20

45 S. schwerinii x aquatica V7533 - - - 2046 S. viminalis 870082 ORM - - - 20

All mixtures were planted as fully random intimate mixtures. At the time of plantinga map was prepared of the planting pattern so that subsequently each stool could beidentified.

Disease assessmentsStarting in mid June and at two week intervals leaf samples were collected. At least100 leaves were taken from approximately ten stools of each variety growing inmono-culture and in mixtures. The percentage number of leaves with rust wasdetermined and the percentage cover of each leaf assessed using a disease assessmentkey (McCracken & Dawson 1992). These assessments were carried out each yearfrom 1996 (first year growth from freshly coppiced stools), 1997 (second yeargrowth), 1998 (third year growth), 1999 (first year growth from freshly coppicedstools) stools and 2000 (second year growth). By using the number of infected leavesand the percentage rust cover, a cumulative rust score was calculated (McCracken &Dawson 1992).

38

Pest assessmentsIn 1999 and 2000 the same leaf samples which had been scored for rust were assessed visually forwillow beetle damage. Herbivory was scored as the percentage of each leaf removed and a mean scorewas calculated for each variety in mono-culture and in mixtures.

RESULTS & DISCUSSION

DiseaseIn general the build up of disease was slower in mixtures than it was in the mono-varietal plots. Thedifferent Salix spp. varieties had a range of susceptibilities to rust, which in turn influenced the effectexerted by inclusion within a mixture. Planting density had no discernible impact on the incidence orseverity of disease. The effect of numbers of components in a mixture on rust development on anyindividual Salix variety within a mixture tended not to be consistent. However there are other benefitsin having larger numbers present within a mixture. Pathotype diversity on an individual variety wasnormally greater when that variety was in a mixture (McCracken et al. 2000). This increased diversity(Fig 1) in turn increased competition between pathotypes, which appeared to reduce selectionpressures, which may have resulted in the development of super races of Melampsora epitea var.epitea. Furthermore, yield compensation for stools that had died within a mixture, due to disease orother causes, was more effective in the mixtures with a larger number of components.

Fig 1: Melampsora epitea pathotype diversity (Shannon-Weaver Index) on Salix spp. varieties grownas mixtures and as mono-culture (McCracken et al. 2000).

The pattern of rust development on individual Salix spp. varieties when included within a mixture waslargely determined by their level of susceptibility to rust.

Salix burjatica GermanySalix dasyclados x aquatica V7511Salix dasyclados x caprea V794

S. burjatica Germany has been grown in N. Ireland for almost twenty years. Rust was first recorded onthis variety at the end of the growing season on 1987. In subsequent seasons rust was observed, usuallytowards the middle or end of the summer, on the upper leaves and at low levels, inflicting little or nodamage to plants (McCracken & Dawson 1992). However in 1996 S. burjatica Germany was severelyinfected with rust, to a degree that, following premature leaf fall there was extensive rod and stooldeath in the mono-varietal plots. In the period from 1996 – 2000 this variety has been so badly affectedthat it has been almost totally killed in some of the mono-varietal plots. Its behaviour in mixtures hasbeen erratic. The build up of rust has been slowed down on plants growing in the larger (10 – 20

0

0 .2

0 .4

0 .6

Diversity

M o n o M ix

L S D = 0 .0 6 8

39

varietal mixtures) although still causing significant damage and loss of stools. While S. dasyclados xaquatica V7511and S. dasyclados x caprea V794 have not been grown in N. Ireland for as long as S.burjatica Germany the pattern of their demise has been similar.

It has been concluded that where a variety is, or becomes particularly susceptible to rust, especially inconditions of high disease pressures, its inclusion in a mixture may slow down the effects of the diseasebut is probably not enough to maintain it as a productive variety.

Fig 2: Melampsora epitea rust disease development in 1999 on S. dasyclados x aquatica V7511grown in mono-culture and included in 5, 10, 15 and 20-way mixtures.

S. mollissima-undulata ‘SQ83’

In 1996 S. mollissima-undulata SQ83 was severely affected by rust, starting early in the growingseason and followed by significant death of rods when in mono-culture. In that year inclusion withinall mixtures had a significant effect in reducing disease. In subsequent years the disease followed adifferent pattern. It tended not to become evident until much later, often late August and consequentlyits effect on growth was less. There was a rapid increase in rust scores in mono-culture which weresignificantly less in mixtures.

S. viminalis ‘77082’, ‘78118’, ‘78183’, ‘870146 ULV’, ‘77683’, ‘78101’, ‘78195’,‘Gigantea’, ‘Gustav’, and ‘870082 ORM’.

In each year from 1996 – 1999 S. viminalis ‘77082’ carried low levels of disease, especially early in thegrowing season, up until September. Only in 1996 were significant levels of rust observed in July. Innone of the years were any differences recorded in the pattern of rust development in the mixtures or onplants grown in mono-culture.

In all years S. viminalis ‘78118’ had only low levels of rust, normally only evident as the seasonprogressed. Inclusion in mixtures appeared not to affect the pattern of disease development.

S. viminalis ‘78183’ had low levels of disease in all seasons with a general trend of there being morerust on plants grown in mono-culture.

S. viminalis ‘870146 ULV’ had virtually no rust in any year with the exception of 1996, when levels ofdisease on mixtures was less than in mono-culture.

0

5000

10000

08/06/99 22/06/99 07/07/99 21/07/99 03/08/99 17/08/99 31/08/99 14/09/99DATE

MEA

N A

CC

UM

ULA

TIVE

RU

ST S

CO

RE

POLY X5 POLY X10 POLY X15

POLY X20 MONO D1

40

S. viminalis ‘77683’. In all years only infrequent pustules were observed at the end of the growingseason in late September. This may have been partly due to this variety being very variable and almostconstituting a mixture in its own right.

S. viminalis ‘78101’ tended to have a few pustules of rust, even in early July. However there was onlya slow increase in the amount of disease as the season progressed and at no point was there a differencebetween the amount of disease on this variety whether grown in mono-culture or within a mixture.

S. viminalis ‘78195’ had low levels of disease which only became apparent towards the end of August.No significant differences were observed between disease levels in mixtures or mono-culture.

S. viminalis ‘77669’ had low levels of disease in all years.

S. viminalis ‘Gigantea’ had no significant levels of rust observed in any year.

S. viminalis ‘Gustav’ had low to moderate rust in each year. In 1997 it was observed early while in1998 and 1999 rust pustules were not recorded till much later. The general trend at each recording datewas for reduced disease levels in mixtures compared to mono-culture.

S. viminalis ‘870082 ORM’ had only very low levels of rust in any year.

Salix viminalis mixturesEach of the straight S. viminalis varieties grown in the trial had only low susceptibility to rust, andoften were only affected late in the season. There were however different patterns of diseasedevelopment. It can therefore be argued, that even a straight S. viminalis mixture in which itscomponents may be susceptible to the same Melampsora epitea pathotypes will still act as a genuinemixture in terms of reducing disease and reducing the selection pressures on the pathogen. This is thetopic of on-going research. A S. viminalis trial has been established at the Northern Ireland Horticultureand Plant Breeding Station, Loughgall in order to evaluate disease development within a mixedplantation where all of the components are straight S. viminalis selections.

S. schwerinii x viminalis x dasyclados V7531S. schwerinii x aquatica V7533S. schwerinii x aquatica V7534

Varieties with S. schwerinii parentage tend to display a high level of resistance to rust. Since plantingthe trial in 1996 rust has only been observed at minimal levels (occasional pustules on < 0.1% ofleaves) in any year, irrespective of age of growth.

There was obviously no immediate disease reduction advantage in including such varieties within amixture. It can be argued however that growing them in mixtures reduces the potential diseasepressures and hence ensures their long-term sustainability over the full 25 – 30 year life of a plantation.

Yield

The mean yield from 100m-2 of any mixture was significantly greater than the mean yield from anequivalent area of the constituent varieties grown in mono-culture (Fig. 3). With the exception of thefive-way mixture, which was significantly lower, there were no significant yield difference between the10, 15 and 20 way mixtures. These trends were consistently observed at each of the three densities.

Five-way mixtureThe yield from the five-way mixture was low because of the death or weak growth of three of thecomponents – S. burjatica Germany, S. dasyclados x aquatica V7511, S. dasyclados x caprea V794,due to the impact of the continuous rust infections. The two remaining varieties had made majorcompensation for the extra space. Variety 17, S. viminalis 77082 produced almost 50% more drymatter in this mixture compared to its growth in mono-culture. Similarly Variety 4, gave a 20%increase in yield.

41

Fig. 3: Dry matter yield (kg) from fixed area (100m2) of mixtures of Salix spp.varieties compared to an equivalent area of the components grown asmono-culture.

Ten-way mixtureYield from the ten-way mixture was significantly higher than the equivalent from the constituentvarieties. This was despite the demise of S. burjatica Germany, S. dasyclados x aquatica V7511, S.dasyclados x caprea V794, whose growth was significantly affected by rust. S. viminalis 77082, S.schwerinii x viminalis x dasyclados V7531 and S. viminalis 870146 ULV all contributed significantlymore dry matter yield than had been predicted.

Fifteen-way mixtureThe increased yield was due mainly to the increased contribution from S. schwerinii x viminalis xdasyclados V7531, S. viminalis 870146 ULV, S. viminalis 77683, S. viminalis 78195 and S. schweriniix aquatica V7534. S. viminalis x aquatica V7503 and S. viminalis x caprea V789 both contributedsignificantly less in this mixture than might have been predicted. Susceptibility to rust may have beena contributory factor. However their poor performance in mixtures was due more to their lack ofability to compete with other components. It has been observed that in the second growth cycle S.viminalis x caprea V789 has had a low survival rate in the 15 and 20-way mixtures.

Twenty-way mixtureThe major contributors to the yield from this mixture were S. viminalis 77683, S. viminalis 78195, S.schwerinii x aquatica V7534, S. schwerinii x aquatica V7533 and S. viminalis 870082 ORM.

Willow beetle damageThe levels of damage to leaves from the various varieties in mono-culture are shown in Fig.4. Overall,damage levels were moderate. However, five varieties exhibited damage, which was significantlygreater than the others. These were S.viminalis 78118, S.viminalis 870146 ULV, S.viminalis x capreaV789, S.viminalis 77683 and S.viminalis Gigantea. Preferential feeding of Phratora vulgatissima ondifferent varieties of willow is documented in the literature (Sage and Tucker, 1998; Kearns, 1934), and

0

50

100

150

200

250

Yield

X 5 X 10 X 15 X 20Mixture

Mono Mixture

42

the results in the present trial indicate that these five varieties could be at risk of significant damage inyears when willow beetle populations are high.

When the results from the mixture plots were examined, these varieties showed a trend towardsreduced damage with increasing numbers of other varieties in the mixtures. Although this result didnot achieve statistical significance, it does suggest that the inclusion of damage-susceptible varieties inmixed plots may offer them some protection from willow beetle attack. It is likely that this would havebecome more obvious if beetle populations had been higher during the period of this study.

Fig. 4: Willow beetle damage to varieties in mono-culture. The dark bars indicate those varieties thatshow significantly higher levels of damage.

CONCLUSIONS AND RECOMMENDATIONS

It is clear that there was significant yield benefits from growing willow in inter- andintra-species mixtures. This is partially due to reduced disease levels on somevarieties when grown in a mixture. It is also due to physiological interactions, whichencourages some varieties to grow more vigorously. Furthermore if a variety dies outof a mixture either due to disease, competition or some other factor then the remainingvarieties can compensate for their loss. For this reason larger mixtures are morebeneficial. In this current trial 60% of the five-way mixture was killed and while thetwo remain varieties did make major compensations it was not sufficient. However inthe twenty-way mixture where only 15% was lost the impact on yield was much lessobvious. The authors also consider than random intimate mixtures give much greateropportunity for compensation compared to either line mixtures or mosaics. The useof short-run mixtures is a compromise to the needs of commercial planting.

When planting willow plantations it is therefore recommended:

012

34567

89

10

% le

af d

amag

e

3 4 13 17 26 14 23 28 32 48 12 18 21 29 34 1 6 9 45 46

Variety Number

43

• Plant inter- and intra-species mixtures of willow.• Plant as diverse a mixture as possible.• Plant mixtures containing least 6 – 10, if possible.• Avoid varieties which are rust susceptible or have become rust susceptible

in other locations.• Mixtures should be as random as commercial practice permits.

REFERENCES

Dawson, W.M. & McCracken, A.R. (1993). The effect of Melampsora rust on the growth anddevelopment of Salix burjatica ‘Korso’ in Northern Ireland. European Journal of ForestPathology 24, 32 – 39.

Kearns, H.G.H. (1934). The control of insect pest of basket willows. Annual Report of Long AshtonResearch Station, 1934, 3-5.

McCracken, A.R. & Dawson, W.M. (1992). Clonal response in Salix to Melampsora rusts in shortrotation coppice plantations. European Journal of Forest Pathology 22, 19 – 28.

McCracken, A.R. & Dawson, W.M. (1990). The effect of Melampsora sp rust on growth andevelopment of Salix in Northern Ireland. Proceedings of Joint Meeting IEA Task 5 ActivityGroups on the exchange of Genetic Material and Joint Trials on Alnus, Populus and Salix andPests / Disease Management, Poplar Research Centre, Geraardsbergen, Belgium 11 – 14September 1990, Ed. By V. Steenackers. Geraardsbergen: International Energy Agency. pp15 – 26.

McCracken, A.R. & Dawson, W.M. (1994). Experiences in the use of mixed-clonal stands of Salixas a method of reducing the impact of rust disease. Norwegian Journal of Agricultural Science 18(suppl.) 101 – 109

McCracken, A.R. & Dawson, W.M. (1997). Growing clonal mixtures of willow to reduce the effectof Melampsora epitea var. epitea. Europena Journal of Forest Pathology 27, 319- 329.

McCracken, A. R., Dawson, W. M., Watson, Sally &. Allen, Chanel Y (2000). Pathotypecomposition in Melampsora epitea populations occurring on willow (Salix) grown in mixedand mono-culture plantations. European Journal of Plant Pathology. (In press).

Sage, R.B. & Tucker, K. (1998). The distribution of Phratora vulgatissima (Coleoptera:Chrysomelidae) on cultivated willows in Britain and Ireland. European Journal of ForestPathology, 28, 289-296.

Wolfe, M. S. (1985). The current status and prospects of multiline, cultivars and variety mixtures fordisease resistance. Annual Review of Phytopathology 23, 251 – 273.

44

Breeding Strategies – The Way Forward

KEVIN LINDEGAARD

IACR-Long Ashton Research Station, Department of Agricultural Sciences, Universityof Bristol, Long Ashton, Bristol BS41 9AF, UK

E-mail: kevin.lindegaard@bbsrc.ac.uk

Abstract

The European Willow Breeding Programme was initiated in 1996 to producenew superior willow varieties for biomass energy. To date, the crossingprogramme has focussed on inter-species combinations between a base populationof approximately 100 varieties obtained from the UK National Willows Collectionheld at IACR-Long Ashton. In total, 750 crosses have been made and over 40,000unique genotypes have been screened in a seedling nursery. Selected individualshave been evaluated in an extensive array of breeding trials and superior genotypeshave been advanced to yield trials. Currently, over 50 genotypes are in advancedobservation and yield trials. Of these it is expected that a small proportion will bemultiplied for commercial use in the biomass industry from 2004 onwards.

The current methods used to select willows in the field are unsatisfactory andneed urgent refinement. Future selection strategies should attempt to combine thebreeders’ observations with quantitative data and yield estimates. Suchinformation will enable multiple trait selection strategies to be employed. Theconstruction of a genetic map of willows and the identification of markers linkedto traits of interest are extremely important developments. The use of molecularmarker technology will facilitate the rapid selection of desirable genotypes at theseedling stage and accelerate the introgression of valuable traits in backcrossprocedures.

In future hybridisation schemes, short term breeding initiatives should becoupled with longer-term recurrent selection strategies. Other long-term prospectsmight include the production of inbred lines by anther culture, polyploid breedingand the application of GM technology.

Key words: Salix, willow, Short Rotation Coppice, biomass,breeding, interspecific cross, recurrent selection.

INTRODUCTION

The European Willow Breeding Programme (EWBP) was established in 1996 toproduce new willow varieties principally for biomass energy production(Lindegaard and Barker, 1997). As with all plant breeding programmessuccess in willow breeding depends on the creation of variability throughhybridisation combined with effective methods of assessment andselection of the progeny in the field. During the first five years of theprogramme various breeding and selection strategies have beenemployed. The purpose of this paper is to review the relative success ofthese strategies and consider where procedures may be improved andwhether new approaches are necessary to accelerate the production ofvarieties yet further.

45

Crossing Programme and Field Evaluations 1996-2000

Between 1996-2000, over 750 crosses have been made using a base population of100 individuals belonging to 20 distinct willow species from North andCentral Europe, Russia, Asia and North America which are maintained inthe UK National Willows Collection held at IACR-Long Ashton. 40,000unique willow genotypes have been screened in a seedling nursery andselected on the basis of yield estimations, resistance to disease and pestsand erect growth habit. Between 5-10% of these selections have beenadvanced initially to field- based observation trials located at IACR-LongAshton, from which elite breeding lines have been identified andsubsequently included in smaller multi-site observation and yield trials.The breeding scheme being followed is demonstrated in Figure 1 and theprogress of the breeding programme to date is illustrated in Figure 2.Already, over 50 genotypes are in advanced observation and yield trials.Of these it is expected that a small proportion will be multiplied forcommercial use in the biomass industry from 2004 onwards. It is hopedthat these will help increase the average yield to around 15 dry tonnes perhectare per year.

The Way Forward

In only a very short time giant strides have been made in the willow breedingprogramme. For instance, the success rate of the crosses performed has increased yearon year and currently stands at 79.5%, almost four times that achieved in 1996.However, there are a number of shortcomings that have not received sufficientattention. These include field selection strategies and the development of medium- tolong-term breeding initiatives. These areas will be considered in turn below alongwith key research requirements and the potential for novel applications of SRCwillows.

Selection Strategies

Until now, field selections from observation trial 1 have been based solely on thevalue judgements made by the breeder. In future, if we are to be absolutely sure thatall the best genotypes are selected for advancement we should not merely rely on“breeders’ eye” alone, but must support this expertise with quantitative data such asmeasurements of height, stem diameter, number of shoots etc. It is hoped that from2000 onwards the use of one-hand computer callipers (Savcor Ltd) will provide arapid non-destructive technique for estimating the yield of genotypes in observationtrial 1. The technique used will be that performed on the national trial network byForest Research (see Armstrong, this volume). The information gained from such anexercise will pave the way for more informed selection procedures for SRC willows.

As with most crops, willow breeding and selection schemes attempt to select formany traits simultaneously. Characters of interest include yield, disease and pestresistance, growth habit and wood quality. Trying to select genotypes that fulfil such

46

diverse criteria from observations alone is extremely demanding and subject to error.Therefore, in future, it would be sensible to consider the utilisation of a multiple traitselection strategy that aims to maximise genetic gain. There are a number of methodsthat endeavour to deal with this problem, such as index selection, independent culling,tandem selection and principal component analysis (Bisoffi and Gullberg, 1996) aswell as cluster analysis (Ares and Gutierrez, 1996). These all require detailedmeasurements of several traits of interest. Cotterill and Dean (1990) suggest thatindex selection is the most efficient method for successful tree breeding. However,Bisoffi and Gullberg (1996) report no examples of index selection being used inpoplar breeding. Instead the use of principal component analysis (Bisoffi, 1990) andcluster analysis (Ares and Gutierrez, 1996) have shown promise and should thus beconsidered most relevant to willow breeding.

A distinct, albeit related concept involves the development of a tree ideotype(Dickmann, 1985). Dickmann and Keathley (1996) propose an ideotype selectionprogramme in poplar. In this example, the breeder selects a subset of geneticallycorrelated characters and these are combined as a single quantitative trait. Thisworking ideotype is then treated as one of two to three variables in an independentculling process, in which clones that do not achieve a certain predetermined standardare eliminated. Although there are only a few examples of ideotype breeding inpractice, Koski and Dickmann (1992) suggest encouraging results from the use of amodelling approach with SRC willows and poplars.

In the medium term, the construction of a genetic map for willow and theidentification of markers linked to traits of interest will facilitate the rapid selection ofdesirable genotypes at the seedling stage and accelerate the introgression of valuabletraits in backcross procedures. This technique, termed marker-assisted selection, willimprove the efficiency of the initial stages of the willow breeding process (see Hanleyet al, this volume).

Medium-Long Term Breeding Initiatives

Figure 3 illustrates the various hybridisation pathways so far utilised in thewillow breeding programme. The vast majority of the crosses fall intoonly three categories: interspecific crosses between two species (type b),three-way species crosses (type f) and four-way species crosses, alsoknown as double crosses (type g). Together these comprise over 90% ofthe crosses attempted. These types of crosses are beneficial because theymaximise heterozygosity and the breeder may select specific genecombinations demonstrating heterosis. Hybrid vigour in such crosses isdue to the trapping of dominant and epistatic gene action withouteliminating or masking favourable recessive genes (Stettler, Zsuffa andWu, 1996). However, in many cases this approach will result in barriers tocontinuing gains from genetic improvement because of a reduction incrossability caused by differences in ploidy levels and pollen-pistilincongruity in certain species (Mosseler, 1990). As a result of this,continuing solely with this approach would necessitate retracingsuccessful crosses with progressively poorer parents due to the exhaustionof the germplasm available.

47

Hence, it is imperative that we back up such productive short-term approacheswith longer-term strategies based on population improvement. Long-termbreeding strategies aim to increase genetic diversity and create newcrossing opportunities for the future (Kang, Lascoux and Gullberg, 1996).This can be effectively achieved using recurrent selection schemes whichinvolve the concentration of desirable alleles during repeated cycles ofhybridisation and selection (Jenkins, 1940; Hull, 1945). Several forms ofrecurrent selection exist but the application best suited for tree crops istermed reciprocal recurrent selection. Baudouin et al (1997) reviewed thebreeding strategies of six tropical tree crops (cacao, coffee, rubber tree, oilpalm, coconut and eucalyptus) and showed that all but rubber treeimprovement were based on this method. Also this is the preferredstrategy for long term poplar breeding (Stettler, Zsuffa and Wu, 1996;Bisoffi and Gullberg, 1996).

Figure 4 illustrates a generalised model for reciprocal recurrent selection forgeneral hybridising ability (RRS-GHA) as used in the Italian poplarbreeding programme (Bisoffi, 1990). This breeding system has theadvantage of capturing both additive and non-additive gene effectstogether with the simultaneous improvement of two base populations andtheir hybrid offspring. The downside of this strategy is that it is labourintensive and demanding of field space.

Other long-term initiatives are also required for the continued success of thebreeding programme. These include developing protocols for the efficientchromosome doubling of triploid hybrids, improving the crossability ofrecalcitrant cross combinations using embryo rescue techniques, thedevelopment of inbred lines using anther culture and the possibleapplication of GM technology.

My own unpublished results suggest that the triploid hybrid variety S. x stottii‘Ashton Stott’ could be doubled with colchicine using a root immersiontechnique based on that performed by Morgan (1975). This and othermethods, such as that carried out by Zhang et al (1997) should be testedmore fully.

Research Requirements

Now for the first time the opportunity exists for researchers to test the near-market varieties produced by the EWBP. Such collaboration will provide detailedinformation on elite varieties prior to their commercial release. This is far moresensible than carrying out research on established varieties or older non-registeredclones. The result of the latter is that by the time the experimental data is acquired thevarieties are liable to be outclassed or obsolete and therefore unlikely to be planted bybiomass growers.

The superior selections created by the breeding programme should form thebedrock of future disease and pest research such as the work on mixed clonalplantations. A major flaw exists in our current breeding strategy; the breeding processis aimed at identifying superior genotypes on the basis of their performance in pure

48

stands. Once superior individuals are identified we advise that they should be plantedin mixtures. A high yielding clone may be completely unsuitable for a mixturebecause it may be too competitive or, if it is slow to establish, not competitiveenough. Research is needed to identify genotypes that are high yielding but alsopossess characters conducive to synergism.

Other crucial questions remain unanswered. For instance we have sparseknowledge on the physiological basis of heterosis. Fundamental research is requiredinto the physiological composition of the highest yielding varieties, such as theirwater use efficiencies, leaf area indices, respiration rates and concentrations of plantgrowth regulators.

Novel Industrial Uses

Willow varieties that are unsuitable for use as a biomass crop may have attributesthat can be exploited in other novel industrial applications such as:♦ raw materials for the forest products industry (Deka, Wong and Roy, 1992;1994),♦ natural dyes (Stott, Bridle and Timberlake, 1971)♦ pharmaceuticals and other biologically active compounds (Gebhardt, 1992; Meier,

1992; Julkunen-Tiitto and Meier, 1992),♦ livestock feeding (McCabe and Barry, 1988), and♦ pollution control (Riddell-Black, Pulford and Stewart, 1997; Edwards, 2000).

The anticipated benefits from such novel applications are numerous and includethe potential to provide the willow grower with value-added products, the provision ofsustainable sources of natural products, the diversification of UK arable farming, ruralemployment opportunities and numerous environmental benefits.

Conclusion

The European Willow Breeding Programme has achieved a great deal in its firstfive years. By the end of the next five years the first products of this initiative will bein the marketplace. Nevertheless, many important strategies and research initiativesneed to be implemented now to ensure the long-term success of the breedingprogramme and the biomass industry as a whole.

ACKNOWLEDGEMENTS

The European Willow Breeding Programme is funded by a partnership comprisingIACR-Long Ashton Research Station, Svalöf Weibull AB and Murray Carter. IACRreceives grant-aided support from the Biotechnology and Biological SciencesResearch Council of the United Kingdom.

49

References

Ares A and Gutierres L (1996) Selection of Poplar Clones for the Lower Valley of theColorado River, Argentina. Forestry 69, No.1: 75-82.

Baudouin L, Baril C, Clément-Demange A, Leroy T and Paulin D (1997) RecurrentSelection of Tropical Tree Crops. Euphytica 96: 101-114

Bisoffi S (1990) The Development of a Breeding Strategy for Poplars. InternationalPoplar Commission (FAO), 35th Committee Meeting, Buenos Aires 19-23 March1990. FO:CIP:BR/90/9. 21pp.

Bisoffi S and Gullberg U (1996) Poplar Breeding and Selection Strategies. In:Biology of Populus and it’s Implications for Management and Conservation pp.139-158. Ed. Stettler, R F, Bradshaw, H D Jr, Heilman, P E, and Hinckley, T M.NRC Research Press, Ottawa, Ontario, Canada.

Cotterill P P and Dean C A (1990) Successful Tree Breeding with Index Selection.CSIRO Division of Forestry and Forest Products. 80 pp.

Deka G C, Wong B M and Roy D N (1992) Suitability of Hybrid Willow as asource of Pulp. Journal of Wood Chemistry and Technology 12: 2: 197-211.

Deka G C, Wong B M and Roy D N (1994) Variation of Specific Gravity, FibreLength and Cell Wall Thickness in Young Salix Clones. Journal of WoodChemistry and Technology 14: 1: 147-158.

Dickmann D I (1985) The Ideotype Concept Applied to Forest Trees. In: Trees asCrop Plants. Ed. Cannell M G R and Jackson J E. Institute of Terrestrial Ecology,Huntingdon, England, pp. 89-91.

Dickmann D I and Keathley D E (1996) Linking Physiology, Molecular Genetic andthe Populus Ideotype. In: Biology of Populus and it’s Implications forManagement and Conservation pp. 139-158. Ed. Stettler R F, Bradshaw H D Jr,Heilman P E, and Hinckley T M. NRC Research Press, Ottawa, Ontario, Canada.

Edwards R R (2000) The Potential for the Use of Willow (Salix spp.) in Buffer Zonesfor Reducing Nitrate and Atrazine Pollution. PhD Thesis. University of Bristol.297 pp.

Gebhardt K (1992) Basis and Methods of Breeding Pharmaceutically ValuableWillows. (in German) Holzzucht 46: 1-4: 9-14.

Hull F H (1945) Recurrent Selection for Specific Combining Ability in Corn. Journalof the American Society of Agronomy, 37: 134-145.

Jenkins M T (1940) The Segregation of Genes Affecting Yield of Grain in Maize.Journal of the American Society of Agronomy, 32: 55-63.

50

Julkunen-Tiitto R and Meier B (1992) Variation in Growth and Secondary PhenolicsAmong Field Cultivated Clones of Salix Myrsinifolia. Planta Medica 58: 77-80.

Kang H, Lascoux M and Gullberg G (1996) Systematic Tree Breeding. In: TreeBreeding. Ed. Mandal A K. Indian Council of Forestry Education. Jabalpur, India.

Koski V and Dickmann D I (1992) Tree Ideotype. Biomass and Bioenergy, 2: 1-6: 71-75.

Larsson S (1997) Commercial Breeding of Willow for Short Rotation Coppice. In:Aspects of Applied Biology 49, Biomass and Energy Crops pp. 215-218. Ed.Bullard M J, Ellis R G, Heath M C, Knight J D, Lainsbury M A and Parker S R.The Association of Applied Biologists.

Lindegaard K N and Barker J H A (1997) Breeding Willows for Biomass. In: Aspectsof Applied Biology 49, Biomass and Energy Crops pp. 155-162. Ed. Bullard M J,Ellis R G, Heath M C, Knight J D, Lainsbury M A and Parker S R. TheAssociation of Applied Biologists.

McCabe S M and Barry T N (1988) Nutritive Value of Willow (Salix sp.) for sheep,goats and deer. Journal of Agricultural Science, Cambridge, 111: 1-9.

Meier B (1992) Pharmaceutical Requirements on Willow Production. (in German)Holzzucht 46: 1-4: 5-9.

Morgan W G (1975) A Technique for the Production of Polyploids in Grasses.Euphytica 25: 443-446.

Mosseler A (1990) Hybrid Performance and Species Crossability Relationships inWillow (Salix). Canadian Journal of Botany 68: 2329-2338.

Nilsson G (1988) Willow Breeding Program at Svalöf AB. International EnergyAgency Proceedings from Willow Breeding Symposium August 31-September 1,1987, pp. 25-27. Research Notes 41, Swedish University of Agricultural Sciences,Uppsala, Sweden.

Riddell-Black D, Pulford I D and Stewart C (1997) Clonal Variation in Heavy MetalUptake by Willow. In: Aspects of Applied Biology 49, Biomass and Energy Cropspp. 155-162. Ed. Bullard M J, Ellis R G, Heath M C, Knight J D, Lainsbury M Aand Parker S R. The Association of Applied Biologists.

Stettler R F, Zsuffa L, and Wu R (1996) The Role of Hybridisation in the GeneticManipulation of Populus. In: Biology of Populus and it’s Implications forManagement and Conservation pp. 139-158. Ed. Stettler R F, Bradshaw H D Jr,Heilman P E, and Hinckley T M. NRC Research Press, Ottawa, Ontario, Canada.

Stott K G, Bridle P and Timberlake C F (1971) Anthocyanins in Willow Bark. LongAshton Research Station Annual Report. University of Bristol, 159-160, 171-173.

51

Zhang Z, Li F, Zhu Z and Kang X (1997) Doubling Technology of PollenChromosome of Populus tomentosa and its Hybrid. Journal of Beijing ForestryUniversity (English Ed.), 6: 2: 9-20.

52

Figure 1:The willow

breeding scheme follow

ed by the European Willow

Breeding Programm

e (Elaborated from N

ilsson (1988) andL

arsson (1997). The cycle from

initial cross to registered variety may take as long as 15 years but genotypes show

ing superiorgrow

th can be fast-tracked through the system, reaching the m

arketplace in 8-10 years.

P1 x P

2

Seedling Nursery

21

34R

egistration and Release

of New

Variety

Multi-site

Yield Trials

5-10% selected

5-10% selected

5-10% selected

1-2 varietiesselected

Year 10-15

Years 6-9

Years 2-5

Year 1

ObservationTrial 2

ObservationTrial 2

ObservationTrial 1

Multip

n Plantations

53

igure 2: Progress of the European W

illow B

reeding Programm

e from 1996-2000. A

lready over 50 varieties are in advanced observation andyield trials.

1997

Obs I

Obs II

514+6018

1998775

20002

1999

Yield

25501

3

1996N

ursery

6530

9037

11602

7593

Crosses

7812504

/18

107

256

245

9878Selection

Selection

Selection

Selection

Com

mercial

Productionand Release

54

Figure 3: Pathways used in the crossing program

me 1996-2000: intraspecific crosses (a); interspecific crosses producing F

1 hybrids (b); full-sibcrosses producing F

2 hybrids (c); first generation backcross (d); second generation backcross (e); three-way species cross (f); four-w

ayspecies cross or double cross (g). (B

ased on Figure 2, page 100, from Stettler, Zsuffa and W

u,1996).

SpeciesSpecies

SpeciesSpecies

xx

F FF

3-wa y

4-way hybrid

(double

Backcros

AB

CD

aa

aa

bb

c

d

Backcross

e

f

g

55

Figure 4:R

eciprocal Recurrent Selection for G

eneral Hybridising A

bility (RR

S-GH

A). This breeding strategy involves the phenotypic selection

of suitable parent plants from tw

o populations’ (a) followed by interspecific hybridisation (b). The F

1 families are tested to identify

progeny with high G

HA

(c) and elite lines are selected and advanced through the breeding scheme possibly leading to new

varieties(d). F

1 hybrids with high G

HA

undergo full sib mating (e) and the progeny are selected on the basis of phenotype (f). The process is

then repeated several times until there is no further genetic gain. (B

ased on Figure 1, page 99, from Stettler, Zsuffa and W

u,1996).

Species A Species B

A/

B/

F1

New

varieties

A1

B1

A1

/B

1/

A2

B2

aa

bc

d

f e

New

varieties

F1

56

Page 57

Prof. Theo VerwijstHead of Department/Chair of Short Rotation ForestryDepartment of Short Rotation ForestrySwedish University of Agricultural SciencesP.O. Box 7016SE-750 07 Uppsala, SwedenE-mail: theo.verwijst@lto.slu.se

BREEDING, PEST & DISEASE CONTROL: AN EUROPEAN VIEWPOINT

The relation between plant breeding on the one hand and pest and disease control on the other hand is particularlystrong when dealing with genera consisting of species that hybridise easily, because such genera in general attractmany different kinds of insects and pathogens. Both Salix (willow) and Populus (poplar) are genera in whichhybridisation is common.Biomass yield, for which the major selection in willow breeding is performed, is

determined by interactions between plants and an environment that consists of biotic and

abiotic factors. Abiotic conditions vary through time, and plants may to a certain degree

be well adapted to a specified range of abiotic conditions. However, pests and diseases are

biotic factors that also influence on plant performance and these may not only vary in time

and space but, in contrast to abiotic factors, also have the ability to co-evolve with plants.

Potentially pathogenic organisms may go unnoticed for longer periods in a willow stand,

until a virulent strain develops.

Susceptibility to pathogens in poplars and willows is, for certain groups of pathogens, highly clone specific. Thishas a number of implications for willow breeding because willow agronomy today largely is based on the use ofmonoclonal stands.

Willow is a boreal-arctic genus and distributed over the entire northern hemisphere, but some indigenous willowspecies can be found in Japan, India, Africa and Central and South America. Willow has been introduced inAustralasia and in New Zealand, and for many of the species, material has been transferred far beyond its naturalrange. For instance, the commercial willow energy forestry systems currently in use in Sweden are based on Salixviminalis introduced during the 16th century from continental Europe for the purpose of basket making, and onhybrids between S.viminalis and some other willow species (S. burjatica, S. schwerinii), that have been introducedfrom Siberia more recently.

Willow also is a complex taxonomical genus. Depending on the taxonomic school (lumpers or splitters) we findabout 350 to 500 willow species world-wide, which is roughly 10 times as much as in Populus. It has been knowfor a long time that ‘Once you grow poplar, diseases will come’. The same is true for willow. Populus is a genusvery closely related to Salix, and furthermore, there is relatively broad experience of breeding and pest- and diseasecontrol in poplar. Consequently, we may learn a lot from the existing poplar experience.

The world-wide transfer of willow material has consequences in the same way as we know it has for poplar.Locally, plant resistance against pathogens is much higher in hybrids because these represent a chemicalcomposition that does not seem to be attractive or manageable for the local set of pathogens that has evolvedtogether with the local species. However, in the long run, pathogens will co-evolve and may have devastatingeffects on monoclonal stands. Especially pathogens with a short generation time may develop virulent pathotypesrapidly and current trends in poplar breeding have almost given up the concept of breeding for complete diseaseresistance. Instead, breeding for tolerance - for being able to perform relatively well despite a certain level ofpathogen infestation - is aimed at in modern breeding programmes. We have seen the result of poplar breedingstrategies in Western Europe: Using a very narrow genetic base in the operational clones, many stands becamewiped-out by leaf rust varieties that managed to adapt to the poplar clones. A clone that is resistant today will besusceptible tomorrow, and consequently we may better target a broad tolerance in our willow clones, using breedingprograms that continuously supply us with new clones.

Two other important experiences have been gained by the world-wide transfer of willow and poplar material: Whilemoving European clones to the USA, it was discovered that some very well performing clones performed relativelypoorly in the USA because the presence of pathogens (insects) that do not live in Europe. Especially in poplarcultivation we also have seen introductions of pathogens to continents where these pathogens did not occur

Page 58

previously. In poplar cultivation, special rules have been designed to avoid such spread of pathogens. In willowbreeding, measures will have to be taken to diminish risk for disease spreading, and preservation of naturalpopulations will be an issue in the future.

Willow breeding therefore is not a national or continental issue, but a global issue. Willow breeding programs areongoing in Sweden, the U.K., Belgium, Croatia and Poland, but also outside Europe in New Zealand, Chile and inthe United States. World-wide exchange of germplasm is for the good or the worse of our national programs: It maygive good traits in hybrids, but is also increasing the risks of disease spreading.

If we ask willow breeders about breeding goals, we definitely get the answer that breeding is primarily towardsyield increase. Quality aspects are beginning to be considered for shrub-formed willows. In most willow breedingprograms an early screening for leaf rust resistance (Melampsora epitea) is implemented. One of the problems withleaf rust is the interaction with winter frost: No proper onset of dormancy is reached, leading to freezing duringwinter and a high mortality.

During recent years resistance, against several species of leaf beetles has come into focus. Willows are known to bethe host for many insect species but few studies have been performed to quantify the effect of defoliation by insectson stem wood production. The few figures from the literature indicate a production loss between 20% and 40%after severe defoliation. The main problem is though to understand the population dynamics of the actual insects tobe able to judge if and how contra-measures should be taken. Slightly more than 10 years ago, problemscommenced with leaf beetles in commercial plantations in Sweden. While some plantations became entirelydefoliated, a student was assigned to the problem. Next spring this student had to travel thousands of miles tocollect a handful of leaf-beetles and had to set up laboratory experiments to get material for his thesis. Five yearslater, around the time of the defence of the thesis, we again started to get extensive leaf beetle outbreaks.

In Sweden we have observed completely defoliated fields adjacent to fields of the same clonewhere hardly any insect damages could be found. There are good indications for a control of leafbeetle populations by generalist predators, and that provision of key habitats for those predatorsclose to willow stands is a key factor in the control of leaf beetles. In other words, it is time forintegrated pest management schemes as we see for many other agricultural crops.

The progress made by current willow breeding programmes has to be integrated with growing systems design(practical planting practice) and practical management of Willow-SRC.We see in Sweden that harvest frequency recommendations have to be changed due to the fact that the newlyreleased clones produce much faster than the ones released only 10 years ago. In other words, there is need for acontinuous exchange of information between willow breeders and those actors who are running the later parts of thechain.If we consider the yield capacity of a willow clone over a period of 25 years, we find that

that the top performers during good years usually have a below average performance over

many years. Consequently, characteristics such as drought resistance and water use

efficiency are now coming into focus as new breeding criteria. Another relevant question

is if we want to target the energy market only, or if we want our willow breeding

programmes to produce clones that are suitable for vegetation filter purposes or pulpwood

production. If new markets will be addressed, other breeding criteria will come into focus.

The use of more or less random multi-clonal mixtures, which indeed may delay the spreading of disease, has beendiscussed for almost a decade and promising experimental work on mixtures has been carried out in several places.Such systems seem to work at least during shorter periods. However, it is a major challenge to compose asustainable clone mixture in which the different clones do not out-compete each other during a time span coveringseveral cutting cycles.

The development of new molecular methods in breeding is likely to speed up the process of clone selection.Although those methods are unlikely to prevent pathogens from co-evolving with the new clonal material, they maygive us - in the longer run - a much higher annual release of marketable clones. They may come to provide us with abroad genetic base in operational plantations and a variety in our willow plantations that is less prone to pests anddiseases than the current willow stands.

Page 59

Yield Models for Short Rotation Coppice of Poplar and Willow:Insect and Disease Surveys 1999

Alan Armstrong, Mark Broadmeadow, Sam Evans, Paul Henshall, David Lonsdale,Robert Matthews, Nigel Straw and Ian Tubby.

Forest Research, Alice Holt Lodge, Wrecclesham, Farnham, Surrey, GU104LH

Introduction

Surveys of insects, disease and damage caused by other agents were repeated in all the SRCplots in 1999, using the standard protocol employed in previous years. The early season surveywas carried out over the period 24 May-21 July, and the end-of-season survey during1 September–19 October.

All plots at Phase I sites, except at Balbirnie, were harvested during winter 1998-1999 and werein their first season of the new rotation in 1999. Plots at Phase II sites were due to be harvestedin winter 1999-2000, and were in their third rotation year at the time of the surveys. Thedifferences in shoot age and structure between Phase I and Phase II sites were reflected in someof the results of the insect and disease surveys. Plots at Balbirnie had grown less well than atother Phase I sites and were being left for an extra year before being harvested. Consequently,survey results from Balbirnie have been included as part of the Phase II site data in the analyses.

The overall incidence of insect infestation and disease across all SRC plots are summarised inTables 1-4. The Tables give the overall average of the plot mean scores for each damagecategory, for the extensively planted poplar and willow varieties at Phase I and Phase II sitesseparately. The most important categories of damage were the same as in previous years,namely:

• leaf rust caused by Melampsora spp.• defoliation by insects• diseases characterised by spots and blotches on leaves.

In addition, leaf distortion and chlorosis were recorded frequently as symptoms of severe rustinfection and other leaf diseases, although a small amount of chlorosis was also noted inassociation with herbicide damage in a few Phase I plots. Shoot dieback was evident at 25-45%of Phase II sites, mainly following heavy rust infection in 1998, and large numbers of aphids(Tuberolachnus salignus) were found again on willow stems at some of the sites that wereinfested with willow aphid in 1998. Frost caused significant damage early in the season atBonython in the extreme south-west.

Leaf rust (Melampsora spp.)

Rust infection amongst the extensive poplar varieties ‘Beaupré’, ‘Ghoy’ and ‘Trichobel’increased at Phase II sites in 1999, compared with 1998, but decreased at Phase I sites(Figure 1). Harvesting the Phase I sites appears to have reduced subsequent infection, althoughnot substantially. Severe rust infection still occurred at some Phase I sites, notably at Trefeinon,Bore Place and Bigbrook Farm (where mean RUST-L plot scores for ‘Beaupré’ were 4.7, 4,4and 4.0, respectively). Across all sites, 85-92% of plots of the extensive poplar varietiesrecorded rust in September-October 1999 (66-84% in Sept.-Oct. 1998).

Page 60

Rust infection of the extensive willow varieties showed a similar pattern to previous years, with‘Germany’ and ‘Q83’ infected severely and ‘Jorunn’ recording little rust (Figure 2). ‘Germany’and ‘Q83’ recorded higher rates of rust infection in 1999 than in 1998, at Phase I and Phase IIsites. There was no apparent effect of harvesting on subsequent rust incidence amongst thewillow plots, and average infection rates have increased progressively over the course of thetrials (Figure 2). Average rust infection of ‘Germany’ and ‘Q83’ was higher at the Phase II sitesthan at the Phase I sites in 1998 and 1999. This difference would appear to reflect generallyhigher rates of rust infection in the areas where the Phase II plots are located, rather than aninfluence of differences in rotation age. Across all sites, 80-88% of plots of the extensivelyplanted willow varieties recorded rust in September-October 1999 (70 –81% in Sept.-Oct.1998).

Geographic distribution of rust on poplar

Average mean plot scores for rust infection (RUST-L) for the extensive poplar varieties at eachsite are illustrated in Figure 3 (9 plots/site). The relative susceptibility of the poplar varieties torust was similar to previous years (‘Beaupré’ > ‘Ghoy’ > ‘Trichobel’, Figure 1), and wasconsistent between sites, such that plot scores for the different varieties could be combined togive an overall picture of rust distribution.

The distribution of rust in 1999 was similar to that observed in 1998, but average scores weregenerally higher. Heavily infected sites were Dunnington (Yorks), Mawdesley (Lancs), CharityFarm and Harper Adams (E. Midlands), Ceredigion and Taironen (Wales), and Alice Holt,Loseley, Charlwood and Bigbrook Farm (S. England) (Figure 3). Loyton Bampton (nearBigbrook) would probably also have scored highly for poplar rust had not the ‘Beaupré’ plotsdied in the previous year, and at Bonython also, in the extreme south-west, average rust scoreswould have been higher had not the most susceptible stools been eliminated in the previous year.

Moderate rust infection was recorded at Sunnybrae (average RUST-L score = 2.0), which is thefirst time that significant poplar rust has been observed in north Scotland.

Geographic distribution of rust on willow

The distribution of rust on willows was analysed separately for ‘Germany’, ‘Jorunn’ and ‘Q83’,because of the wide differences in relative susceptibility of the varieties (Figures 4-6). ‘Jorunn’showed very low rates of rust infection (Figures 2, 5), as observed in previous years.

Leaf rust was most widespread on ‘Germany’, following the pattern seen in 1997 and 1998, andsevere infection (average score > 4.0) was recorded at sites in north Scotland, N. Ireland,Yorkshire, the east Midland, south Wales and south and south-west England (Figure 4). Many ofthese sites recorded equally high rates of rust infection on ‘Germany’ in 1998.

The distribution of rust on ‘Q83’ in 1999 was very similar to that recorded in 1998. High ratesof infection were recorded for many sites in England and Wales, but infection rates were low inN. Ireland and Scotland (Figure 6). Severe rust infection on ‘Q83’ (scores > 4.0) was recordedat Dunnington (Yorks), Delamere (Cheshire), Soham (Cambs) and at Bigbrook Farm andLoyton Bampton in the south-west.

Page 61

Rust infection on poplars at the Intensive sites

Poplar varieties at the Intensive sites showed the same patterns of relative susceptibility to rustas were observed in 1997 and 1998. Amongst the Intensive sites in southern England and S.Wales, rust was most prevalent at Alice Holt, Loyton Bampton and Trefeinon, and at these sitesthe TD/DT varieties were heavily infected, the DN varieties were moderately infected, and the Tvarieties were least infected (Table 5). The main exception to this pattern was the continuingresistance of varieties ‘690394’ (‘Hoogvorst’) and ‘690386’ (‘Hazendans’) at Trumpington andTrefeinon. The resistance of these two varieties to rust disappeared at Loyton Bampton in 1998,and at Alice Holt in 1999. Rust infection on ‘690394’ and ‘690386’ was so severe on these andthe other TD/DT varieties (except ‘Raspalje’) at Loyton Bampton in 1998 that, in combinationwith damage from frost, the stools died before assessments started in 1999.

Relatively high rates of rust infection occurred on ‘TT32’ at Alice Holt in 1999, as was observedin 1998.

Rust infection was less severe at the northern Intensive sites and differences in susceptibilitybetween varieties were less apparent (Table 5). However, the new Belgian varieties ‘690394’and ‘690386’ retained their high degree of resistance and recorded very low rates of rustinfection. The DN varieties ‘Ghoy’, ‘Gaver’ and ‘Gibecq’ also recorded very little at rust atBalbirnie (E. Scotland) and Thorpe Thewles (Yorks).

Rust infection on willows at the Intensive sites

The rank ordering of willow varieties at the southern Intensive sites in terms of susceptibility torust was very similar to that observed in 1997 and 1998. Highest rust scores were recorded for‘Germany’, the S. burjatica x viminalis varieties ‘Stott 10’ and ‘Stott 11’, and the S. triandra xviminalis varieties ‘Q83’ and ‘ST248155’ (Table 6). Pure S. viminalis x viminalis varietiesshowed moderate amounts of rust infection, whereas hybrids between S. viminalis and speciesother than S. triandra and S. burjatica recorded little rust (Table 6). ‘Spaethii’ (S. spaethii)developed a moderate amount of rust, and ‘Bebbiana’ (S. sitchensis) was almost free ofinfection.

Overall rates of rust infection were lower at the northern Intensive sites and differences betweenthe most and least affected varieties were less marked. ‘Germany’ recorded the highest rustscores at Thorpe Thewles and Loughgall. ‘Q83’, ‘ST248155’, ‘Stott 10’ and ‘Spaethii’ were alsomoderately infected at Loughgall.

Differences between pure and mixed plots

Rust infection in the pure and mixed plots showed a similar pattern to that reported in 1997 and1998. The rate of infection on varieties that were less affected by rust in pure plots tended toincrease when the variety was planted in mixture with varieties susceptible to rust.Consequently, because the amount of rust on susceptible varieties (eg ‘Beaupré’, ‘Germany’ and‘Q83’) was similar in pure and mixed plots, the overall mean rust scores for the mixed plotswere higher than for the average of the pure plots.

Page 62

Leaf diseases other than rust (‘spots and blotches’)

Symptoms of diseases that caused spot and blotch damage to leaves were more developed by thetime of the autumn survey. On average, 6.8-12.5% of the leaf area was affected by September-October 1999 (Tables 1-4). Disease was particularly noticeable at certain sites (Table 7).Infection causing more than a 20% loss of leaf area was recorded in 18% of the poplar plots and8% of the willow plots by September 1999. These proportions were lower than in 1998 (cf. 25%for poplars, 14% for willows).

Maximum rates of damage reached 39-41% leaf-area-lost amongst plots at Phase I sites and 64-83% leaf-area-lost amongst plots at Phase II willow sites (Table 7). High rates of damageoccurred on all of the extensive poplar varieties, but most of the high rates recorded for willowsoccurred on ‘Germany’. The higher maximum rates of damage recorded at Phase II sites, andgreater number of Phase II plots with high scores (Table 7), indicates a greater development ofleaf diseases other than rust in the taller, denser crops before cutting back. This is a reversal ofthe situation observed in 1998:

Percentage of plots with >20% leaf-area-lost because of diseasespots and blotches

1998 1999

Phase I Phase II Phase I Phase II

Poplars Willows

30.221.9

15.512.8

9.06.3

30.08.7

The majority of high plot scores for spot and blotch damage occurred in England and Walessouth of Yorkshire. Within this area there was no geographical pattern in disease incidence.

TAPHRINA AUREA ON POPLAR

As in previous years, T aurea infection was restricted to the poplar variety ‘Ghoy’, with smallamounts also occurring on ’Gaver’, both DN (P. deltoides x nigra) hybrids. Higher rates ofinfection were observed in the June-July survey, because infected leaves tended to fall before theautumn.

Infection rates at Phase II sites were similar to rates observed in 1998, but infection at Phase Isites was lower. All of the ‘Ghoy’ plots at Phase II sites recorded some T. aurea, and many plotsrecorded >10% leaf area affected. The highest mean plot scores for T. aurea occurred atMawdesley (10-39% leaf area affected) and Delamere (8-34% leaf area affected), both Phase IIsites. In contrast, only 10% of Phase I plots of ‘Ghoy’ were infected with T. aurea and in noneof these was more than 10% of the leaf area damaged. In 1998, Phase I and Phase II sitesshowed equivalent rates of infection. The reduction in infection at Phase I sites indicates anadverse effect of harvesting on subsequent development of T. aurea or less suitable conditionsfor development of T. aurea in the early stages of the coppice cycle.

Page 63

There were no significant differences in the rate of infection with T. aurea between pure andmixed plots of ‘Ghoy’.

Defoliation by insects

Average rates of defoliation by leaf-chewing and skeletonising insects across all plots weretypically low (8-10%, Tables 1-4), but a small number of plots were more severely affected.Between 2% and 3% of poplar and willow plots recorded defoliation rates greater than 20%,which was similar to 1998, but maximum amounts of defoliation recorded were lower than inprevious years (Tables 8 and 9). The highest rates of defoliation occurred in June-July amongstpoplars at Alice Holt and Llandovery 18, and amongst willows at Charity Farm and Balbirnie(Table 8).

The damage at Alice Holt and Llandovery 18 represent increases from 1998. At Alice Holt, only‘690386’ was heavily defoliated in 1998 (by 23-48%), whereas a much wider range of varietieswere affected in 1999 (Table 8). Defoliation at Alice Holt was caused primarily by brassy beetle(Phratora spp.). At Llandovery 18, the three plots of ‘Beaupré were defoliated by 33% in 1998,but ‘Ghoy’ and ‘Trichobel’ were much less affected. Alice Holt and Llandovery 18 are Phase IIsites and the increase in defoliation runs parallel with the increase in growth of the crop.

The highest rates of defoliation in June-July 1999, and the greatest number of plots with >20%defoliation, occurred at Phase II sites, rather than at Phase I sites (Table 8), which suggests thatpopulations of Phratora and other defoliating insects were adversely affected by the reduction inshoot material available after harvesting. At most sites, rates of defoliation were too low, andexamples of heavy defoliation too scarce, to determine whether defoliation rates were reducedsignificantly after harvesting. However, defoliation of ‘Jorunn’ at Aller Court, which reached70-80% in 1997 and 1998, during the second and third years of the crop rotation, was reduced toless than 15% in June-July 1999. Defoliation increased to 24-32% by September (Table 9).Surveys in 2000 should indicate whether defoliation at Aller Court is truly increasing again, andwhether there is a correlation between defoliation rate and crop development.

Aphids infesting stems (Tuberolachnus salignus)

Infestations of aphids on stems are recorded primarily during the September-October survey,and mainly affect willows. In most cases, the infestations can be identified as colonies of thelarge willow aphid, Tuberolachnus salignus L.. Infestations have been restricted to a smallnumber of sites each year, but at some of these sites the aphids have been extremely abundant.In 1999, 13 sites in England and Wales, and Castlearchdale in N. Ireland, recorded plot meanscores >1.0 for aphids on willow stems (Table 10). This compares with 4 sites in 1996, 2 sites in1997 and 8 sites in 1998.

Aphid infestations in 1999 were most severe at Llanrwst , Aller Court and Long Ashton. Aphidswere also present on willow stems at these sites in 1998, particularly at Llanrwst and AllerCourt. The infestation at Llanrwst led to the death of two of the ‘Germany’ plots before theassessments in 1999. The data provide strong evidence that populations of willow aphids over-wintered at these sites. The population at Aller Court does not appear to have been affected byharvesting during the intervening winter. At this site, all of the aphids occurred on the new greenstems in 1999, whereas at the Phase II sites the aphids occurred predominately on the older 2-year stem.

Page 64

No significant differences could be detected in aphid populations between pure and mixedwillow plots. Apart from the death of ‘Germany’ and some dieback on ‘Q83’ at Llanrwst, heavyinfestations of stem aphids in 1998 were not associated with shoot dieback in 1999.

Very few aphids (Pterocomma spp.) were recorded on stems of poplar in 1999, following thepattern of previous years. Moderate numbers occurred only at Loseley (max. plot scores 1.0-2.5).

Frost damage and shoot dieback

Very little frost damage was observed in 1999, and only in the June-July survey. Light damagewas recorded on the poplars and willows at Balbirnie, on poplars at Alice Holt and on willows atBonython. One plot of ‘Ghoy’ was severely damaged (plot mean score = 4.1) at Bonython.

Shoot dieback was recorded in 45% of poplar plots and 41% of willow plots at Phase II sites in1999, but in less than 5% of plots at Phase I sites which had been cut back during the winter.Serious dieback (plot mean scores >3.0) at Phase II sites occurred almost exclusively in plots ofTD and DT poplar varieties that had suffered severe rust infection in 1998, especially at AliceHolt, Ceredigion, Charity Farm, Delamere, Llandovery 16, Loseley and Mawdesley. Similarly,dieback was also evident on ‘Germany’ at Ceredigion and Mawsdesley, and on this variety and‘Q83’ at Delamere. However, dieback on willows was not as severe as on poplars.

Dieback amongst the poplars at Ceredigion might not be associated solely with previous rustinfection, however, as ‘Trichobel’ at this site also recorded moderate (2.0-3.0) shoot dieback butwas not being heavily infested with rust in the previous autumn.

Moderate shoot dieback noted previously on the willow varieties ‘Dasyclados’ and ‘V789’,which is not associated with rust, was observed again in 1999 at Loyton Bampton and Balbirnie.

Page 65

TABLE 1. Overall plot mean scores for each damage category for ‘Beaupré’, ‘Ghoy’ and

‘Trichobel’ at Phase I sites in 1999. (n=201 plots)

June-July 1999 September-October 1999

Overallmeanplot

score

Max plotscore

% plotsaffected

Overallmeanplot

score

Max plotscore

% plotsaffected

(1) Incidence scores

rust – leaves1

rust – stools2

distortion2

chlorosis2

stem lesions1

leaf hoppers1

hail damage1

terminal bud death3

aphids – leaves1

aphids – 1 year stem1

leaf galls 1shoot dieback1

stem borers1

frost damage1

spittle bugs1

aphids – 2 year stem1

aphids – shoot tip3

caterpillars in tips3

0.010.010.260.070.040.020.01

00.01

00.01

0000000

0.331.001.391.711.000.430.170.060.13

-0.240.110.11

-----

5%5%75%39%19%20%16%2%12%

-10%1% 1%

-----

1.071.060.210.160.180.060.020.010.010.01

00000000

4.783.001.563.002.351.960.830.350.130.720.110.22

------

85%82%52%51%27%21%11%7%11%2%7%1%------

(2) % leaf area lost (0 – 100%)

disease spots & blotchesleaf chewersskeletonizingblotch leaf mineslinear leaf minesTaphrina

3.5%4.8%3.5%0.1%0.1%0.1%

12.9%11.9%8.7% 1.3% 2.2%4.2%

100%100%100% 15% 18%4%

6.9%4.7%3.7%0.1%0.1%

0

40.6%9.9%18.1%2.6%1.5%1.9%

100%100%100%13%20%3%

1 incidence scored 0-52 incidence scored 0-320%3 incidence scored 0-1

Page 66

TABLE 2. Overall plot mean scores for each damage category for ‘Germany’, ‘Jorunn’ and

‘Q83’ at Phase I sites in 1999. (n=197 plots)

June-July September-October

Overallmeanplot

score

Max plotscore

% plotsaffected

Overallmeanplot

score

Max plotscore

% plotsaffected

(2) Incidence scores

rust – leaves1

rust – stools2

distortion2

chlorosis2

aphids – 1 year stem1

stem lesions1

terminal bud death3

leaf galls1

leaf hoppers1

hail damage1

shoot dieback1

aphids – leaves1

stem borers1

aphids – 2 year stem1

caterpillars in tips3

frost damage1

aphids – shoot tip3

spittle bugs1

0.090.160.250.05

00.05

00.020.010.01

00

0.0100

0.020.01

0

1.982.001.371.07

-0.500.110.310.280.460.500.200.56

--

1.211.000.17

22%21%60%28%

-28%2%15%10%8%3%8%5%--

2%3%5%

1.631.410.320.330.200.100.060.050.030.010.010.010.01

00000

5.003.002.892.874.442.000.891.091.090.440.780.170.280.110.06

---

80%76%68%58%15%30%22%14%10%7%5%15%4%1%1%---

(2) % leaf area lost (0 – 100%)

disease spots & blotchesleaf chewersskeletonizingblotch leaf mineslinear leaf minesTaphrina

2.6%4.6%3.3%0.1%0.1%

0

4.3%9.6%13.1%2.1%2.4%

-

100%100%100%12%12%

-

6.8%5.9%3.2%0.2%

00

39.0%22.6%10.0%13.9%1.8%

-

100%100%100%12%4%-

1 incidence scored 0-52 incidence scored 0-33 incidence scored 0-1

Page 67

TABLE 3. Overall plot mean scores for each damage category for ‘Beaupré’, ‘Ghoy’ and

‘Trichobel’ at Phase II sites in 1999. (n=227 plots)

June-July September-October

Overallmeanplot

score

Max plotscore

% plotsaffected

Overallmeanplot

score

Max plotscore

% plotsaffected

(3) Incidence scores

rust – leaves1

rust – stools2

distortion2

shoot dieback1

chlorosis 2leaf hoppers1

stem lesions2

hail damage1

aphids – leaves1

aphids – 2 year stem1

terminal bud death3

aphids – 1 year stem1

stem borers1

leaf galls1

aphids – shoot tip3

frost damage1

spittle bugs1

caterpillars in tips3

0.100.210.200.570.060.010.090.02

00000

0.040

0.0200

1.892.561.504.940.700.222.000.370.090.060.060.170.391.150.114.110.110.06

25%29%55%45%42%12%24%15%10%1%1%2%6%12%3%2%1%1%

2.291.770.250.220.120.120.100.070.010.010.01

0000000

5.003.002.593.691.312.131.671.260.301.000.270.500.110.06

----

92%92%60%25%44%16%22%13%11%3%4%1%4%2%----

(2) % leaf area lost (0 – 100%)

disease spots & blotchesleaf chewersskeletonizingTaphrinalinear leaf minesblotch leaf mines

4.0%4.7%5.7%3.8%0.1%0.1%

30.0%21.5%38.8%39.0%2.4%1.3%

100%100%100%35%15%16%

12.5%5.2%4.2%0.5%0.1%

0

82.5%15.0%22.9%13.6%1.9%0.4%

100%100%100%18%8%2%

1 incidence scored 0-52 incidence scored 0-33 incidence scored 0-1

Page 68

TABLE 4. Overall plot mean scores for each damage category for ‘Germany’, ‘Jorunn’ and

‘Q83’ at Phase II sites in 1999. (n=237 plots)

June-July September-October

Overallmeanplot

score

Max plotscore

% plotsaffected

Overallmeanplot

score

Max plotscore

% plotsaffected

(4) Incidence scores

rust – leaves1

rust – stools2

aphids – 2 year stem1

distortion2

chlorosis 2shoot dieback1

aphids – 1 year stem1

leaf hoppers1

stem lesions1

leaf galls1

terminal bud death3

aphids – leaves1

hail damage1

stem borers1

aphids – shoot tip3

frost damage1

spittle bugs1

caterpillars in tips3

0.400.51

00.140.120.33

00.040.100.020.040.010.010.020.010.050.01

0

4.003.000.281.191.674.390.061.422.830.610.830.390.351.330.562.000.44

-

44%47%3%60%57%41%2%15%26%15%14%10%8%8%5%6%4%-

2.121.590.460.320.170.170.170.110.110.040.030.030.020.010.01

000

5.003.005.002.812.373.225.001.572.170.830.671.060.500.781.56

---

88%86%25%62%37%28%8%20%24%12%10%14%8%10%1%---

(2) % leaf area lost (0 – 100%)

disease spots & blotchesleaf chewersskeletonizingblotch leaf mineslinear leaf minesTaphrina

3.0%5.4%3.4%0.1%0.1%

0

12.5%40.6%11.8%3.5%1.7%

-

100%100%100%11%7%-

8.1%5.9%3.3%

000

63.8%14.7%11.9%0.4%

--

100%100%100%4%--

1 incidence scored 0-52 incidence scored 0-33 incidence scored 0-1

69

TABLE 5. Susceptibility of poplar varieties to Melampsora rust, September 1999.(Mean RUST-L score: 0 = no rust, 5 = severe rust damage)

Variety Parentgroup1

S. England S. Wales N. England Scotland N. Ireland

Alice Holt

(3)2

Loyton

(1)

Trump’ton

(1)

Trefeinon

(1)

ThorpeThewles

(1)

Balbirnie

(3)

Loughgall

(1)

BeaupréBoelareUnal710091710092710151

Raspalje

6903942

6903862

GhoyGibecqGaver

TT321

Columbia

FritziTrichobel

TDTDTDDTDTDT

TD

TDTD

DNDNDN

TB

T

TT

4.613.673.704.944.624.68

4.43

2.053.26

2.992.172.02

3.47

2.86

1.231.46

******

2.72

**

3.551.840.55

1.98

1.78

1.571.30

1.030.900.860.780.770.86

0.80

0.010.01

0.670.900.52

1.93

0.94

0.930.67

4.744.783.464.374.494.20

2.92

0.110

2.873.271.62

2.71

1.53

0.930.83

0.440.590.220.880.190.72

0.16

0.020

0.100.130.05

0.65

0.55

0.230.29

0.260.210.130.090.480.13

0

00

000

0.66

0.87

0.140.59

1.972.441.862.561.701.78

1.38

00

1.531.521.43

1.42

1.66

0.801.13

1 Parent species: D = deltoides; N = nigra, T = trichocarpa; B = balsamifera

2 Rotation year. Phase I sites are in the first year after harvest during winter 1998-1999 (1); Phase II sites are inthe third year of the rotation (3)

70

TABLE 6. Susceptibility of willow varieties to leaf rust, September 1999.(Mean RUST-L score: 0 = no rust, 5 = severe rust damage)

Variety Parentspp.1

S. England S. Wales N. England Scotland N. Ireland

Alice Holt

(3)2

Loyton

(3)

Trump’ton

(1)

Trefeinon

(1)

ThorpeThewles

(1)

Balbirnie

(3)

Loughgall

(1)

Q83ST248155

Stott 10Stott 11

Germany

Spaethii

OrmJorrJorunn

UlvV789

DelamereDasyclados

BjornTora

Bebbiana

TVTV

BVBV

B

Sp

VVVVVV

VVVCa

ACiVCaCiV

VScVSc

Si

3.112.88

4.244.23

4.92

1.78

1.211.090.20

0.220.20

0.260.03

0.010.04

0.02

4.121.37

3.462.55

4.10

2.68

1.791.030.63

1.330.41

0.280.33

0.080.04

0

3.062.69

0.180.74

2.59

1.34

0.730.650.09

0.210.10

0.770.32

00

0

1.670.82

0.820.99

2.47

1.07

00.05

0

0.030.01

0.250.09

00

0

1.121.49

0.030.41

2.06

0.30

0.600.04

0

00.24

1.280.85

00

0

00

0.360.08

0.12

0.20

0.130.200.05

00

00

00

0

2.262.22

2.251.16

3.06

2.27

0.670.510.48

0.420.46

0.230.31

0.200.02

0

1 Parent species: A = Salix aurea; B = S. burjatica; Ca = S. caprea; Ci = S. cinerea; Sc = S. schwerinii; T = S.triandra; V = S. viminalis: Sp = S. spaethii; Si = S. sitchensis

2 Rotation year. Phase I sites are in the first year after harvest during winter 1998-1999 (1); Phase II sites are inthe third year of the rotation (3).

71

TABLE 7. Sites recording >20% leaf area lost because of disease spots and blotches inSeptember-October 1999.

Poplars Willows

Site Variety1% leaf

area lost Site Variety1% leaf

area lost

Phase I sites:Roves Farm

LoytonGilderbeck

“TrefeinonBore Place

Thorpe ThewlesTrefeinonGilderbeck

BeaupréRaspaljeTrichobel

m - TrichobelBoelareGhoy

Boelare710092Beaupré

4125

21 - 2523 - 24

2523232120

Phase I sites:Trefeinon

MyerscoughHayburn

Aller CourtBore Place

Myerscough“

Bore PlaceRoves Farm

GermanyGermanyGermanyGermanyGermany

Q83m - Q83Jorunn

Q83

25 - 3925 - 3721 - 3425 - 28

2624212221

Phase II sites:Llandovery 16

“Dunnington

“Mawdesley

“Carruchan

Charity Farm“

Delamere“““

GwentAlice Holt

“““““

BonythonBalbirnie

“Llangoed

Long Ashton“

Loseley“

Charlwood

Ghoym - Ghoy

m - TrichobelTrichobel

m - TrichobelTrichobel

GhoyGhoy

m - Ghoym - Trichobel

TrichobelGhoy

m - GhoyGhoy

Columbia690394

TrichobelGhoy

GibecqFritzi

Beaupré710151710092Ghoy

Beauprém - Beauprém - Trichobel

GhoyGhoy

53 - 8335

35 - 5324 - 2627 -53

3434 - 5230 – 49

3625 - 4126 - 3326 - 30

2936383331222222313027282928282226

Phase II sites:Charity Farm

“Loyton

Dunstall CourtAlice Holt

““

BonythonCharity FarmMawdesley

Gwent“

m – GermanyGermanyGermanyGermany

V789Delamere

UlvGermany

Q83GermanyGermany

Q83

70643232

24 – 322120

20 – 292524

21 - 2420

72

Wensum m - Beaupré 21

1 m- = clone in mixture plot

73

TABLE 8. Sites recording 20% or more defoliation by leaf chewing and skeletonising insectsin June-July 1999.

Poplars Willows

Site Variety1% leaf

area lost Site Variety1% leaf

area lost

Phase I sites:Loughgall

““

Friars Court

690394Columbia690386710151m-Ghoy

2221212020

Phase I sites:Friars Court

CastlearchdaleFriars Court

m - Jorunnm - Jorunn

Jorunn

22 - 282421

Phase II sites:Alice Holt

Llandovery 18“

Alice Holt“

Dunstall CourtLong Ashton

Alice Holt““““

Dunstall CourtAlice HoltLlanrwst

Alice Holt“““““

TT32Beaupré

GhoyGhoy

RaspaljeBeaupré

TrichobelBeaupré690386Boelare710092710151Ghoy

GibecqTrichobelColumbia

Fritzi PauleyUnal

690394Gaver710091

36 - 5928 - 4424 - 3623 - 3527 - 3226 -32

3225 - 31

2920 - 2822 - 27

2622 - 2520 - 2522 - 24

222222222120

Phase II sites:Charity Farm

“Balbrinie

“““““

Charity FarmMawdesley

Llanrwst

m - GermanyGermanyBebbiana

ToraStott 11Stott 10Spaethii

UlvJorunn

m - GermanyQ83

4834 - 43

3328

26 - 2823

22 - 2322232321

m- = clone in mixture plot

74

TABLE 9. Sites recording 20% or more defoliation by leaf chewing and skeletonisinginsects in September-October 1999.

Poplars Willows

Site Variety1% leaf

area lost Site Variety1% leaf

area lost

Phase I sites:Trefeinon

““““““““

GibecqTT32

710091690386710151690394Ghoy

RaspaljeColumbia

25 – 3331262524

20 - 24222120

Phase I sites:Aller Court

Castlearchdale“

Aller Court

Jorunnm – Jorunn

m – GermanyQ83

24 - 3222 – 27

2720 - 22

Phase II sites:Charity FarmLlandovery 18

Alice Holt“

GhoyGhoyTT32Ghoy

20 – 3526

20 - 2524

Phase II sites:SunnybraeBalbirnie

CeredigionBalbirnie

“Mawdesley

m – Q83Ulv

m – Q83Tora

Dasycladosm - Germany

2621 – 25

22222020

1 m- = clone in mixture plot

75

TABLE 10. Maximum plot scores for aphids (Tuberolachnus salignus) on stems of willowsin September-October 1999. Sites with plot scores <1.0 not shown.

Plot scores: 0 = no aphids, 1 = a few individual aphids on stems, 2-3 = moderateinfestation, 4-5 = severe infestations, ie most of stem covered with aphids. *indicates that aphids were present also in 1998

Willow varietySite

Germany Jorunn Q83 Other varieties

Phase I sites:* Aller Court* Trumpington Friars Court Myerscough Castlearchdale

1.3 - 1.51.8 - 2.7

1.82.3-

3.1 - 3.41.0

1.3 - 2.3-

1.0

3.4 - 4.4-

1.3 - 2.11.0-

-2.9---

Phase II sites:* Llanrwst* Long Ashton Loseley* Harper Adams Charity Farm Delamere Dunnington Alice Holt Mawdesley

4.03.0 – 4.01.1 – 2.8

-1.7

1.4 – 2.31.7 – 2.21.1 – 2.9

1.4

4.02.3 – 5.01.1 – 2.3

2.82.5

1.2 – 2.31.5 – 2.3

--

4.01.7 – 3.81.0 – 1.3

1.11.0

1.3 – 1.71.2 – 1.71.4 - 1.6

-

-------

2.2-

(- indicates plot scores < 1.0)

Aphids were present on willow stems at Dunstall Court and Demontfort in 1998, but not in1999.

76

77

78

79

80

81

82

83

84

85

86

Pests and Diseases in Commercial Scale Short Rotation Coppice

Barbara HiltonFuel Supply Manager

ARBRE ENERGY LIMITED

E-MAIL: BARBARA.HILTON@FIRST-RENEWABLES.COMIntroduction

ARBRE Energy currently has 1,100 hectares of short rotation coppice (SRC) established as part of the long-termprimary fuel source for its power plant near Selby in North Yorkshire. In order to reduce both financial and carboncosts in the transportation of the fuel, all of the coppices have been planted within a 45 mile radius of the powerplant. It is intended that between 1,500 to 2,000ha will be planted in total for the ARBRE 1 plant with plantingcompleted either in 2001 or 2002.

All of the SRC planted is willow apart from two 0.5ha areas of poplar, one planted in 1996 and the other in 1997. Upto seven varieties of willow are planted in an intimate mix at each site as recommended by Long Ashton ResearchCentre. The majority of these varieties have Salix viminalis, the Osier willow, as parental stock and all of them areon the Forestry Commission's approved list i.e. they have been through a plant breeding programme to assesssuitability for UK conditions, have acceptable resistance to or tolerance of diseases such as rusts and can producehigh yields. As it is always preferable to have a planting mix as genetically diverse as possible, a number of thenewest varieties are now from the following crosses: S. viminalis x schwerinii and S. viminalis x burjatica.

The first coppices planted were two 2.5ha trial plots on Yorkshire Water land in 1993. Commercial planting forARBRE started in 1996 although farmers' interest in the crop only began in earnest in 1998 leading to 300ha beingplanted in 1999 and 675ha in 2000. Therefore the majority of the 1,100ha of SRC are still within the early years oftheir first harvest cycle.

Various pests have occurred at a number of the coppice sites to date, the majority causing damage in theestablishment year as opposed to damaging the more mature willow. Where necessary and where feasible, treatmenthas been carried out.

Disease Occurrence

Rust

Only one site to date has been damaged by rust, Melampsora spp. In 1996 an 11ha site was planted primarily withthe new varieties of willow available at the time but also trial plots of more obscure varieties were planted plus 0.5haof poplar, Beupre and Boelare.

In 1997 the poplars suffered a severe rust infestation which in turn affected much of the adjacent willow. All leaveswere shed well before the end of the growing season. Fortunately there was no long-term damage as both the poplarsand willow fully recovered the following year. Any affect on yield was unknown.

Apart from this one site, rust has not been a problem within the ARBRE coppices probably due to the willowvarieties used and the planting regime of intimate mixing. Spraying against rust would not be carried out due to bothfinancial and environmental cost.

Pest Occurrence

Leatherjackets

On ARBRE coppice sites, spraying with Dursban is carried out with the post-planting weed control wherever SRC isplanted on ex-grassland and long term set-aside land in order to control leatherjackets, Tipula paludosa.Leatherjacket damage becomes evident when stems are found lying on the ground following attack just below thesoil surface. The automatic and early use of Dursban prevents most damage but it has been noted that if theheadlands remain un-sprayed, the leatherjackets can sometimes migrate into the edges of the crop. One applicationof Dursban proved ineffective at one late planted site this year due to the very dry conditions both at the time ofspraying and for some time afterwards. It was felt that the leatherjackets had migrated well down into the soil as it

87

became drier and, due to lack of rain, the Dursban was not washed into the soil and therefore remained ineffective atthe surface. A second application was made when leatherjacket damage was noted. Leatherjacket damage onlyoccurs during the establishment year.

Slugs

Slugs have been a problem on a very small number of wet, heavier soil sites where they can strip the leaves on youngstems within a few weeks of planting. Underground damage can also occur to the newly emerging shoots. Drazapellets have been applied to one ARBRE site where the damage was serious enough to warrant their use. Slug traps,plastic covers placed randomly across the field with pellets underneath to attract and kill the slugs, have also beenused at one other site to assess the number of slugs present and whether Draza application was necessary which iswas not. One of the most effective and benign forms of slug control is efficient ground preparation. If the soils areproperly consolidated by rolling after planting, the slugs have difficulty populating a site as they prefer to movebetween and under soil 'cobbles' on unconsolidated land.

Willow beetles

One severe infestation of willow beetle has occurred to date at one of the ARBRE coppices, unfortunately the samesite as suffered the rust damage. Beetle damage was first noticed late in the 1998-growing season when both the blue(Phyllodecta vulgatissima) and brassy (P. vitellinae) beetles were identified. Population counts of the beetles werecarried out during the spring of 1999 as recommended by The Game Conservancy Trust and when counts of 100beetles/m-2 were recorded the decision to spray, using Permasect 25, was made. Adult willow beetles tend to over-winter under the bark of mature trees and similar habitats within a few hundred metres of coppice plantations (Sage,et al. 1999). During late winter, early spring as temperatures increase, the adult beetles move into the edges of thecoppice and begin to feed before mating and then moving further into the coppice. If the population reaches criticalnumbers during this edge-feeding stage, approximately 100m-2, it should prove relatively easy to spray the edges ofthe coppice from the headlands rather than attempt to overspray the entire crop. Overspraying an entire plantationwould prove extremely costly not only financially but also ecologically as the pesticides used are not specific andwould damage many non-target and beneficial insects.

Ideally an orchard sprayer would have been used to spray the pesticide from the headlands laterally into the coppice.Unfortunately however, due to the proximity of a watercourse alongside the most badly affected area of the coppice,current guidelines prohibited the use of an orchard sprayer. A hand lance was used from a tractor-towed trailerwhich meant the pesticide was only sprayed 6m into the crop rather than the hoped for 15m. The spraying did havesome effect although the population increased again towards the end of the season but no further spraying wascarried out. Harvesting of the site was due during the winter of 1999/2000 but was delayed until spring 2000 forvarious reasons. Willow beetle numbers were exceptionally high prior to harvest although no counts were made. Asa small area of two-year-old coppice, approximately 1ha in size, was to be left unharvested it was felt that, followingharvesting of the main plantation, the willow beetles would migrate to the standing willow and this area could thenbe sprayed. Before this could be carried out however, the farmer chose to overspray the whole harvested area withpesticide.

All of the ARBRE coppice sites have been monitored for willow beetles this year and they have been found atvirtually all of the sites although not in numbers high enough to warrant any action.

Rabbits

ARBRE originally fenced each coppice site in its entirety against rabbits using British Standard fencing. In anattempt to reduce overall establishment costs and as fencing was one of the two highest costs, the other beingcuttings, it was decided from the 2000 planting season onwards that fencing would only be erected if and wherenecessary. Advice was taken from each grower as to what degree of rabbit infestation he had in the vicinity of theproposed coppice site and the location of the main rabbit harbourage. If the grower was unsure, an ex-gamekeeperwho was able to assess areas where rabbit damage might occur walked the site. Rabbits do still gain accessparticularly on the lighter soils where they are able to dig under the buried and out-turned fence. Rabbits within aplantation e.g where warrens are present in hedges within a plantation, are the responsibility of the grower and,generally, they use gassing or lamping to control numbers or, preferably, eradicate the problem.

Other pests

88

Occasionally other pest species can cause localised damage to a particular coppice site. Only one site has been deer-fenced, one of the original 1993 sites. Deer fencing is far too expensive to use so, at one other site where deer werepotentially a severe problem, sacrificial planting i.e. higher density planting, was established adjacent to thewoodland which held the deer. Nowadays, growers themselves are responsible for controlling or preventing deerfrom damaging their coppices. Hares are able to gain easy access over rabbit fencing but, to date, they have notcaused significant damage to any of the ARBRE coppices although this may not be the case in other parts of thecountry.

Rooks have pulled newly emerged shoots from the ground at one site when they moved from an adjacent field wherethey had been feeding on other crops. Geese have grazed the tops from establishing coppice where the plantationwas located next to a lake where geese were breeding. Moles are prevalent within some of the coppices but damageis rare and sporadic as they generally manage to avoid the newly planted cuttings. Once the cuttings are rooted theyare unlikely to be dislodged by mole activity. Cattle have broken through fencing to graze the coppice. Badgers canbreak through rabbit netting if it is placed across their runs and, at one site, a badger sett had been dug within thecoppice but fortunately it was noticed before a harvesting machine drove over or, more likely, into it.

Recommendations

SRC is not a high value crop. Two of its many positive aspects, particularly in promoting it to the general public, arethe low pesticide usage involved in its establishment and management and its ecological benefits. All pest controlmethods should take these points into account.

The two main pests within the ARBRE coppices have been rabbits and willow beetles although the latter has onlycaused significant damage at one site. In both cases the actual economic cost of the damage they cause is unknown.Rabbit presence is relatively easily identified although estimating population numbers takes more time and effort.From a developer and grower's point of view the following questions need answers:

• What is the cost of rabbit damage compared to the cost of fencing?• Is there a threshold population under which it is not necessary to fence?• Could electric fencing be used in places for short periods after planting and cut-back?• If rabbits graze young shoots, what proportion of the cuttings will survive?• What long-term effect on yields is caused by rabbit damage?• Are there willow varities which prove unpalatable to rabbits?

Willow beetle numbers can again be monitored relatively easily by random counts within the edges of coppices.Unfortunately, to decide whether to spray and also the appropriate time to spray, monitoring needs to be carried outat frequent intervals through late winter and spring. For a developer in ARBRE's current situation, withresponsibility for 1,100ha possibly rising to 2,000ha, this is not feasible. Individual farmers with a 10 or 20ha site oftheir own may not have the time to carry out this level of monitoring effectively. Willow beetle damage is notrestricted to one particular year within a harvest cycle, it can occur at any time so all coppices need to be monitored.The Game Conservancy Trust have recommended choosing sites with care to try and avoid those areas where beetlecolonies are likely to be found naturally (Tucker & Sage. 1999). Most farmers tend to choose their poorer land forSRC and this is frequently adjacent to watercourses and therefore prone to flooding which precludes many othercrops. Native willows are often present probably harbouring willow beetles ready to migrate into the adjacent,establishing SRC. As willow beetles are virtually ubiquitous in the ARBRE coppices, concern for the future withregard to this pest is now apparent.

As with rabbits, there are questions that need answering:

• What is the economic cost of willow beetle damage if they are left totally uncontrolled?• Would the coppice recover from a severe, uncontrolled infestation?• What is the effect on long-term yields of beetle damage?• Although counts/m-2 are simple to carry out they are time consuming, would it be possible to develop a simple

visual system for assessing numbers?• Will a benign biological control ever be developed?• Could a specific insecticide be developed that will not harm other insects?• What is the cost of planting sacrificial strips of willow around an SRC site taking into account the extra costs at

planting and, potentially, not harvesting these strips for spraying post-harvest?

References

89

Sage R, Fell D, Tucker K, Sotherton N. 1999. Post hibernation dispersal of three leaf-eating beetles (Coleoptera:Chrysomelidae) colonising cultivated willows and poplars. Agricultural and Forest Entomology 1, 61-70.

Tucker K, Sage R. 1999. Integrated Pest Management in Short Rotation Coppice for Energy - A Grower's Guide.The Game Conservancy Ltd, Fordingbridge, Hampshire, SP6 1EF.

90

COMMERCIAL SRC BIOMASS PLANTATIONS – WILL THEY STILL BE GOODFOR FARMLAND WILDLIFE?

Rufus Sage and Tracey Rich

The Game Conservancy Trust, Fordingbridge, Hampshire, SP6 1EFE-mail: rsage@gct.org.uk

ABSTRACT

There have been a several studies concerning the ecology of SRC crops in Europe over thelast 10 years or so. They tend to indicate that many pre-commercial SRC plots contain anunder-storey vegetation and insect fauna that from a conservation point of view is moreinteresting than in most other farmland crops. The crop also attracts other animals. Inparticular, the studies indicate that SRC usually contains many more bird species at higherdensities than any other arable crop.

It is crucial however that we understand how the findings of these early studies translate intothe commercial situation. The imperative to produce as much woody biomass as cheaply aspossible may still lead to considerable differences between commercial SRC plantations andthe early pre-commercial plots on which much of this work was based.

The first major biomass energy project to plant commercial SRC plantations in the UK isProject ARBRE in Yorkshire. The DTI/ETSU are funding an ecological monitoringprogramme in a representative sample of these sites that aims to assess the ecological impactof the ARBRE SRC plantations on wildlife, in what is primarily an arable area. Inconjunction with other similar studies the project should provide general guidance on theecological impact of future commercial SRC developments.

Over four years, wildlife surveys will be undertaken in 12 large SRC plots, all on previouslycultivated ground and in 12 paired arable control plots. A year-by-year comparison of theabundance and diversity of key wildlife groups in the SRC plots compared to that in thecontrols can thus be made. This should allow general conclusions about the ecological impactof SRC cropping in this region and the links between wildlife potential and plantation designand management to be drawn. Results from the first year of monitoring in 2000 aresummarized.

Introduction

Early studies concerning the ecology of SRC crops in Europe over the last 10 years or soprovide information on the potential of this new crop type to provide habitat for farmlandwildlife, especially insects, plants and birds. By measuring aspects of the environmentprovided by this new crop-type some of the work also provides guidance on maximising thispotential and on what may limit it (Coates & Say,1999; see Sage 1998 for a review).

91

For example several studies identify differences between the types of plantations in which thiswork was undertaken and commercial biomass plantations. Early monitoring programmeswere usually based in plots of SRC planted for research and demonstration purposes or forother small scale uses such as rods for basket making. Many of these contained small patchesof different age-classes and varieties and some were managed without biomass production inmind at all. It is therefore important that we understand how the findings of these earlystudies relating to wildlife use translate into the commercial situation where regular pesticideapplications in large, mainly single age class plantations, containing a handful of varieties arelikely.

Project ARBRE in Yorkshire is the first major commercial SRC development in the UK.They have now planted several hundred hectares of SRC within a 30-mile radius of theirpower station. This paper describes the design stage of a new four-year DTI/ETSU fundedecological monitoring programme in a representative sample of the ARBRE sites, that aims tobridge this gap in our knowledge of the ecolgical impact of this new crop (Anon 2000). Firsthowever, the paper summaries our existing knowledge of the potential for SRC crops toprovide habitat for new and existing wildlife on farmland, and highlights some of thedifferences that commercial plots may have on particular wildlife groups

Wildlife in pre-commercial SRC plantations

Plants

Once an SRC crop is established, because weeds are relatively difficult to control and becausethey may be less threatening to the crops potential than in, say, a cereal, there could be scopefor these crops to tolerate the presence of a ground flora. The likelihood and agronomicconsequences of this are discussed elsewhere (e.g. Sage 1999, Sage & Tucker 1998). In termsof biodiversity, a ground flora would clearly be a good thing. Certainly, many of the earlySRC plots in which ground flora surveys were undertaken contained an abundance of weedsor other plants, despite the shady environment (Gustaffson 1988; Sage 1995; Sage & Tucker1998; Coates & Say 1999). Most of these studies also identified a process of change in thecomposition of the flora as stable perennial species replace the invasive ruderal species whichcharacterise other crops and young SRC plantings.

There is therefore considerable evidence from the UK and Sweden that in many pre-commercial SRC plots an understorey vegetation has been tolerated and that over time, from aconservation point of view, it has become more interesting. A ground cover within SRCplantations would have biodiversity implications in terms of the plants themselves, insects andbirds and possibly other wildlife groups. There is relatively little information on theinteraction between weeds and SRC crops in established plantations but one study indicates areasonably high tolerance of well-established coppice to weeds (Sage 1999). Wouldcommercial plots display similar characteristics in the nature and development of the groundflora to these pre-commercial plots or would a desire to maintain weed free conditions and thecontinued invasions of typical farmland weeds prevail?

Insects

In terms of general insect biodiversity of tree species, Kennedy & Southwood (1984), whocompared insect and spider species on British trees found more on willow than any other tree

92

or genus, suggesting that invertebrate diversity in SRC crops, at least in the canopy, could behigh. Although numbers have been substantially lower in surveys of the canopy of precommercial SRC plots these studies do suggest that the diversity in native willows may inpart be reflected in SRC crops (Sage & Tucker 1997; Coates & Say 1999). Ground faunasurveys in SRC crops identify a community change in predatory insects with no increase inoverall species (Coates & Say 1999). This may mean a mixture of SRC and other cropping ina given area would lead to increase in ground invertebrate diversity through the addition ofspecies. There is also evidence that if SRC contains a ground flora it will further increase theinvertebrate biodiversity potential of this crop. A high profile insect group, butterflies willuse sheltered headlands and rides in and around SRC plantations (Sage 1996).

Like most arable crops however, SRC is also prone to damage from certain insect pest species(Sage & Tucker 1998). The use of insecticides in commercial SRC plantations would clearlydiminish the insect fauna. Herbicides too have the indirect effect of reducing herbivorous andassociated insects by removing the non-crop plants on which some of them live. It remainsunclear the extent to which insecticides will be used in commercial SRC largely because ofdifficulties in accessing standing coppice. In addition access headlands and rides, common inthe early survey plots, may not be a common feature of commercial plantations and henceprovide poor habitat for butterflies and other wildlife that use these grassy margins.

Birds

Most studies of higher-order animals in SRC crops concentrate on birds. Studies of songbirdsin pre-commerical SRC indicate that bird densities and species diversity would increaseconsiderably compared to other arable crops and improved grasslands (Kavanagh 1990;Goransson 1994; Sage & Robertson 1996; Coates & Say 1999). SRC crops share some of thehabitat features of traditional forms of coppiced woodland that are considered to be attractiveto wildlife, for example in terms of structure, edge habitat, periodic harvesting (Fuller &Henderson 1992). Populations of farmland hedgerow species would probably increase andsome new woodland and scrub species not dependant on old or dying trees would colonise.Migrant species, particularly uncommon in studies of arable farmland (e.g. Arnold, 1983),would increase dramatically if willow SRC were grown.

The issue of scale in particular is pertinent to studies of commercial plantations as many birdsuse edge habitats (e.g. boundaries between woodlands and fields) and small SRC plots are ‘alledge’. It is likely therefore that larger commercial SRC plots will contain lower densities ofbirds overall and this needs quantifying. However it is reasonable to conclude that if plantedon arable or improved grassland, commercial SRC plantations would lead to substantialincrease in the overall abundance of breeding songbirds using that land. It is also likely thatsome birds that are dependant on open farmland habitats with low cover, would not benefitfor SRC crops. Others, such as skylark and meadow pipit have however been commonlyrecorded from cut SRC (Sage & Robertson 1996). It would be valuable to investigate theextent to which cut SRC can provide alternative nesting and brood rearing habitat for thesedeclining birds. Like improved farmland, unimproved farmland habitats such as haymeadows or rough pasture also tend to contain low densities of breeding birds (Lack 1992).While SRC plantings would thus lead to increases in birds if planted on these uncommonhabitats, they usually also support important botanical and invertebrate communities and lesscommon birds and consevation. Overall, a mixture of SRC cropping and other land uses willprobably produce the greatest benefits in terms of bird diversity and abundance but

93

particularly large plantations may contain relatively few birds and may displace certainimportant species. The new study aims to address these issues.

Wildlife monitoring at project ARBRE

Background

While most of the studies outlined above suggest that new and existing farmland wildlifespecies would use SRC crops, as indicated most survey work has been undertaken in a smallnumber of relatively small experimental plantations. Table 1 provides a list of the probabledifferences between commercial SRC plantations and early SRC plots. Would edge effects inthese small plots mean that wildlife use of larger commercial plots be less? Were the smallnumber of plots used in some of the surveys representative? Would emerging problems withpests and disease lead to increasingly intensive management methods, for example morepesticides, and hence limit the crop’s value to wildlife?

Table 1.0 Likely differences between early and commercial SRC plantations (From Anon2000).

Previous Plots ARBRE Commercial Plots

Size Generally small, <5ha. Increasingly large, 10haminimum planned forProject ARBRE.

Age Structure Often mixed age classes. Typically even aged.

Planting Design Often containing clonaltrials and intimate structuraldifferences.

Typically uniform plantingsof 4-6 clones.

Agricultural Inputs Little use of fertilizers orinsecticides.

Fertilizers and insecticidesmay be used.

Land Quality Typically ex-pastoral or lowquality land.

Grade 1 or 2 arable landmay be typical for ProjectARBRE

Growth Rates Many plots consisted oftrial, relatively slowgrowing clones, 3-5 yearrotations typical.

Novel, fast growing clones,2-3 year rotations may bepossible.

Experimental design

Project ARBRE in Yorkshire is the first major commercial SRC development in the UK. It isbased on a 10 MW wood-fired power station requiring around 2000 ha of SRC. By 2000 theproject had planted several hundred hectares of SRC within a 30-mile radius of their powerstation. The project provides the opportunity to make an early assessment of the likelyecological impact of a commercial SRC operation in the UK and hence enable us to refine

94

earlier work on the details of wildlife use in SRC. This work is essential as SRC has thepotential to become a major form of land-use in certain areas. In addition, no previous studyhas properly quantified the ecological impact of SRC plantations compared to the land-use itreplaces as adjacent control, or before and after plots, have not been used (e.g. Coates & Say1999, Sage & Tucker 1998). ARBRE allows the use of control plots (otherwise similar non-SRC fields nearby) in a replicated, paired comparison.

The project therefore aims to monitor appropriate flora and fauna within and around a suitablenumber of SRC plantations and adjacent control sites and to use this information to assess theimpact on the ARBRE SRC plantations on wildlife in the area. In conjunction with othersimilar studies in the future the project should provide general guidance on the ecologicalimpact of future commercial SRC developments. There is already a new project beingconsidered that has similar aims for commercial SRC on grassland sites (ETSU Pers. Comm.).

The basic experimental design is described in the project ARBRE ecological managementproject (Anon 2000). The trial consists of 12 SRC plots, all on previously cultivated ground,six planted in 1998/99 and six 1999/2000. Plot sizes range from 5 – 27 hectare. All plotshave similar design, all are on previously cultivated ground (not grass or other) and all plotswithin 25 miles of a central point (the power station). As indicated each SRC plot has apaired control plot that is as far as possible similar to the SRC plot prior to it being plantedwith SRC. The maximum separation between SRC & control plot is 500 m.

ARBRE uses a standard planting strategy for its SRC sites so all study plots have a similardesign. Six willow varieties are used. These are Tora, used in 30 % of each plot, Jorunn 20%, Jorr 20%, Orm 15%, Ulv 10 %, Bowles hybrid 5 % (some slight between year variation inproportions exists). These are all varieties based on the willow Salix viminalis so that cropspecies diversity is low. They are planted using standard double staggered rows andharvested every three-years.

Monitoring at each of the 12 ARBRE sites will be based on repeated surveys of plants, insectsand birds in both the SRC and control plots. A year-by-year comparison of the wildlife in theSRC plots compared to that in the controls can thus be made. While there is scope to measurea variety of other wildlife groups, plants and insects underpin the ecology of these habitats(Sage & Tucker 1998). We suggest that of the higher taxonomic groups birds are by far theeasiest and most valuable group to monitor. Dedicated butterfly surveys would be undertakenin the plot headlands (Sage et. al., 1994). Monitoring of other groups such as small mammalswould require considerable additional effort and within the broad constraints of this projectcould not be undertaken adequately. Comprehensive data on weed and insect pestpopulations will be recorded incidentally.

Preliminary results

The results from the first year of monitoring provide us with some information of the effect ofvery young SRC plots on wildlife (Anon 2001). This report also describes refinements to andthe success of the various sampling and surveying methodologies that will be used during thefour-year study. There were subtle differences in the wildlife communities found in andaround the recently planted and cut SRC fields (which contained relatively little willowgrowth, particularly in early summer) and the adjacent controls. Overall, the abundance anddiversity of plants was already greater in the SRC fields, particularly the older plots planted in1999. The insect community recorded from the young SRC shoots in several plots already

95

included the occurrence of certain chrysomelid beetles, the main defoliating pests of this crop.Eighteen butterfly species were recorded in the first year of the study with a tendency formore individuals in headlands of the SRC plots than in the headlands of the arable controls.Butterflies require sheltered areas and the young SRC would have provided this later in thesummer. More bird species were recorded from the SRC fields than the adjacent controls.This included several species typical of open farmland habitats. The overall density ofsongbirds at 1.5 per ha of SRC is lower than in other studies of mature SRC. Although wewould expect this to increase as the project progresses it is possible that the effect of scale (i.e.large plots) is already having a reducing effect on bird densities.

REFERENCES

Anon (2000) ARBRE monitoring – Ecology of Short Rotation Coppice Plantations:

Agreement. ETSU B/U1/00627/00/00.

Anon (2001) ARBRE monitoring – Ecology of Short Rotation Coppice Plantations:

Interim Annual Report, 2000. ETSU B/U1/00627/00/00.

Arnold, G.W. (1983) The influence of ditch and hedgerow structure, length of hedgerow andarea of woodland and garden on bird numbers on farmland. Journal of Applied Ecology 20,731-750.

Coates, A. & Say. A. (1999) Ecological Assessment of Energy Coppice. ReportB/W5/00216/00/REP. ETSU, Harwell, Oxford.

Fuller, R.J. & Henderson, A.C.B. (1992) Distribution of breeding songbirds in BradfieldWoods, Suffolk, in relation to vegetation and coppice management. Bird Study 39, 73 - 88.

Gustafsson, L. (1988) Vegetation succession during the establishment phase of an energyforest on a peat bog in East-central Sweden. Scandinavian Journal of Forest Research 3,371-385.

Göransson, G. (1994) Bird fauna of cultivated energy shrub forests at different heights.Biomass and Bioenergy 6, 49-52.

Kavanagh, B. (1990) Bird communities of two short rotation forestry plantations on cutoverpeatland. Irish Birds 4, 169-180.

Kennedy, C.E.J. & Southwood, T.R.E. (1984) The number of species of insects associatedwith British trees: a re-analysis. Journal of Animal Ecology 53, 455-478.

Lack, P. (1992) Birds on Lowland Farms. HMSO, London.

Sage, R.B. (1995) Factors affecting wild plant communities occupying short rotation coppicecrops on farmland in the UK and in Eire. In: Proceedings of the Brighton Crop ProtectionConference 1995 - The role of weed control in land use change, 7D, 980-985.

96

Sage, R.B. & Robertson, P.A. (1996) Factors affecting songbird communities using newshort rotation coppice habitats in spring. Bird Study, 43, 201-213.

Sage, R.B. & Tucker, K. (1997) Invertebrates in the canopy of willow and poplar shortrotation coppices. Aspects of Applied Biology, 49, 105-111.

Sage, R.B. (1998) Short rotation coppice for energy - towards ecological guidelines.Biomass & Bioenergy, 15, 39-47.

Sage, R.B. & Tucker, K. (1998) The distribution of Phratora vulgatissima (Coleoptera:Chrysomelidae) on cultivated willows in Britain and Ireland. European Journal of ForestPathology, 28, 289-296.

Sage, R.B. & Tucker, K. (1998) Integrated crop management of SRC plantations to maximisecrop value, wildlife benefits and other added value opportunities. ETSUB/W2/00400/00/REP. Harwell Laboratories, Oxford. 347pp.

Sage, R.B., Fell, D.A., Tucker, K. & Sotherton, N.W. (1999) Post hibernation dispersal ofthree leaf-eating beetles (Coleoptera: Chrysomelidae) colonising cultivated willows andpoplars. Agricultural and Forest Entomology 1: 61-70.

Sage, R.B. (1999) Weed competition in willow coppice crops: The cause and extent of yieldlosses. Weed Research, 39, 399-411.