aphid suitability and its relationship to oviposition preference in predatory

Aphid Suitability and Its Relationship to Oviposition Preference in PredatoryHoverflies

Hussein Sadeghi Francis Gilbert

The Journal of Animal Ecology Vol 69 No 5 (Sep 2000) pp 771-784

Stable URL

httplinksjstororgsicisici=0021-87902820000929693A53C7713AASAIRT3E20CO3B2-B

The Journal of Animal Ecology is currently published by British Ecological Society

Your use of the JSTOR archive indicates your acceptance of JSTORs Terms and Conditions of Use available athttpwwwjstororgabouttermshtml JSTORs Terms and Conditions of Use provides in part that unless you have obtainedprior permission you may not download an entire issue of a journal or multiple copies of articles and you may use content inthe JSTOR archive only for your personal non-commercial use

Please contact the publisher regarding any further use of this work Publisher contact information may be obtained athttpwwwjstororgjournalsbritecohtml

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academicjournals and scholarly literature from around the world The Archive is supported by libraries scholarly societies publishersand foundations It is an initiative of JSTOR a not-for-profit organization with a mission to help the scholarly community takeadvantage of advances in technology For more information regarding JSTOR please contact supportjstororg

httpwwwjstororgMon Sep 10 085409 2007

Jotlnzal of Aniinnl Ecolo~~l2000

5c2000 British Ecological Society

Aphid suitability and its relationship to oviposition preference in predatory hoverflies

HUSSEIN SADEGHI and FRANCIS GILBERT School of Biologicctl Sciences Notti1igh~i72 Ui~ile~ U Ksit) Nottingh~171

Summary

1 The fitness collsequeilces of feeding on different aphids (apple blackberry dock elder nettle pea rose and sycalnore aphids) for two species of generalist insect predator (the aphidophagous larvae of Episjllplzus baltent~rs and Sj~rphtts ribesii -

DipteraSyrphidae) were illeasured in the laboratory The relevant literature studies 011prey specialization in the Syrphidae were suinlllarized for the first time 2 Both the literature evidence reviewed and the data from the experiment indi- cated that fitness compollents and overall individual fitnesses are broadly similar among all aphid species with some exceptions 3 We correlated individual fitness to two estimates of ovipositio~l preference an indirect (field distributioll of larvae) and a direct lneasure (egg distributioll under laboratory conditions) 111 the non-resident nligratory E balteat~rssignificant corre- lations were absent but there were indications of a weak preference-performance correlatioil in the largely resident S ribesii

Key-worclJ aphidophagous Syrphidae feeding specialization preference-perfor- inance

Journal o f Alzil~~alEcologj~(2000) 69 771-784

Introduction key area of studies of host-plant specializatioll in herbivores (see Via 1986 T11ompso11 1988) since it

While there is a huge literature on host-plant specia- illflue~lces how shifts onto novel hosts occur and

lization in phytophagous insects (see Schoonhoven consequently how insect species are distributed

Jeriny amp van Loon 1998) very little is kllowll of among potential hosts over evolutionary time

coinparable features of the illteractioll between Numerous studies on a variety of phytophagous

predatory insects and their prey At least some also insects have estimated the preference-performance

show moderate or narrow specializatioll (Tauber amp correlation but there is 110 consensus with results

Tauber 1987 Bristowe 1988 Hodek 1993 Milbrath rangiilg from strong to weak values Reasons put

Tauber amp Tauber 1993 Rank Snliley amp Kopf 1996 forward for this variation in outcome illclude the

Albuquerque Tauber amp Tauber 1997) Most phyto- measures of performallce used (Thompson 1988

phagous insects are specific in their choice of food Nylin amp Janz 1993 McGraw amp Caswell 1996) and

(Berenbaum 1990 Jaenike 1990 Schoonhoven et nl differences in the ability of larvae to move to alter-

1998) and even gelleralists display a hierarchy of native food sources (Price Craig amp Roininen 1995

preferences for different hosts (eg Wiklund 1981) Price 1997) Nothing is know about this in preda-

What influences prey specificity 111predators Two tory i~lsects

aspects are known to be important deteriniilallts of Larval performance depends on food quality and

specialization the foraging behaviour of ovipositing mortality due to natural enemies There is presum-

females among available food types and the subse- ably little difference between the llutritional require-

quent performance (biochemical and physiological ments of predatory insects and the content of their

specializations) of the developing larvae (eg Tauber prey and as a result the eosts of capture and the

amp Tauber 1987) The relationship between these is a toxin content of prey are probably relatively inore importa~lt to the overall costs of feeding (eg Mal- colm 1992) The potentially key role of llatural ene-

Correspondence Dr H Sadeghi School of Biological mies in inoulding oviposition responses of predators Sciences University Park Nottingham University Nottingham NG7 2RD E-mail p l x l ~ s l ~ e r n ~ e s n o t t i ~ ~ g l ~ a - llard1y (see Berdegue et lllla studied at rnacuk 1996) For herbivores feeding on plants there is a

2000 British Ecological Society Journrrl of Animnl Ecologj 69

major discrepancy betweeil what plants provide measured variables soine geileralities can be drawn (varying ainoilg plant species and in time a i d While development tiines and pupal weights are space) and what herbivores require (Schoonhoven often unaffected (eg Mathur amp Sharma 1973 et 01 1998) particularly nitrogen (White 1993) and Phoon 1973 Tawfik Azab amp Awadallah 1974ab toxin content (eg Stockoff 1991) Hainid 1985 Natskova 1985 D u amp Chen 1993)

We study here the relationship betweeil preference aphid-prey species andor prey quality does seein to and performance in two species of aphidophagous affect mortality Several aphids are probably toxic hoverflies (DipteraSyrphidae) In hoverflies like or partially toxic to the larvae of one or inore spe- many other insects ovipositiilg females select among cies of hoverfly (Milne 1971 Schinutterer 1972 food types (see Sadeghi amp Gilbert 2000a Sadeghi amp Kaufmanil 1973 Ruzicka 1975 Malcoln~ 1976 Gilbert 2000b) For a polyphagous syrphid oviposi- 1992 Einrich 1991 D u amp Cheil 1993 Rank et 01 tioilal preference by females has a profouild effect 1996) Other variables have rarely been considered on the performallce of the offspring because syrphid but can be affected by food quality (eg fecundity larvae probably have rather limited dispersal abil- Saidov 1969 ease of capture Ruzicka 1976 Wnuk ities Some indicatioils iillply larvae call inove to 1979) It is therefore unfortuilate that several studies new aphid coloilies to a certain extent Froin his measured oilly developlneilt time or pupal weight samples Banks (1968) suggested that even quite While some aphids were consisteiltly identified as sinall syrphid larvae must move between plants and uilsuitable [eg Megourn vicine aphids by Milne possibly considerable distances Kan (1988ab) sug- (1971) Ruzicka (1975) and Xiong amp Dong (1992)] gested that a single maple- or pea-aphid coloily was several aphids were deemed highly toxic in one insufficieilt to support larval developmeilt to matur- study but suitable in another [eg Brevicor7i~e bias- ity and older larvae are observed actively inigratiilg sicne by Schmutterer (1972) and Ruzicka (1975)l A ainoilg inaple branches or pea plants Most authors possible way of understandiilg these apparent con- agree however that the larvae are more or less tradictions lies in the fact that aphid toxins may be sedentary allnost completely dependent on their a t least ill part sequestered from the host plants (see mothers to choose a suitable host-prey Malcolin 1976 1992 Hodek 1993) which are vari-

A number of statements in the literature about able in toxin content inter- and intra-specifically the degree of specialization of syrphid larvae derive and in one plant individual through time (see from pure observation without ally quantitative Schoonhoven et 01 1998) There are a few explicit data In many such cases species that from their demoilstrations of the way in which variation in field distributioil are apparently specialized seein to host-plant che~llistry affects the suitability of aphids accept ally aphid offered to them in the laboratory as prey for syrphid larvae (see Kaufinann 1973 (eg Dasjsj~lph~rsspp E~tpeorles [1Megcts~~il1h~w] Malcolm 1976 Emrich 199 1 possibly Pnrng~ts loizgi- ~ i i ~ z t i ~ ) e s enti ti is in Schmutterer 1972) Pcti~ctsyil~hzts isittigel Plnryrheiitrs oi~ctis see Goeldlin 1974 Laska 1978) For example Dasy- Two species were studied here Epi s j vph~~sDnltect-sy iph~tsspecies are oilly found on tree aphids and t ~ t s(de Geer) and Sjgtiplzzts iibesii (L) chosen delib- are specialized to life 011 bark in their dark brown erately to be geileralists because an uilderstaildiilg of colouratioi~ spiny appearance and sit-and-wait the basis of variation in performance of generalist behaviour However they can be reared ill the predators will shed light upoil the ilature and evolu- laboratory on all herb aphids that have been tried tion of more specialized predatory feeding habits (Laska 1978) On the other hand there are persis- The two genera involved are not closely related tent reports of the larvae of soine species rejecting (Rotheray amp Gilbert 1989 1999) and hence there is certain aphid prey For example herb aphids are 110 cr priori reason to expect either similarities or dif- rejected by the pine specialists E ~ q ~ e o d e s ferences in their responses beyond the fact that they [Metrrsyr-ph~ts] i~ielseni ~ i i d E[M] iliteils (Laska 1978 are both labelled as extreme geileralists (see Rotheray 1988) and all aphids are rejected by syr- Rotheray amp Gilbert 1989) Field distributions show phid larvae that are specialized to ilon-aphid prey different patterns of prey utilizatioil betweeil the two such as Xc~iithnnc1iw to caterpillars of gregarious study species (H Sadeghi F Gilbert GE Rotheray Lepidoptera (Lyon 1968) or Pcrinsyiph~tsspecies to amp P Laska unpublished data) There are varying larvae of chrysomelid leaf-beetles (Rank et ctl 1996) degrees of specialization among individuals within Snpiiigogctster iiigi(t larvae in Trinidad are said to the populatio~ls of these generalist predators (Sade- thrive 011 Tl~oii~ctspisicrricr but not on TIio~1ctspis ghi amp Gilbert 1999) some individual females prefer p~~bescei1s(Guppy 19 14) particular aphids for oviposition on which their lar-

Rather few studies of syrphid larval develop~neilt vae perform better when offered as prey this specia- compare quantitatively any aspect of larval perfor- lization entails a trade-off in performailce on other mance on different aphids with the notable excep- aphid species tions of Schmutterer (1972) Ruzicka (1975) and The study reported here has two aims First we Malcolin (1976 1992) Althougl~ illrrolviilg disparate compare the illtrillsic suitability of different aphid studies with different aims and very different sets of species as food for the larvae of two species of syr- 771-784

773 H Sadeghi amp F Gilbert

2000 British Ecological Society Jouinrrl of Aizillinl Ecology 69 771-784

phid under laboratory conditions testing the ilull hypothesis that all aphids are equally good as food for these larvae Secondly we test whether there is an overall preference-performance correlatio~l in these species using both field distributio~l (H Sade- ghi et rtl unpublished data) and egg distribution among aphids in laboratory experiilleilts (Sadeghi amp Gilbert 2000a) as our prefere~lce measures

Materials and methods

L A B O R A T O R Y S T U D Y O R G A N I S M S

The two generalist hoverfly species used in this study are very colnlnon in the U K (Gilbert 1993 Stubbs amp Falk 1996) The larvae of both species have been reported from colo~lies of more than 100 species of aphids worldwide (F Gilbert unpublished data) E p i s y ~ p I ~ ~ ~ s baltent~w has a single generation (occasionally two) in the UK with a d ~ ~ l t activity in late summer During a mild winter adult fe~nales may occasionally overwinter but the bulk of the population migrates to southern Europe where pre- sumably it breeds (Rotheray 1989 Gilbert 1993) it has very little ability to tolerate cold (Hart amp Bale 1997a Hart Bale amp Fenlon 1997) S~~1y11ii1s ribesii is oligovoltine with two or three generations per year overwintering as a very cold-tolerant larva (Hart amp Bale 1997b 1998) Adult fe~nales are ilormally ready to lay eggs 7-8 days after emergence The larvae of both species are very generalized in their feeding habits but nonetheless show different field distributions on aphids (H Sadeghi et ell unpublished data) and oviposition prefereilces (Sadeghi amp Gilbert 2000a) E baltent~~swas cultured over the winter and was therefore available duriilg spring for experi- mentation S ribesii could not be cultured reliably for long and hence experiinents were done on this species during the summer

The following aphids were used in these experiments chosen for their availability and because in both field distribution (H Sadeghi ei 01 unpublished data) a i d by experiment (Sadeghi amp Gilbert 2000a) they f o r ~ n a continuun~ from favourite to avoided prey Acj~rtlzosiphon ~ ~ S L I I I (Harris) froin stock culture on broad bean (Vicicr fiber L) Aphis firbae Scopoli from dock (R~riles obrtlsifoli~rs L) Microlophi~llil ccrr~zos~ri7~ (Buckton) from nettle (Urtica dioicrr L) Mircrosiphtli~ rosne (L) from rose (Rosa sp) Aphis srrr~zbilci L from elder (S~1117btrctls nigi0 L) Aphis rllbori117 (Borner) from blackberry (R~rbtls fitlticosiw L) Dre~~ctlzos~~lz~~rlplerr~icrrloides (Shrank) fro111 sycamore (Acer pse~~ilol~lrrtai~~~s L)

L A R V A L P E R F O R M A N C E

Laboratory cultures of both syrphid species were established from gravid females captured in the field around woodland areas at the Nottingham Univer- sity campus and kept in illu~niilated cages in a con- stant e~lvironmeilt of 22 1degC with 16-11 day length Adults were fed on pollen from bee hives (Sigma Ltd Gilli~lgham UK) and on crystalliile sugar placed on Petri dish lids on the floor of the cage Water was placed 011 a soaked pad of cotton wool in a conical flask

To obtain a group of larvae of the same age females were induced to lay eggs on cut sectio~ls of broad bean plant (Vicicr ferbcr var Aq~lc~lillce cla~lclirr) infested with pea aphids For experi~nental purposes a batch of eggs laid over a period of 3 h was selected and placed in a large Petri dish to hatch The percentage mortality of the early larval illstar is nor- nlally higher than the other instars because larvae at this time are delicate and difficult to handle In addition in these experiilleilts there was a low percentage of egg hatching Together these factors decreased the llunlber of larvae available for study To iinprove survival larvae were left in groups for the first 3days and allowed to feed on pea aphids (only one or two aphids are ~lormally consu~ned during this time) At the beginniilg of the fourth day larvae were weighed individually transferred to experimental Petri dishes 14-cm in diaineter and given aphids of the selected species

Each day the larvae were weighed and the same number of similarly sized aphids were weighed and added to the Petri dishes to supply food for 24 h the remaills being weighed again the following day Each larva received only one aphid species As the larvae grew in size the number of aphids offered each day increased to keep pace with their requirements thus aphids were always in excess but always offered in equal numbers among aphid species Aphids are probably easier to catch in this situation than in nature and hence the suitability we inea- sured did not incorporate all of the costs of capture Depending on availability 10-16 replicates were used except in the case of blackberry aphid for which only six larvae were reared because of the scarcity of this aphid in the field

The resulting pupae were weighed and the potential fecundity of the resulting adult females was measured by dissecting them and counting the ovariole ~luillbers of both ovaries In some instances the pupae could not be weighed because their strong attachment to the Petri dish meant they could not be re~noved without injury this led to variable sample sizes in the data Development time was mea-sured to the nearest day Given the labour involved

Cavaiielln sp from hogweed (Heic~clelriiz s~~hol~rlj - ill the experiment it was not feasible to measure it li~ri71 L) and Aphis 110171i DeGeer from apple (Mcrltls more accurately and in any case the exact time of donzestica Borkh) pupation in syrphiiles is difficult to determine since

774 Prefe~m~ce-perfori~znrlce i i ~ predatory h o ~erjies

2000 B r ~ t ~ s h Ecological Society

the process is gradual with 110 fixed points such as the emergence of pupal spiracles through the puparial cuticle (as in eristaline syrphids)

Performance is a fitness measure that should include survival growth rate feeding efficiency pupal inass and fecundity as well as loilgevity of the resultant adult 111 practice however offspring performance has usually been estimated froin only one or two compoilents of performance more than 50 of studies have taken only one or two parts of the insects life-cycle into account It is a problem to decide which compoileilt of larval perforlnance to use in correlation analysis because different aspects of offspring perfor~nailce call produce different results when co~npared to female preference (Thompson 1988) Furthermore as noted by McGraw amp Caswell (1996) different compoileilts of performailce respond differeiltly to changes in the eavironment For such reasons Jai~z Nyliil amp Wedell (1994) recommend that oviposition preference should be correlated with total offspring fitness and not only with particular components Therefore individual fitness ( r ) was calculated as a perfor-mance measure (McGraw amp Caswell 1996) by inte- grating development tiine (D) survival (111 = 1 or 0) and potential fecuildity (V) via the equation r = [Ln (112v] D where Ln = natural logarithms

Fitness measured in the laboratory is of course not the same as fitness in the field particularly because mortality due to ilatural ellenlies is omitted lneasusiilg fitness of individuals in the field would be al~nost impossible The objective of this study was to measure the iiltri~lsic suitability of aphids For all these reasons the i~ldividual fitness of McGraw amp Caswell (1996) was selected as the measure of choice Fitnesses were calculated for surviving females only and for all females iilcludiilg non-survivors Because female larvaepupae callilot be differ- entiated from males we assumed half the lnortality to be female rouildillg up wheil an odd number had died This procedure also led to variable sample sizes ill the data The sex ratio of einergillg adults did not differ sigllificailtly from unity (flt 17 NS)

We used Analysis of Variance for the analysis of differences in performa~lce amoilg aphids for all ineasures except mortality where we used and individual fitiless including non-susvivors where we used a non-parametric ANOVA (ie a Kruskal-Wallis test) Since there was a o (I priori prediction of these differences we report only the overall variance ratio following Day amp Quill11 (1989) we refrain from using multiple coinparisoils

O V I P O S I T I O N P R E F E R E N C E

Ovipositioll preferences alnoilg hosts was iavesti-gated in two ways illdirectly from the field distribu-

tion of larvae (H Sadeghi et 01 unpublished data) and directly from egg distributions under laboratory coilditions (Sadeghi amp Gilbert 2000a)

Briefly the field distribution of larvae was assessed by a regular sampling programme From early May until late October 1996 a weekly sampling programme of aphid coloilies and associated larvae of certain syrphid predators was carried out The site of field sampling was the Nottingham University campus a large area of about 90 ha mainly of grass- lands and a few areas of woodland Plant names used here are from Mabberley (1997) and Clapham Tutiil amp Warburg (1968) Aphid colonies (see Table 4) on plant leavesstems were removed to the laboratory and the type and number of each syrphid species recorded and the aphid identified using Rotheray (1989) and Blackmail amp Eastop (1994) The focus of this part of the study was on only the two species considered here In taking a sample from every five aphid-infested leavesstems the fifth was picked and put into a plastic bag As a sainpling unit 50 aphid-infested leaves were used for sycamore beech elm apple and blackberry for other plants (willow-herb nettle hogweed dock rose and elder) ten 20-cin aphid-infested steins were used The size of sampled aphid colonies was recorded on a 3-point scale (1 = lt 5 2 = 5-40 3 = gt 40) To give a rough estimate of the overall relative avail- abilities of the different aphids the colony sizes of all samples were summed for each aphid species

Ovipositioll preferences were determined in the laboratory using females of known age Females were initially naive having had no previous expo- sure to aphids Aphids were then offered to each individual female on a newly cut section of their host-plant standing in water great care was taken to provide equal aphid densities by eilsuring that all cut sections were about the same size (a 20-cm shoot with three small leaves) and infested with the same number of aphids (of various instars) Each day aphids were presented in a randomized sequence to each syrphid female (ie a no-choice situation with only one aphid species available at any time) Each preselltatioil of an aphid species lasted for 30min The number of eggs laid in each case was counted and the aphid replaced by another aphid species contilluiilg until all aphids had been presented Each day all test aphid species were offered presentations colltinued until all oviposition had finished

Results

P E R F O R M A N C E O F E P I S I R P H U S

B A L T E A T U S

The results of ineasuring the components of performance of E Onltenttrs fed on various aphids are glvell ill Table 1 The larvae of E Onlteutus call

775 H Srrdeghi amp F Gilbert

c2000 British Ecological Society JOLIIII(IofAniina1 Ecologj)69 771-784

776 Preference-perfo1~orr72i~~ce111

predatory lzoi~erflies

c2000 British Ecological Society Jourrzrrl of Arzi17ziil Ecologj) 69

develop to maturity successfully on all the aphid species offered The survival of larvae during development was low for those feeding on apple aphid otherwise apart from 011pea aphids (75) survival rates were close to 10O00 The larval developineiltal period ranged from 6 (elder blackberry dock and nettle aphids) to 14 (apple aphid) days There were sigilificailt differences among the seven aphid species in respect of larval developinent time Larvae fed on apple aphid had the greatest average developinent time again clearly different from all others

The weights of pupae ranged from 225mg on elder aphid to 46mg on dock aphid Mean pupal weights from larvae reared on different aphids differed significantly For the resulting adults meail head widths and wing lengths varied significantly ainong the experiinental groups for both sexes The number of ovarioles of adult females were not significantly different among experiineiltal groups but ranged froin 46 (elder aphid) to 88 (dock aphid)

Individual fitnesses of E bnlteat~rs developiilg on different aphids are given in Table 1 Except in the case of apple aphids there were few obvious differences aillong mean fitnesses for the various aphid- prey treatments

P E R F O R M A N C E O F SI RPHCrS R I B E S I I

Sjyphz~s r~besii larvae were fed six aphid species col- lected from various host plants apple aphid and elder aphid could not be used due to their scarcity in the field at the time of the experiment and were replaced by sycainore aphid The perforinailce measures of 5 ribesii reared on the different aphid species are given in Table 2

There were no significant differences in survival aiuong treatment groups apart froin dock aphid (75) the survival rate on all aphids was close to 100 The larval development period ranged from 9 (rose nettle pea and sycainore aphids) to 11 (dock blackberry nettle and pea aphids) days The inean developine~lt tiines were significantly different among groups slightly longer on blackberry and dock aphids than the very similar but more rapid times for the other four aphid species

Pupal weights ranged from 324mg (dock aphid) to 708ing (sycamore aphid) and were significantly different among aphid-prey groups Larvae fed on rose and sycamore aphids on average resulted in heavier pupae whereas the pupae of larvae reared 011blackberry and dock were lighter In the resulti~lg adults there were significant differences among treatment groups in head width and wing length for both sexes The number of ovarioles of adult feinales ranged from 61 (dock aphid) to 120 (sycamore aphid) but the meail values were not significantly different

Individual fitnesses of survivors (Table 2) differed significantly ainong aphid-prey treatments with larvae fed on sycamore and rose aphids having greater fitness The increased variance associated with the inclusio~l of non-survivors (ie individual females with zero fitness) led to no significant differences ainong the meail fitnesses of all females although a non-parametric ANOVA remailled significant (Krus- kal-Wallis H = 131 P lt 005)

C O M P A R I S O N B E T W E E N S Y R P H I D S

We tested the individual fitness measures (omitting non-survivors) using a two-way Anova to look for differences between the two syrphid species and for differences in response to the various aphids (the interaction term) Fitness of 5 ribesii was consis-tently higher than E bnltentzrs on all aphids

=5025 P lt lt 0001) and there were differ-ences overall among aphids (F434= 511) The significant interaction =392 P=001) showed different responses of the syrphid species ainong aphids E balteritus had inore or less equivalent fitness on all aphids (we could not provide apple aphids to S ribesii) whereas the fitness of S ribesii was more variable and was particularly high on rose aphids

P R E F E R E N C E - P E R F O R M A N C E

C O R R E L A T I O N

The assumption of the field sampling programme was that the relative frequency of larvae on particular aphid colonies would represent the relative preference of that prey by ovipositiilg females The results showed a consistent pattern of preference among aphid species detailed in Table 3 Relative frequencies and the absolute numbers of larvae were highly correlated (r =093 n = 11 P lt 001) so we concentrated 011 testing for correlatio~ls with the absolute numbers The results of the oviposition preference experiments also showed that feinales sig- ilificailtly prefer some aphids over others again the overall preference hierarchy is given in Table 3 It is often expected that differences in oviposition preference measured in the laboratory will be reflected in the distribution of larvae among aphid colonies in the field However testing for a positive (rank) correlation between larval frequency among different aphid colonies and oviposition preference by females varied with syrphid species There may have been agreement between the hierarchies of oviposition preference and larval distribution among different aphid colo~lies for E baltentus (r =054 n =7 P =009 wholly determined by the low oviposition preference for Crivnrielln aphids-without this datum r =084 I = 6 P lt 005) There was no 771-784

Tahle2 The influence o l the species of different aphids on various perlormance measures in the development of Syrplzus ril~c~rii(L) Each entry in the table gives the mean SEM range and sample size (except for survival) Aphids are in order of the fitness achieved by larvae feeding upon them (lasl row) Differences among means are assessed by one-way ANOVA except for individual fitness including non-survivors where a non-parametric Ailova test was used (H distributed as survival diiTereilces are tested using X Overall fitness for each individual was calculated according to McGraw amp Caswell (1996 see Methods) both including and excluding non-survivors

Aphid Measure Sycamore Pea Blackberry Dock Nettle Rosc Tesl statistic

Development time

(days)

Survival () (ol larvae)

Survival () (to adult emergence)

Pupal weight (mg)$

Ovariole number

Fitness r (per day x 100) (of survivors)

Fitness r (per day x 100) (including non-survivors)

--Minimum-maximum sample size) fDue to the attachment of some pupae to the Petri dishes not all individuals were weighed

--- - --

Table3 Hierarchy of aphid species in relation to field distribution (H Sadeghi et oiunpublished data) oviposition preference (fronl Sadeghi amp Gilbert 2000a) and suitability for larval development

Aphid species

El~i~yiplittsbcrlteirtlrs Acjrrhosipiion pi~to7i Aphisfirboe Apliis grosszrlniine Aplzis poi7ii Aphis itrboiiiiii Aphis snr7ib~rci Cal~rrieilrrsp Diepariosipli~trli piniitnrioiili~ Mocio~ij~litrr~iiosne

Aphis grossirlniinc Aphis poi7ii Apliis iiihorui7i Aplzis sanibitci Cnl~iriieilirsp

2000 British Ecological Society Jouri~ni o f Aiiiii~ni Ecoloel 69-

~~ -

Plant species

Field distribution

Relative Number of frequency larvae

Ovipositiol~ preference$

Suitability for larval development

All$ Survivors~

Soiiibtretrs 17igio Herncieiini ~pl ioi~~ j~l i i i~i i Acer pse~rdol~inrifori~rs Rosn sp Uitica riioica

TVicio frrbcr Rzriiies obt~isifolitts El~ilobi~rniliirs~it~rni Adolits doiiiestico Rtrbzrs jiuricosus Srrriib~tctis iiigro Heiiicietir71 splioridjli~ciii Acer pseitdopioiitrriiiis Rosa sp CTrt icn rlioicir Firgtis syil~ntico ~ 1 7 7 1 i Y Sp

Number of larvae divided by the index of aphid relative abundance (H Sadeghi er oi unpublished data) Absolute ~lunlher of larvae found during standard sampling $Mean percentage of total egg load laid on an aphid during no-choice secluential oviposition tests $Mean individual fitness of all individuals tested calculated according to McGraw amp Caswell (1996) IMean individual fitness of those surviving to adulthood calculated according to McGraw amp Caswell (1996)

suggestion of any relationship in S iibesii (r =037 17 =7 NS)

The suitability of different aphid species as larval food showed sinall but significant differences for both syrphid species summarized in Table3 There was no discernible positive relatioilship between either measure of individual fitness and either measure of the frequency of larvae in field for E bnltea-tiis (all rank coefficients were negative) Of the aphids studied blackberry and nettle aphid were the least preferred by adult females in field with only one or two larvae found in all the systeinatic samples Both aphids were suitable as larval prey however with nettle aphid resulting in the highest measured overall fitness In the case of S iibesii testing for a positive (rank) correlation between iildividual fitness and the larval ilumbers in the coloilies of different aphids was significant (r=087 n =5 P lt 005) showing that even for these small numbers there was agreement between performance and frequency of larvae in the field

Testing for a positive (rank) correlation between oviposition preference and the suitability of aphids measured by individual fitness showed a similar pic- ture There was no suggestioil of any positive relationship between oviposition preferences by gravid females and either nleasure of fitness in E balteatus (both coefficients were negative) In S ribesii there were indications of positive relationships (in both cases r =060 11 =6 P= 009) although the power of the test was weak because of the small number of points

The data lead us to suggest that in S ribesii there is a link among the hierarchy of ovipositioil preferences larval frequencies among aphid colonies in the field and subsequeilt offspring fitness In E bal-teatus however the data provide no evidence of such links We do not think this is an artefact of small sample sizes (caused by the labour-intensive nature of the experiments) in part because the sample sizes of E balteatus were always larger than those of S i~besii 771-784

779 H Sndeghi amp F Gilbert

2000 British Ecological Society Jouvnul ojAniinc11 Ecologj 69

Discussion

Malcolm (1992) divided predators of a particular prey species into three categories iilcllrded (those that are unaffected by prey defences and hence can exploit all prey types successfully including the one in question) peripheral (those that suffer significant declines in fitness when reared on the prey species in question) and excluded (those that cannot exploit the prey species because it kills them) Increases (or decreases) in allocation by the prey species to defensive toxins expand (or contract) enemy-free space at the expense (or to the benefit) of peripheral predators only leaving included and excluded predators unaffected The peripheral predators therefore con- stitute a potent evolutionary force on prey defences A community approach is therefore important each aphid species is likely to be attacked by a variety of aphidophagous predators and many aphidophagous predators attack a variety of aphid species as prey Our results show that even generalist aphidophagous syrphids are peripheral predators on some common aphid species

The vast majority of studies (summarized in Table 1 and including our own) use generalist species The overwhelming pattern arising from the results is that inany aphids are rather similar in their suitability as food for syrphid larvae but often with one or two exceptions (eg apple aphid in our work) The rank order of aphid suitability as food differed for different components of larval performance (cf Tables 2 and 3) This underlines the need to integrate components into a single conlposite measure of fitness rather than relying on single components we think that the individual fitness approach of McGraw amp Caswell (1996) provides a better more complete performance measure for ecological studies A further conclusion is that in the laboratory larvae are often able to develop successfully on aphids rarely used in the field For example in the experiments with E baltent~rs nettle aphids were highly suitable as food for larvae whereas field sampling (H Sadeghi et nl unpublished data) and oviposition preference experiillents (Sadeghi amp Gil-bert 2000a) showed that this aphid is rarely selected for oviposition by gravid females In the context of the hierarchy-threshold model of Courtney Chen amp Gardner (1989) nettle aphids are low in the rank hierarchy of acceptability to gravid females why might this be the case

A variety of factors might influence the evolved rank hierarchy of suitability the host-plant as a habitat for larvae the intrillsic suitability of the aphid as food may vary with host-plant (Hodek 1993) body size norillal colony size and density the sequestration or production of toxins defensive behaviour recruitment of ants season (particularly in declining colonies Kan 1988ab Hodek 1993) or various other forms of low profitability the iillpact

of natural enemies (which may vary with aphid species andor host-plant) or the nature and frequency of coillpetition aillong larvae within aphid colonies Thus in our particular case survival of E baltecrt~ls larvae on the nettle host plant in nature may be low because of the effects of the plant itself on the larvae which must move about on its surface Nettle aphids are known to be especially adept a t avoiding capture and hence have high costs of capture not illeasured in the laboratory in Petri dishes or the inortality due to natural enemies might be unusually high when using this aphid and females have been selected to avoid using it Most of these factors remain alnlost unstudied in their effects on syrphid larvae

One well-studied aspect of prey selection in syrphids is that they prefer to lay eggs in larger (Chandler 1968ab H Sadeghi et 01 unpublished data) or developing aphid colonies (Kan 1988ab) and they do not like declining colonies This is probably related to the larval food requirement and may be one reason why blackberry aphids (which often occur at very low densities) are low in the oviposition preference hierarchy This behaviour probably ensures that newly emerged larvae have enough food to develop successf~~llyand is iillportant because periods of food deprivation during the larval stage can result in dwarfed adults (Ruzicka amp Gonzales Cairo 1976) with lowered fecundity or even sterility (Cornelius amp Barlow 1980) Just why aphids from colonies in decline are intrinsically less suitable is hard to say The longer development time and decrease in survival of larvae of E bnlteuttls on apple aphids could be due to this effect since at the time of the experiment the colonies of this aphid were going into decline For some replicates a similar explanation was advanced for the unsuitability of Aphis fkbne when fed to Ezlpeocles (Metasjrpl~us) corollcre larvae (Ruzicka 1975)

Aphids show a wide range of defences against their enemies speed and alertness (eg Dixon 1958 Niku 1976 Brodsky amp Barlow 1986) hiding in galls or waxy secretions (eg Mitchell amp Mak-symov 1977 Evenl~uis 1978) recruitillent of ants that are effective deterrents (Banks 1962 Dixon 1998 p 228) active defence soinetiilles by soldier castes (Ohara 1985b Foster 1990 Dixon 1998 p 102) and toxins (Wink amp Roilier 1986 Nishida amp Fukailli 1989 Malcolin 1990) While the impact of aphid defence has been reasonably well studied in the aphidophagous coccinellids (see Majerus 1994) there is a dearth of such work on aphidophagous syrphids One Japanese coinillunity of aphidophagous syrphid species could be divided into three groups generalists specialists on ant-tended aphids and specialists on aphids that aggressively defend themselves (Mizuno et al 1997) There are particular larval n~orphological (eg in Pnrcrgus hnei~~oirholrs Mizuno et a 1997) 771-784

780 Preje~ence-peifoii~criice iiz p ~ e d a t o i j ~ hoveiflzes

lc 2000 British Ecological Society Journal ojArziii1c11 Ecologj~69

or female ovipositional strategies (eg E~peodes (Metasyiphus) coilficrtei Ohara 1985a) in the syrphids that mitigate the effects of these defences Branquart (1999) came to a siinilar conclusion in studying European aphidophagous syrphids 90 are oligophagous or stenophagous and they mostly feed on well defended aphid prey Adapta- tions of syrphids to aphid defences and particu-larly the presence of ants are an iinportant and unstudied component of syrphid evolution Because of the design of our (and most other) experiments the measured defensive components contributing to suitability consisted only of some aspects of capture efficiency together with the ability of the larvae to digest aphid tissue However there were differences in suitability among aphids in these components and these were related to oviposition preferences in Syrphrts ~ibesii

The role of aphid toxins in defence against their predators has been studied in only a few cases The oleander aphid Aphis neiii (Malcoln~ 1976 1990 1992) appears to synthesize its toxins rather than deriving them from its very poisonous host-plants but there is 110 doubting their effectiveness against its peripheral predators In studies of syrphids differences in quality among aphids are not consistent among studies even using the same syrphid species suggesting that variable host-plant chemistry might be the source of this tri-trophic interaction For example our successful rearings E bnlteatzrs or S iibesii larvae on both elder and dock aphids contrast with those of Ruzicka (1975) vlho found neither aphid to be suitable food for Elrpeodes (Metnsj~r- p h ~ ~ s )coiollae larvae If there was any detrimental effect of these two aphids on test larvae in this study a possible explanation could be due to start- ing the experiment with 3-day-old larvae according to Schmutterer (1972) the resistance of syrphid larvae to toxic aphids increases with larval age although his scanty data do not provide a great deal of support for this statement Most accounts suggest that when larval mortality is large it occurs within the first 4days after eclosion froin the egg (Schmut- terer 1972 Kaufmann 1973 Malcolm 1992) Differ- ential early mortality on different aphids needs further investigation

The size of prey is often considered an important characteristic in determining whether or not it can be handled by predators (Strand amp Obrycki 1996) Kan (1988ab) noted that aphid size is critical for the newly eclosed first-instar larva and part of the reason for ovipositing in young colonies may be to make available small and tender aphids for the first few meals of the first instar The higher illortality and longer development time for E bnltentzu on apple aphid (Aphis poi71i) may be related to the smaller size of this aphid which imposes extra capture costs on older larvae Hodek (1993) reported that apple aphids (Aphis poini) were also less suita-

ble than other aphids as food for coccinellids and caused a decrease in the weight of larvae of Exoclzo- II(S qz4nclT~~ll~tul~1t11s(L)

In this study the main unmeasured components of fitness were differences in capture efficiency among aphids on their host plants illortality due to natural enemies and the risk and effects of larval competition at aphid colonies The proportion of larvae killed by specialist parasitoids can be very high (Rotheray 1989 Gilbert 1993) but no analyses of relative risks at different aphid colo-nies in the field have been carried out such field ineasurements of susceptibility need doing Compe- tition between syrphid larvae can greatly influence the bionomics of the species (Benestad-Higvar 1972 1973) Field sampling (Mizuno et al 1997 H Sadeghi et nl unpublished data) generally shows a wide overlap in the ranges of prey aphid species between the studied syrphid species but the extent to which heterospecific larvae co-occur in the same colonies has rarely if ever been reported in syrphid larval communities and hence the potential for inter-specific competition remains unknown In the population dynamics of a com-munity of adult hoverflies there are signs of competitive effects only among hoverfly species that are generalists as aphidophagous larvae (Gilbert amp Owen 1990 Gilbert 1990) Oviposition strategies might be expected to have adapted to regular and frequent competitive interactions and therefore the females of generalist species might avoid ovi-positing in aphid colonies where eggs or larvae are present Some observations have supported this prediction (Banks 1953 Chandler 1968a Rotheray amp Dobson 1987) However despite some claims based purely on observational data (Hemp- tinne et 01 1993 Hemptinne Doucet amp Gaspar 1994) careful experiments (Chandler 1968b Phoon 1973 Bargen Saudl~ofamp Poehling 1998) showed no signs at all of any such avoidance of competition

Despite the relatively small differences in fitness resulting fro111 feeding on different aphid species and small sample sizes there was evidence for a preference-performance correlation in one of our generalist syrphid species S ribesii but not in the other E bnlteatus As in some insect herbivores (Via 1986 Ng 1988) the existence of individual variation among females may weaken or abolish a preference- perforillance relationship measured at the population level About 60 of individual female E bal-teatus showed few oviposition preferences whilst the remainder differed strongly from one another (Sadeghi 81 Gilbert 1999) associated with trade-offs in larval performance Thus some individuals appear to be adapted to different prey-use strategies with preference for one aphid species entailing a trade-off in performance on another This further underlines the fact that the relationship between 771-784


02000 British Ecological Society Journal of A n i ~ z a Ecology 69

insect predators and their prey is very similar to that between insect herbivores and their host plants (Tauber amp Tauber 1987 Bristowe 1988 Thompson 1988 Albuquerque et nl 1997)

Assuming that the difference between the two syrphid species is real a possible explanation lies in the overwintering strategies of the two species E bal-teatirs is the only temperate representative of a tropi- cal genus and inigration from the South in spring and back again in autumn is the dominating characteristic of its population ecology Thus E balterr-tus has not evolved primarily to the conditions of Northern Europe and therefore possibly not to its aphids In stark contrast S ribesii is the commonest representative of an overwheln~ingly Northern Holarctic genus strongly adapted to overwintering residency there and hence to its aphids This might also explain why the mean fitness of S ribesii is con- sistently higher than that of E bcrlteat~lsand also its greater variation in fitness among aphids One might reasonably predict greater local adaptation in the resident S ribesii and greater genetic uniforn~ity in the migrant E bcrlteatzls

In order to study evolutionary changes in diet breadth life-history components need to be considered in a phylogenetic context an estinlate of the phylogeny now exists for the Syrphidae (Rotheray amp Gilbert 1989 1999) with which to assess such coinponents associated with changes in diet breadth Aside from preliminary phylogeny-free studies (Gil- bert 1990 Gilbert et a 1994) there are few published studies on predators with a phylogenetic focus One exception is a pair of sister-species of aphidophagous Chrysopidae (see Tauber amp Tauber 1987 Albuquerque et 01 1997) where one species is a generalist and the other a specialist derived from it that feeds on woolly alder aphids Evolving to specialize on these aphids entailed the evolutioil of larger eggs and hatchlings larger adults reduced fecundity slower development and a reduced ability to feed on the prey of the generalist Some of these same traits may also be associated with specialization in syrphids (Gilbert 1990) a conclusion supported by the detailed phylogenetic studies of Branquart (1999)

Acknowledgements

This work was supported by a studentship to the first author awarded by the Governinent of Iran We thank Graham Rotheray and several anon-ymous referees for their coininents on the manu-script and Andy Hart Jeff Bale Grahain Holloway and Richard Harrington for help with rearing aphids and syrphids

References

Albuquerque GS Tauber MJ amp Tauber CA (1997) Life-history adaptations and reproductive costs associated with specialization in predacious insects Jotrrrinl of Ar~imnl Ecologj 66 307-3 17

Banks CJ (1953) Stzrdies or1 rile piedoror coniplex ofcei- tail7 Aphirlirlr~e ( H e m ) 1tlit11 partic~ilni refererice to Aphis fabae Stop Uapublished PhD Thesis Univer- sity of Loildoi~

Banks CJ (1962) Effects of the ant Lnsizrs iiiger (L) on illsects preying on small populations of Apl~is fobcre Scop on bean plants Arli~als of Ap~~lier l Bioiogj 50 669-679

Banks CJ (1968) Effects of insect predators on small populatioils of Al11lis frrboe in the field Enfoniologiir Esl~ei~irierlriilise f Al~plicntn 11 169-1 76

Bargen H Saudhof K amp Poehling H-M (1998) Prey fillding by larvae and adult fenlales of E l ~ i ~ r p ~ u shnl-r~ar~rs Erifornologia E~perir~ieiiralis er Applicata 87 245-254

Benestad-Higvar E (1972) The effect of intra- and interspecific larval coinpetitioil for food (14yzus pe i s i ca~) on the development at 20degC of Sjrpii~rsribesii and Sy i -phlrs co~~ol lae (Diptera Syrphidae) Eizrorilol~l~ngn17 71-77

Benestad-H$gvar E (1973) Food con~petition in larvae of S j i l~h~ i srihesii (L) and Sj~rplitts coi~ollrre (Fabr) (Dipt Syrphidae) 1Voislc Er~toiliologislc Tirlsslii~ift 20 3 1 5- 321

Berdegue M (1996) Is it enemy free space The evidence for terrestrial insects and freshwater arthropods Ecolo-gicrri Ei i to i l l~~ogj 21 203-217

Berenbaum MR (1990) Evolution of specialization in insect-umbellifer associations Arzr~~ialReiien of Ellto- r71010gj35 3 19-343

Blackman RL amp Estop VF (1994) Aphicl~0 1 1 the Worlrls Trees or1 Iderit$cntiori a i d Irfoiitintiori Guide CAB International (NHM) London

Branquart E (1999) L ~ e - h i ~ t o i j ~ stinfegies of l~overflies ~~ i r i~~rerlireio~r~Ior~oe (Di]1rei~rlSjrplii11r1e)PhD Thesis Facult6 Ulliversitaire des Sciences Agronollliques de Gembloux Belgium

Bristowe C M (1988) What makes a predator specialize Trerlds in Ecology amp Evol~itior-r3 1-2

Brodsky LM amp Barlow CA (1986) Escape responses of the pea aphid Acjrrio~ipiori ] I ~ Y L I ~ I I (Harris) (Ho1nopteraAphididae) il~flueilce of predator type and temperature Cnr~orliiril Jouii~ol of Zoologj 64 937-939

Chandler AEF (1968a) The relationship between aphid infestations and oviposition by aphidophagous Syrphi- dae (Diptera) Aiinrrl~ of Applied Biology 61 425434

Chandler AEF (1968b) Sollle factors influencing the occureilce and site of oviposition by aphidophagous Syrphidae (Diptera) Ailizals of Applierl Biology 61 435-446

clap hall^ AR Tutin TG amp Warburg EF (1968) Excursioii F101rr of the Biitislz I~les Canlbridge Univer- sity Press Cambridge

Cornelius M amp Barlow CA (1980) Effect of aphid con- sumptioil by larvae on development and reproductive efficiency of a flowelfly S y y h u s corollrre (DipteraSyr-phidae) Carmilini~ Er7tori701ogi~r 112 989-992

Courtney SP Chen GK amp Gardner A (1989) A general inode1 for individual host selection Oilcos 55 55-65

Day RW amp Quinn GP (1989) Coinparisoils of treatments after an ailalysis of variance in ecology Ecologi-cnl Mor~ogriiplis 59 433-463 771-784

782 Pieference-peifo~n~cincein

predatory zoveifiies

2000 British Ecological Society Journnl of Aizin~al Ecology 69 771-784

Dixon AFG (1958) Escape responses shown by certain aphids to the presence of the coccinellid beetle Aili~lin ilecen~p~c~ictieto of 28(L) Joziinnl Aiiin~nl Ecology 259-28 1

Dixon AFG (1998) Aphid Ecologj 2nd edn Chapinan amp Hall London

Du Y-Z amp Chen X-Z (1993) I~lfluence of different aphid prey on the developnlent of Metnsyrpl~us corolliie (Dip Syrphidae) [in Chinese with English sumn~ary] Cliii~ese Jozci11nl of Biologicnl Contiol 9 1 1 1-1 13

Emrich BH (1991) Acquired toxicity of the lupin aphid 14ncrosiphzci~1 nlbifrons and its influence on the aphidophagous predators Coccinella septenpzinctntri Epi- ~)~rpi~ir ciirrieii [in Gernlan biilteatzis and CIii)~so~~erln with English summary] Zeitsclir~$t fiii Pj7iinzei1kiiiil-Icl~eiteri ze~id Pjn~izer~schzetz 98 398-404

Evenhuis NL (1978) Diden internze~liii (Diptera Syrphi- dae) as a predator of S e I ~ i ~ o l a ~ I ~ ~ ~ z e spineti (Hemiptera Aphididae) and prey specialisation in other aphidophagous syrphid larvae [in Dutch] E~~tonzologisclieBericli-ten An~stei~ilni~z 38 129-1 3 1

Foster WA (1990) Experimental evidence for effective and altruistic colony defence against natural predators by soldiers of the gall-forming aphid Pei~~l~liigzaSII~I -othecne (HemipteraPemphigidae) Belini~iouinl Ecologjl amp Sociobiology 27 421430

Gilbert F (1990) Size life history phylogeny and feeding specialisation in insect predators Insect Life Cycles Genetics Eiol~etion rind Coor~ilinotion(ed F Gilbert) pp 101-124 Springer Verlag London

Gilbert F (1993) Hoveijies Naturalists Handbooks 5 2nd edn Richmond Press Surrey

Gilbert F amp Owen J (1990) Size shape competition and community structure in hoverflies (DipteraSyrphidae) Jouiriai of Anin~iti Ecology 59 21-39

Gilbert F Rotheray GE Zafar R amp En~erson P (1994) The evolution of feeding strategies Plzyloge-rieticr nizd Ecologj (eds P Eggleton amp R Vane-Wright) pp 324-343 Acadenlic Press London

Goeldlin P (1974) Contributio~l to the systeinatics and ecology of the Syrphidae (Diptera) of western Switzer- land [in French] Mitteilzrrzgerz tier Sclz~~~eizeriscl~er~ Africa

field Bzelletii~ de la Section Regionale de IOuert-Pniaenrctiqzie Orgnnirilfion Internietionnle de Lutte Bio- logique 17 101-1 11

Hodek I (1993) Habitat and food specificity in aphidophagous predators Biocontiol Science amp Teclznologj~ 3 91-100

Jaenike J (1990) Host specialization in phytophagous insects Annual Reviei~ of Ecologj~ amp Systematics 21 243-273

Janz N Nylin S amp Wedell N (1994) Host plant utilization in the Comma butterfly source of variation and evolutionary implications Oecologiil 99 132-140

Kan E (1988a) Assessment of aphid colonies by hoverflies 1 Maple aphids and Epirj~iplus baiteatus (de Geer) (DipteraSyrphidae) Journnl of Ethologjl 6 39-48

Kan E (1988b) Assessment of aphid colonies by hoverflies 11 Pea aphids and 3 syrphid species Bernsj~iplzus seiniizis (Wiedemann) Metnsyrplius jiequens Matsu-lnura and Slil~usvitril~enrzir(Meigen) DipteraSyrphi- dae) Journal of Etkologj~ 6 135-142

Kaufinann T (1973) Biology of Pnrngus boiboniz~r (Dip-tera Syrphidae) as predator of Toxoptein nuraiztii (Homoptera aphididae) attacking cocoa in Ghana Ameiicnn M i i h n d Naturalirt 90 252-256

Laska P (1978) Current knowledge of the feeding specialisation of different species of aphidophagous larvae of Syrphidae [in French] Ai~rznle~ Ecoiogierle Zoologie -Ai~iinicle10 (3) 395-397

Lyon J-P (1968) Contribution to the biological study of Xaiztlzond~us conlptus Harris [in French] Aniznle~ Des Epil~hytie~19 683-693

Mabberley DJ (1997) The Plnnt Book 2nd edn Cam- bridge University Press Cambridge

Majerus MEN (1994) Ladybirds Collins New Natural- ists London

Malcolm SB (1976) Aiz investigotiorz ofpiant-derived mr -

iliac glj~cosider (is n possible bitsis for ieposernatisnz irz the aphidophagozis I z o v e r ~ ~ ~ Ischiodon aegyptius (Wie-deniilnn) (Di l~ fe in Sj~rphidne) Unpublished MSc Thesis Rhodes University Grahamstown South

E1ztor1oiogiscIzei7 Geselischaft 47 151-252 Guppy PL (1914) Breeding and colonising the syrphid

Circulilr of tile Department of Agric~eltzire Tiininild amp Tobago 10 217-226

Hamid S (1985) Aphid control potential of El~iryrl~lius billtentus (DeGeer) Pioceerling~ of tlie 5111 Polcismil Congress of Zoologj 153-1 58

Hart AJ amp Bale JS (1997a) Cold tolerance of the aphid predator Epis~~rp11urbalteatzis (DeGeer) (Diptera Syr- phidae) Physiologicni E~ltoiioogjl 22 332-338

Hart AJ amp Bale JS (199713) Evidence for the first strongly freeze tolerant insect found in the UK Ecolo-gicnl Entoizoiogj~ 22 242-245

Hart AJ amp Bale JS (1998) Factors affecting the freeze tolerance of the hoverfly Syiplzzis ribesii (DipteraSyr-phidae) Jozirnnl of Insect PIiyriology 44 2 1-29

Hart AJ Bale JS amp Fenlon JS (1997) Developillental threshold day-degree requirements and voltinism of the aphid predator E l ~ i r y r ~ ~ ~ u ~ (Dipterabielteictus Syrphidae) A~zi~nls o f Applied Biologj~ 130 427-437

Hemptinne J-L Dixon AFG Doucet J-L amp Peter-sen J-E (1993) Optinla1 foraging by hoverflies (Dip- tera Syrphidae) and ladybirds (Coleoptera Coccinellidae) mechanisms Ezii~opean Jozirnal of Ento- ~ ~ ~ o l o g y 90 451-455

Hemptinne J-L Doucet J-L amp Gaspar C (1994) How do ladybirds and syrphids respond to aphids in the

Malcolm SB (1990) Chemical defence in chewing and sucking insect herbivores plant-derived cardenolides in the monarch butterfly and oleander aphid Cl~ernoecol-ogy 1 12-21

Malcolm SB (1992) Prey defence and predator foraging Nnturnl Eneinies the Population Biology of Pierlators Pnritsites niid Di~eoses (ed M J Crawley) pp 458- 475 Blackwell Scientific Oxford

Mathur T amp Sharma J (1973) Studies on the predaceous behaviour and bionomics of Xnrztl~ograi~~rnasclcteilaie Fabr and Splzneropl~orinsp (Diptera Syrphidae) Uni-veisitj~ of Uclnipur Resenre11 Journni 11 52-53

McGraw JB amp Caswell H (1996) Estimation of individual fitness from life-history data Anzericnrz h~ntumlist 147 47-64

Milbrath LR Tauber MJ amp Tauber CA (1993) Prey specificity in Cliryrol~nan interspecific comparison of larval feeding and defensive behavior Ecology 74 13841393

Milne W (1971) Fnctois affecting nplzid populntioiz~ orz broad beans Unpublished PhD Thesis Imperial Col- lege University of London

Mitchell RG amp Maksymov JK (1977) Observations of predation on spruce gall aphids within the gall Ento-n~o~~lzogn22 179-1 86

Mizuno M Itioka T Tatematsu Y amp Ito Y (1997) Food utilization of aphidophagous hoverfly larvae (Diptera Syrphidae Chamaemyiidae) on herbaceous


02000 British Ecological Society Journal of Aniilznl Ecoiogy 69 771-784

plants in urban habitat Ecologicitl Rr~ei~ici an 12 239-248

Natskova V (1985) Influence of basic ecologic factors on the feeding capacity of some aphidophages in the larvae period [in Bulgarian with Russian and English summaries] Elrologij~ii15 35-42

Ng D (1988) A novel level of interactions in plant-insect systems hntitre 334 611-613

Niku B (1976) Some consequences of the drop reaction of Acj~i~tzo~ipzoizpisiti-il for the larvae of Sjipilits corocte [in German with English sun~mary] Ei i to i~~ol~ l~ngi~ 257-263

21

Nishida R amp Fukami H (1989) Host-plant iridoid-based chemical defence of an aphid A ~ ~ ~ i ~ t I ~ o ~ i ~ ~ h o i ~ 25 91-100iiil11101ii-

Ruzicka 2 (1976) Prey selection by larvae of Metnsyiphit~ coi~ollcie(Diptera Syrphidae) Actci Ei~toniologio Bolie- r-i~osloi~ctccl73 305-3 1 1

Ruzicka 2 amp Gonzales Cairo V (1976) The effect of larval starvation on the develop~nent of M E ~ ~ I S J ~ I ~ I L I Scor-ollae (Diptera) Ve~ti~ilcCesltosloi~eiislte Sl~oleciiosti Zoologiclte 40 206-21 3

Sadeghi H amp Gilbert F (1999) Individual variation in oviposition preference and its interaction with larval perfornlance in an insect predator Oecologict (Brrlinj 118 405-41 1

Sadeghi H amp Gilbert F (2000a) Oviposition preferences of aphidophagous hoverflies Ecoiogiccil Ei~tor-i~ologj~

cus against ladybird beetles Jo~tri~ctlof Ci~er-izicctl Ecol-ogj 15 1837-1845

Nylin S amp Janz N (1993) Oviposition preference and larval perforlnance in Polj~gor~ictc-cilbitr-i~Ecologicnl Etito- Wlologjl 18 394-398

Ohara K (1985a) Observations on the oviposition behaviour of Metctsj~ipl~uscorfintei (Diptera Syrphidae) and the defensive behaviour of soldiers of Pseitdoiegi~lct bnr~ibucicoia(Homoptera Pemphigidae) Esctltict 1985 99-106

Ohara K (1985b) Observations on the prey-predator relationship between bci~~~b~tcicolciPse~tcloregi~~n (Homo-ptera Pemphigidae) M e r i ~ ~ r l ~ h i t ~and coifiittei (Diptera Syrphidae) with special reference to the behaviour of the aphid soldiers Esciliiii 1985 107-110

Phoon CG (1973) Biology nriti ecologj~ of sor-iie n~~zi(io- 11hitgoits S~~iplzitiite Dipteict) Unpublished MSc Thesis University of Malaya Kuala LUITIPLI~ Malaysia

Price PW (1997) hlsect Ecologj~ 3rd edn Wiley New York

Price PW Craig TP amp Roininen H (1995) Working towards theory on galling sawfly population dynamics Popitlntion Dyizni~~ics Ne11 Appionc11es oriel Sjli~tl~esis (eds N Cappuchino amp P W Price) pp 321-338 Aca- demic Press New York

Rank N Smiley J amp Kopf A (1996) Natural enemies and host plant relationships for chrysomeline leaf beetles feeding on Salicaceae Czi~yioiiiri(oeBiologj~r ~ 2 Ecoiogicol Studies (eds P H A Jolivet amp M L Cox) pp 147-1 7 1 SPB Acadenlic Publishing Amsterdam

Rotheray GE (1988) Third stage larvae of six species of

Sadeghi H amp Gilbert F (2000b) The effect of egg load and host deprivation on oviposition behaviour in aphidophagous hoverflies Ecologiciil E~lto~-iioogj 25 101-108

Saidov AKh (1969) The effect of larval and inlaginal feeding on fertility in syrphids [in Russian] Noiiclirij~e Ti~tci~Tnshlceiitsii Gositdctrstl~ertnj~i Uiiiveisitet III [I Leiiirlit 1969 120-124

Schmutterer H (1972) On the food specificity of polyphagous predatory syrphids of East Africa [in German with English summary] Zeitschrfr Jyii Aiigel~~i~iinte Eiitoil~ologie71 278-286

Schoonhoven LM Jermy T amp van Loon JJA (1998) It~sect-Plonr Biologj~ Chapman amp Hall London

Stockoff BA (1991) Starvation resistance of gypsy moth Ljil~nr~tiin elispoi L (Lepidoptera Lymantridae) trade-offs among growth body size and survival Oeco-logic iBeilir~) 88 422-429

Strand MR amp Obrycki JJ (1996) Host specificity of insect parasitoids and predators Bioscieiice 46 422-429

Stubbs AE amp Falk SJ (1996) Biitii11 Hoi~eiPies it11 Nlit-itrnted Iceitt$cntioi~ Guide 2nd edn British Entomolo- gical amp Natural History Society London

Tauber CA amp Tauber MJ (1987) Food specificity in predacious insects a comparative ecophysiological and genetic study Ei~ohttior~cli~~Ecoiogj~1 175-186

Tawfik MFS Azab AK amp Awadallah KT (1974a) Studies on the life-history and description of the imma- ture stages of the Egyptian aphidophagous syrphids 111 Xnrltliogrni~ii~li~ Wied B~tlletin cle la Soci-negjll~tiiii~i ire Eiitoi~iologiq~tecle iEglpte 58 73-83

Tawfik MFS Azab AK amp Awadallah KT (197413) Studies on the life-history and description of the imma- ture stages of the Egyptian aphidophagous syrphids

Gazette 39 153-159 Rotheray GE (1989) Aphid Pietlittor~ Cambridge Natur-

alists Handbooks 7 Cambridge University Press Cambridge

Rotheray GE amp Dobson J (1987) Aphidophagy and the larval and pupal stages of the syrphid Platj~cl~eii~itsjitl-viieiztiis (Macquart) Enton~ologists Gctzette 38 245-251

Rotheray GE amp Gilbert F (1989) The phylogeny and systenlatics of European predacious Syrphidae (Dip- tera) based on larval and puparial stages Zoologicitl Jo~triinl of the Lini~eoii Societjl 95 29-70

Rotheray GE amp Gilbert F (1999) Phylogeny of Palaearctic Syrphidae (Diptera) evidence from larval stages Zooiogicnl Jourr~ol of tlie Liii~~ectr~ Societj 127 1-1 12

Ruzicka 2 (1975) The effects of various aphids as larval prey on the development of 14etnsjlipIi~c~ coiollne (Dipt Syrphidae) Er~roi-i~ol~l~i~gi~20 393402

aphidophagous Syrphidae (Diptera) E~ito~-i iologi~t~ IV Spzcteiopl~oi~inflci~~icc~~tdciZett Bitlletiri cle 10 SociPtP E~itor-iiologique cie IEgjyte 58 103-1 15

Thompson J N (1988) Evolutionary ecology of the relationship between oviposition preference and perfor-mance of offspring in phytophagous insects Eiironiologiii Ev~~er~inierititli~ 47 3-14et A l ~ ~ ~ l i c a t i ~

Via S (1986) Genetic covariance between oviposition preference and larval performa~lce in an insect herbivore Ei~ol~itioii40 778-785

White TCR (1993) Tlie Iiiii(ieqitote Eiiviioiii~erit Nitro- geii ctr~n tile Ah~tr~riitiice of Ai~iil~nlsSpringer-Verlag Berlin

Wiklund C (1981) Generalist vs specialist oviposition behaviour in Pcipilio n~clcl~ctor~ (Lepidoptera) and func- tional aspects on the hierarchy of oviposition preferences Oikos 36 163-170

Wink M amp Romer P (1986) Acquired toxicity-the advantages of specialising on alkaloid-rich lupins to Mnci~osipliiri~ictll~ifi~oiis(Aphididae) Nc~tu~~n~isser~scIzcf-teii 73 210-212

784 Wnuk A (1979) Episjlrpl7us bnltentu~ (DeGeer 1776) Ine (Diptera Syrphidae) [in Chinese with English sum-

Preference- (Diptera Syrphidae) as an aphid (Homoptera aphi- mary] Clzirzese Jourrznl of Biological Control 8 6-8

pe~forinancein predatory

doidea) predator [in Polish with Russian and English summaries] Alcadenzin Roliiiczo 1111 Hugonir Ko11nfnjn W K r n k o ~ ~ ~ i e Rozprnrvn Hnbilitacyjr~n 72 1-64

1zoveifiies Xiong H amp Dong H (1992) Experiments on rearing and Received 24 hrovernber 1999 revision received 9 February greenhouse release of the larvae of Metos~~rpizuscoi01- 2000

02000 British Ecological Society Jouirinl of Ariirznl Ecologj~69 771-784

Jotlnzal of Aniinnl Ecolo~~l2000

5c2000 British Ecological Society

Aphid suitability and its relationship to oviposition preference in predatory hoverflies

HUSSEIN SADEGHI and FRANCIS GILBERT School of Biologicctl Sciences Notti1igh~i72 Ui~ile~ U Ksit) Nottingh~171

Summary

1 The fitness collsequeilces of feeding on different aphids (apple blackberry dock elder nettle pea rose and sycalnore aphids) for two species of generalist insect predator (the aphidophagous larvae of Episjllplzus baltent~rs and Sj~rphtts ribesii -

DipteraSyrphidae) were illeasured in the laboratory The relevant literature studies 011prey specialization in the Syrphidae were suinlllarized for the first time 2 Both the literature evidence reviewed and the data from the experiment indi- cated that fitness compollents and overall individual fitnesses are broadly similar among all aphid species with some exceptions 3 We correlated individual fitness to two estimates of ovipositio~l preference an indirect (field distributioll of larvae) and a direct lneasure (egg distributioll under laboratory conditions) 111 the non-resident nligratory E balteat~rssignificant corre- lations were absent but there were indications of a weak preference-performance correlatioil in the largely resident S ribesii

Key-worclJ aphidophagous Syrphidae feeding specialization preference-perfor- inance

Journal o f Alzil~~alEcologj~(2000) 69 771-784

Introduction key area of studies of host-plant specializatioll in herbivores (see Via 1986 T11ompso11 1988) since it

While there is a huge literature on host-plant specia- illflue~lces how shifts onto novel hosts occur and

lization in phytophagous insects (see Schoonhoven consequently how insect species are distributed

Jeriny amp van Loon 1998) very little is kllowll of among potential hosts over evolutionary time

coinparable features of the illteractioll between Numerous studies on a variety of phytophagous

predatory insects and their prey At least some also insects have estimated the preference-performance

show moderate or narrow specializatioll (Tauber amp correlation but there is 110 consensus with results

Tauber 1987 Bristowe 1988 Hodek 1993 Milbrath rangiilg from strong to weak values Reasons put

Tauber amp Tauber 1993 Rank Snliley amp Kopf 1996 forward for this variation in outcome illclude the

Albuquerque Tauber amp Tauber 1997) Most phyto- measures of performallce used (Thompson 1988

phagous insects are specific in their choice of food Nylin amp Janz 1993 McGraw amp Caswell 1996) and

(Berenbaum 1990 Jaenike 1990 Schoonhoven et nl differences in the ability of larvae to move to alter-

1998) and even gelleralists display a hierarchy of native food sources (Price Craig amp Roininen 1995

preferences for different hosts (eg Wiklund 1981) Price 1997) Nothing is know about this in preda-

What influences prey specificity 111predators Two tory i~lsects

aspects are known to be important deteriniilallts of Larval performance depends on food quality and

specialization the foraging behaviour of ovipositing mortality due to natural enemies There is presum-

females among available food types and the subse- ably little difference between the llutritional require-

quent performance (biochemical and physiological ments of predatory insects and the content of their

specializations) of the developing larvae (eg Tauber prey and as a result the eosts of capture and the

amp Tauber 1987) The relationship between these is a toxin content of prey are probably relatively inore importa~lt to the overall costs of feeding (eg Mal- colm 1992) The potentially key role of llatural ene-

Correspondence Dr H Sadeghi School of Biological mies in inoulding oviposition responses of predators Sciences University Park Nottingham University Nottingham NG7 2RD E-mail p l x l ~ s l ~ e r n ~ e s n o t t i ~ ~ g l ~ a - llard1y (see Berdegue et lllla studied at rnacuk 1996) For herbivores feeding on plants there is a






























Results


B A L T E A T U S























Development time

(days)



Pupal weight (mg)$

Ovariole number




--- - --


Aphid species




~~ -

Plant species

Field distribution




All$ Survivors~










Discussion




















Acknowledgements


References






































































21








































































Results


B A L T E A T U S























Development time

(days)



Pupal weight (mg)$

Ovariole number




--- - --


Aphid species




~~ -

Plant species

Field distribution




All$ Survivors~










Discussion




















Acknowledgements


References






































































21
































































Development time

(days)



Pupal weight (mg)$

Ovariole number




--- - --


Aphid species




~~ -

Plant species

Field distribution




All$ Survivors~










Discussion




















Acknowledgements


References






































































21











































--- - --


Aphid species




~~ -

Plant species

Field distribution




All$ Survivors~










Discussion




















Acknowledgements


References






































































21













































Discussion




















Acknowledgements


References






































































21
























































Acknowledgements


References






































































21





























































































21











































aphid suitability and its relationship to oviposition preference in predatory

Documents