Self-organized asymmetries in ant foraging: a functional response to food type and colony needs

The dominant paradigm to explain asymmetries in the spatialdistribution of foraging animals is that they track the spatialheterogeneity of their environment. However, in social insects,endogenous spatial asymmetries can emerge within a uniformenvironment as an outcome from the self-organizing processof trail recruitment. We studied how self-organized asymmetries contribute to the exploitation of different food sources (carbohydrateor proteins) in colonies of the aphid-tending ant Lasius nigervarying in their nutritional needs (presence or absence ofbrood). Colonies with brood fed on sucrose sources exhibita higher mobilization of foragers than the other experimentalgroups. Foraging patterns differ greatly according to food type: colonies strongly focus their activity on only one dropletof sucrose, whereas they show a rather homogeneous distributionof foragers between proteinaceous sources. In addition, thepresence of brood in the colony enhances the asymmetry of collectiveforaging for both types of food. These spatial differencesin self-organized foraging patterns allow efficient exploitationof natural resources and play a role in the competitive strategy of this widespread palearctic ant.     

On foraging,recruitment systems and optimum number of scouts in eusocial colonies

Summary A numerical model of an eusocial colony foraging for food showed that, for each set of values of resource density, resource size and recruitment system employed, a given optimal proportion of scouts in the colony maximize the amount of resources retrieved by a colony during a fixed period. The model predicts that ants using mass recruitment systems should have larger colonies with small foragers, and should forage on large food sources. Retrieval of small food sources by small colonies is best achieved with large workers using individual foraging strategies. For mass foragers, several food sources are best retrieved using democratic decision-making systems in recruitment, whereas for very large food sources at very low mean food patch density, autocratic decision-making systems are optimal. Some of the experimental evidence available is discussed in the light of these findings, as they confirm the prediction that large colonies with small workers have mass recruitment systems, whereas workers of small colonies with large workers are generally lone foragers.     

Cladoceran filtration rate-body length relations: model improvements developed for a Microcystis-dominated hypertrophic reservoir

Filtering rates were measured for zooplankton species in Situon single-celled Chlorella and on four Microcystis colony sizefractions (5–20, 20–40, 40–60 and 60–100µm) in a hypertrophic reservoir. Natural-log-transformedfiltration rates of five cladoceran species, one copepod andone rotifer were included in an all-food-particle, all-speciesmultiple regression model which explained 43% of the variancein filtration rate as a function of animal body length. An additional14% and 7.6% of the variance was attributable to food type andzooplankton species respectively, with temperature accountingfor <4% of the variance. Restricting the filtration ratemodel to cladocerans alone explained 51% of the variance asa function of animal length, 16% as a function of food type,7.5% as a function of species and only 0.2% as a function oftemperature. In linear filtration rate models for each foodtype, cladoceran body length explained 70% of the variance whenfeeding on Chlorella and between 57 and 67% of the varianceon the four Microcystis colony fractions. Models describingcladoceran filtration rates on Chlorella and the 5–20µm Microcystis colony fraction were significantly differentfrom the three models on larger colonies due to cladoceran responsesto increasing food particle size. Accordingly, a combined modelfor Microcystis colonies >20 µm was developed. Inclusionof food quality factors such as cyanophyte colony size seemsjustified in models aimed at estimating clearance rates, resourceutilization and phytoplankton grazing losses in plankton orecosystem studies when applied to eutrophic or hypertrophiclakes where large cyanophyte particles are abundant.     

Omnivory in cyclopoid copepods: comparisons of algae and invertebrates as food for three, differenfly sized species

We examined the importance of algal versus invertebrate preyfor three cyclopoid copepods by comparing reproductive successfor females fed on three diets: exclusively algal food, exclusivelyinvertebrate prey and a combination of algal plus invertebrateprey. The three cyclopoid species represent a distinct gradientin body size: Tropocyclops prasinus mexicanus (small body size),Diacyclops thomasi (medium body size) and Mesocyclops edax (largebody size). Our results provide strong evidence for the potentialimportance of algae in the diet of adult cyclopoid copepods.However, the role of herbivory is dependent on species bodysize. We found that algae are the key dietary factor for reproductivesuccess of the small T.p.mexicanus. In contrast, the availabilityof invertebrate prey had a greater influence on reproductionby the two larger species. Overall, cyclopoid copepods shouldbe generally considered as having the potential to derive theirenergy from either animal or algal sources. This has importantconsequences for interpretations of the forces influencing zooplanktoncommunity diversity as well as for general food web theory.     

Fungal Assciations: II. Cultural Studies on Rhizoctonia solani Kuhn, Fusariutn solani Snyder and Hansen, and Other Fungi, and Their Interactions

Several types of interaction between Rhizoctonia solani, Fusariumsolani, and Phoma foveata were found when these fungi were grownon potato-dextrose agar. After being used by Rhizoctonia a potatomash medium gave better growth of Rizoctonia and Fusarium thanit did when the medium was initially used by Fusarium; and thiswas so whether the reaction of the spent medium was readjustedor not. It is suggested that potato mash medium used by Fusariumcontains a thermostable factor(s) affecting the subsequent growthof Rhizoctonia or Fusarium. The range of pH values suitable for Rhizoctonia growth was narrowerthan that for Fusarium, optimum values being approximately 5•9for the former and 7•8 for the latter. In mixed culturesof the two fungi on potato-dextrose agar adjusted to differentpH values, the fungus for which the reaction of the medium wasmore suitable usually became visually predominant after sometime. A study of various carbon sources showed that poor growth ofRhizoctonia was obtained when pectin was used as the sole sourceof carbon. On a pectin-agar medium, the rate of growth of aRhizoctonia colony increased on the sector which lay towardsan adjacent Fusarium colony; also, after the two fungi camein contact, there was more rapid growth of Rhizoctonia roundthe Fusarium colony than elsewhere. On a synthetic liquid mediumwith pectin as the carbon source better Rhizoctonia growth wasobtained when Fusarium-spent medium was added to it than whenRMzoctoma-spent medium was added. Rhizoctonia showed partial deficiencies in thiamine, biotin,and inositol. Both the extract of Fusarium mycelium, grown onvitamin-free medium, and the Fusarium-spent medium, stimulatedthe growth of Rhizoctonia on vitamin-free medium.     

Body Wall Morphology of the Sertellid Cheilostome Bryozoan,Reteporellina evelinae

Autozoids in R. evelinae are budded as in most cheilostomes,but some lateral zoid buds systematically become empty kenozoidsthat wrap around onto the basal side of the colony. Each kenozoidcontributes partly to the formation of a younger autozoid. Budsuppression and kenozoid formation lead to an erect, branching,stony colony. Kenozoids calcify heavily and may bud other kenozoidsor autozoids in old colony parts; these probably are adaptationsto protect the otherwise delicate basal sides of autozoids. Large frontal avicularia are produced from two areolae (pores)in frontal walls of autozoids. Interior walls delimiting theavicularium contact the frontal cuticle (epitheca), from whichthe mandible differentiates. The hypostegal cavity of the autozoidremains in communication with the perigastric cavity only throughavicularia. Rare lateral avicularia similarly differentiatefrom pores connecting autozoids and basal kenozoids. All zoids heavily thicken secondarily; cuticles between adjacentzoids simultaneously are thrown into interlocking longitudinalfolds which probably strengthen the colony mechanically. Despitecontinuing carbonate secretion, the cystid epidermis is tenuousand incomplete; the peritoneum is absent or poorly-developed.This situation is usual among gymnolaemate Bryozoa. The bryozoanbody cavity is therefore usually a pseudocoel in the morphologicalsense, but developmental evidence indicates it represents adegenerate coelom. The ovicell develops as a compressed evagination from the bodycavity of the autozoid distal to the zoid which deposits itsegg in the ovicell. Both inner and outer walls calcify, th outermore heavily than the inner. Secondary calcification is achievedby engulfment of the ovicell by a fold of hypostegal coelomwhich calcifies basally. There is no nutritive median vesicle,so embryos develop without food. The polypide is unique in lackinga vestibule; this may be associated with the absense of a medialvesicle.     

Foraging strategies as a function of season and rank among wild female chimpanzees (Pan troglodytes)

Among mammals, female reproduction is generally thought to befood limited, and dominance should theoretically afford high-rankingfemales with access to better food resources. Although the importanceof dominance rank among female chimpanzees (Pan troglodytes)has been debated in the past, mounting evidence suggests thatrank is very important among females (P. t. schweinfurthii)at Gombe National Park, Tanzania. In this study, we investigatedthe influence of season and dominance rank on female foragingstrategies. We found that high-ranking females spent less timeforaging and tended to have a narrower diet breadth and higherdiet quality than subordinate females. In this way, subordinatefemale foraging strategies were consistent with how femalesin general adapted to periods of food scarcity. The resultsof this study therefore suggest that low-ranking females mayface persistent "food scarcity" as a result of interferencefood competition. We also provide evidence that subordinatesmay forage less efficiently because they occupy lower qualityhabitats or avoid associating with dominant females in sharedareas.     

Natural variation in the sexually selected feather ornaments of crested auklets (Aethia cristatella) does not predict future survival

We evaluated whether sexually selected crest and auricular plumefeather ornaments of crested auklet (Aethia cristatella) adultscovaried with individual local survival over 11 years (1991–2001).Crested auklets (n = 364 total) were captured near breedingsites, marked with color rings, and local survival estimateswere based on color ring resightings at a breeding colony. Survivalestimates and relationships among local survival and crest length,auricular plume length, mass and tarsus were evaluated usingthe program MARK. The best models included four groups, definedby sex and ease of resighting, that differed in resighting rate(p) but not local survival rate (). This model structure effectivelyexplained sources of variation in local survival and resightabilityamong individuals. The best fitting model showed local survivalrate varying annually, while accounting for differences in resightabilityof marked individuals between the sexes and groups ([t], p[sex*easeof resighting]). Annual local survival varied from 0.940 ±0.029 (SE) in 1993–1994 to 0.767 ± 0.034 in 1997–1998and averaged 0.859 ± 0.019. We found no evidence thatcrested auklet local survival covaried with continuous variationin individuals' ornaments. Simulations indicated that our dataset was sufficient to detect a relationship between local survivaland a covariate that equaled or exceeded a range of 8%. Theimplications for competing sexual selection mechanisms of empiricallymeasured survival–ornament relationships are controversial,but our study emphasizes that survival estimates for such investigationsmust control for confounding factors such as resighting rateas well as have sufficient statistical power and time scaleto be biologically meaningful. Our results are most consistentwith the idea that the conspicuous variation in crested auklet'sshowy ornaments is arbitrary with respect to individual viabilityas quantified by their long-term survival.     

The short-term regulation of foraging in harvester ants

In the seed-eating ant Pogonomyrmex barbatus, the return ofsuccessful foragers stimulates inactive foragers to leave thenest. The rate at which successful foragers return to the nestdepends on food availability; the more food available, the morequickly foragers will find it and bring it back. Field experimentsexamined how quickly a colony can adjust to a decline in therate of forager return, and thus to a decline in food availability,by slowing down foraging activity. In response to a brief, 3-to 5-min reduction in the forager return rate, foraging activityusually decreased within 2–3 min and then recovered within5 min. This indicates that whether an inactive forager leavesthe nest on its next trip depends on its very recent experienceof the rate of forager return. On some days, colonies respondedmore to a change in forager return rate. The rapid colony responseto fluctuations in forager return rate, enabling colonies toact as risk-averse foragers, may arise from the limited intervalover which an ant can track its encounters with returning foragers.     

Regulation of worker activity in a primitively eusocial wasp, Ropalidia marginata

Ropalidia marginala, a tropical, primitively eusocial, polistinewasp, is unusual in that the queen (the sole egg-layer) is neitherthe most behaviorally dominant nor the most active individualin the colony. The queen by herself rarely ever initiates interactionstoward her nest mates or unloads returning foragers. There arealways a few workers in the colony who are more dominant andactive than the queen. Absence of the queen from her colonydoes not affect colony maintenance activities such as foragingor brood care, but it always results in one individual becomingvery aggressive and dominant. The dominant worker becomes thenext queen if the original queen does not return. The queendoes not appear to play any significant role in colony activityregulation. Instead, colony activities appear to be regulatedby several mechanisms including dominance behavior toward foragers,feeding of larvae, and the unloading of returning foragers,all mediated by workers themselves. Regulation of colony maintenanceappears to be based on direct evaluation of the needs of thecolony by the workers themselves. The queen however has perfectreproductive control over all workers; workers never lay eggsin the presence of the queen. It appears therefore that themechanisms involved in regulation of worker activity and workerreproduction are separate in R marginata. These findings contrastwith other primitively eusocial species where the queen actsas a "central pacemaker" and controls both worker activity andworker reproduction.     

Studies on the Nutrition of Coprinus lagopus Fr., especially as affecting Fruiting

Growth and fruiting of Coprinus lagopus in pure culture havebeen studied on a medium containing glucose, dl--alanine, mineralsalts and thiamin, and the effects of changes in concentrationof certain components of this medium and of various substitutionsinvestigated. It is demonstrated that fruiting led to withdrawalof materials from the whole of the colony. No evidence was obtainedof the existence of a minimal quantity of substrate for fruitingto take place. The distribution of mature sporophores on a colonywas affected by an internal mechanism. A theory as to its possiblenature is suggested.     

Carbohydrates,invertase activity,growth and dimorphism inSporisorium reilianum

Sporisorium reilianum, the fungus that causes sorghum head smut, was grown with sucrose, lactose, trehalose or raffinose in liquid suspension or on a solid medium. Liquid culture media were analyzed for hydrolysis products of these carbohydrates to determine extracellular enzyme activity of the fungus. Increased amounts of glucose and fructose in the culture medium ofS. reilianum grown with sucrose or raffinose indicated that invertase (-fructofuranosidase, 3.2.1.26) activity was present. No evidence of extracellular galactosidase or trehalase activity was found. Enhanced sporidial colony formation on carbohydrates that can be hydrolyzed to hexoses, and specific forms of mycelial growth on lactose, trehalose or on a carbohydrate-deficient medium might suggest that mycelial growth is a way of foraging for food sources. However, the rapid and profuse mycelial growth on the host cell wall glycoprotein appears to be in response to abundant food supply (probably of a different type). Therefore availability of different kinds of carbon sources in the environment of the growing fungus might determine dimorphism and associated pathogenesis byS. reilianum.Technical Article No: 30699 from the Texas Agricultural Experiment Station.     

Experimental evaluation of herbivory in the ctenophore Mnemiopsis leidyi relevant to ctenophore-zooplankton-phytoplankton interactions in Narragansett Bay, Rhode Island, USA

Herbivory of Mnemiopsis leidyi and its interactions with phytoplanktonand non-gelatinous zooplankton were examined in small-scalemicrocosm experiments. Clearance rates for M. leidyi incubatedwith phytoplankton were generally negative, but ranged up to4.5 1 ctenophore^–1 day^–1 when the large (80 µmø) diatom Ditylum brightwelli was offered as a food source.These highest ingestion rates would provide Mnemiopsis withonly 21 % of its daily carbon requirements for respiration.Mean shrinkage of M. leidyi was 8.2–51% when incubatedwith phytoplankton. Although M. leidyi neither fed activelyon phytoplankton, nor satisfied its nutritional needs on sucha diet, the chain-forming diatom Skeletonema costatum becameentangled in mucus strands and balls produced by M. leidyi inthe absence of zooplankton. Attachment onto mucus occurred atphytoplankton concentrations commonly observed in NarragansettBay and may be important in the formation of "marine snow" duringsummer M. leidyi pulses; phytoplankton sinking rate and the"package size" available to herbivores would also be affected.The experiments support our previous hypothesis based on fieldobservations in Narragansett Bay that M. leidyi indirectly regulatesphytoplankton abundance there during the summer bloom as a consequenceof predation on zooplankton. The extent to which M. leidyi influencedphytoplankton dynamics in the microcosms was dependent on therelative abundance and physiological state of the three trophiclevels. A food web diagram for M. leidyi is presented.     

Food resources, chemical signaling, and nest mate recognition in the ant Formica aquilonia

Animals such as social insects that live in colonies can recognizeintruders from other colonies of the same or different speciesusing colony-specific odors. Such colony odors usually haveboth a genetic and an environmental origin. When within-colonyrelatedness is high (i.e., one or very few reproductive queens),colonies comprise genetically distinct entities, and recognitionbased on genetic cues is reliable. However, when nests containmultiple queens and colonies comprise multiple nests (polydomy),the use of purely genetically determined recognition labelsmay become impractical. This is due to high within-colony geneticheterogeneity and low between-colony genetic heterogeneity.This may favor the use of environmentally determined recognitionlabels. However, because nests within polydomous colonies maydiffer in their microenvironment, the use of environmental labelsmay also be impractical unless they are actively mixed amongthe nests. Using a laboratory experiment, we found that bothisolation per se and diet composition influenced the cuticularchemical profiles in workers of Formica aquilonia. In addition,the level of aggression increased when both the proportionsof dietary ingredients and the availability of food were altered.This suggests that increased aggression was mediated by changesin the chemical profile and that environmental cues can mediaterecognition between colonies. These results also suggest thatthe underlying recognition cues are mutable in response to extrinsicfactors such as the amount and the composition of food.     

Reproductive decision-making in the termite, Cryptotermes secundus (Kalotermitidae), under variable food conditions

Relatedness concepts have dominated the discussion on the evolutionand maintenance of eusociality in social insects. In the diploidtermites, explanations based on relatedness asymmetries havebeen less relevant than in the Hymenoptera; ecological factorshave been claimed to be paramount. Yet, relevant quantitativestudies investigating the role of ecological factors are lacking.We examined the influence of ecological factors on reproductivetactics in the drywood termite, Cryptotermes secundus. In thisspecies, caste development is very flexible, with individualshaving the option to remain at the natal nest as helpers/workersor to develop into dispersing reproductives (sexuals). An importantecological factor expected to influence this "decision" is foodavailability; C. secundus nests in a piece of wood that servesas food and shelter, with individuals never leaving the nestto forage. Thus, a reduction in the amount of food parallelsa reduction in the nests' longevity. Therefore, we tested theinfluence of food availability on caste-developmental decisionsin natural colonies, as well as in two experiments in whichwe simulated a gradual and a sudden decline in the amount ofavailable food. In all trials dispersing sexuals occurred moreoften in colonies with diminished food resources than in colonieswith abundant suitable food. Thus, regardless of how food declines,individuals seem to switch their tactic from being a helperto becoming a dispersing reproductive if nest conditions deteriorateand the nests's longevity decline.     

The benefits of stealing from a predator: foraging rates, predation risk, and intraspecific aggression in the kleptoparasitic spider Argyrodes antipodiana

In the trade-off between food and safety, the role of aggressiveintraspecific interactions has not been extensively examined.Here I present information on this system using a kleptoparasiticspider, Argyrodts antipodiana, and its host spider and potentialpredator, Eriophora pustulosa. A. antipodiana can feed eitherat a potentially dangerous site (the hub of its host's web withthe host), or at a relatively safe site (on food bundles aroundthe edge of the host's web). I found that A. antipodiana cangain food very quickly when feeding with the host, apparentlyby exploiting the host's ability to digest the prey. Thus A.antipodiana follows predictions based on foraging models inthat it accepts a higher predation risk at the hub because ofthe higher food payoff. A. antipodiana also aggressively competesfor access to more food. However, aggressive competition increasesthe predation risk from the host, especially at tile hub wherethe host is very close. Consequently, A. antipodiana modifiesits level of intraspecific aggressiveness in accordance withits position on the web: at the hub, where the cost of aggressionis high (due to predation risk), A. antipodiana reduces itsaggressiveness, but it is aggressive away from the hub whencompeting for food bundles. The ability of A. antipodiana tochange interaction intensity as a function of its position onthe web enables it to exploit a rich, but risky, food sourceand provides a new angle for examining food and safety tradeoffs in light of intraspecific competition for food     

The influence of forager number and colony size on food distribution in the odorous house ant, Tapinoma sessile

Stomodeal trophallaxis plays a major role in ant colony nutrition and communication. While the rate of food distribution at the individual level (worker to worker) is rapid, factors affecting the rate of food distribution at the colony level remain poorly understood. We used the odorous house ant, Tapinoma sessile (Say), as a model species to investigate the factors affecting the rate of spread of liquid carbohydrate food throughout a colony. To track the movement of the food we used protein marking and double antibody sandwich enzyme-linked immunosorbent assay, DAS-ELISA. Increasing colony size while keeping the number of donor workers constant significantly decreased the number of individuals testing positive for the marker. After 8 h of trophallactic interactions with ten donors, 92 ± 5% of recipient workers tested positive in a colony of 125 and 38 ± 5% tested positive in a colony of 1,000. Interestingly, as colony size increased and the percentage of workers testing positive decreased, the proportion of workers actually receiving food increased. Food originating from a single donor fed approximately 12 individuals in colonies comprised of 125 recipients and approximately 38 individuals in colonies comprised of 1,000 recipients. Thus, the per capita consumption of food decreased as colony size increased, most likely because the amount of food reaching the colony was limited. Increasing the number of donors while keeping colony size constant significantly increased the number of recipient ants testing positive for the marker. As the number of donor workers doubled, the percentage of recipients testing positive more than doubled suggesting that the number of individuals receiving food increases with increasing colony size, while the per capita amount of food decreases. When food was available ad libitum and in close proximity to the nest, numerous workers fed directly at the food source. This dramatically increased the rate and the extent of food distribution to both the workers and the queens and colony size had no significant effect on the spread of the marker in the workers or the queens. The rate and the extent of food distribution at the colony level may depend on a number of factors including the number of successful foragers, the size and density of the recipient colony, and the recipient caste.     

Queen number and genetic relatedness in a neotropical wasp, Polybia occidentalis

We found that genetic relatedness among Polybia occidentalisworkers was .26±0.057, a value high enough to make altruisticbehavior by workers relatively easy to explain. This comparativelyhigh level of relatedness can be attributed to close relatednessamong queens of .57±0.077 and to great variation amongcolonies in numbers of queens. The harmonic mean of queen numberis 3.1 queens per colony, which is much lower than the arithmeticmean of 10.6 queens per colony. These results are consistentwith a colony cycle called cyclical oligogyny, that is characterizedby a reduction in queen number from colony initiation to colonyreproduction. We did not find any evidence that one or a fewqueens monopolized egg laying or that there was any inbreeding,both of which have been hypothesized to increase relatednessamong workers. Another factor that can increase relatednessamong workers and the brood they rear is withincolony segregationon the basis of relatedness. We found that combmate pupae aresignificantly more closely related to each other (r = .41) thanthey are to pupae in other combs (r = .33), but we have notinvestigated whether workers take advantage of these relatednesspatterns. This distribution of relatedness among combs willoccur if queens do not lay eggs randomly throughout the nest,but concentrate their egg laying on one or a subset of the availablecombs.     

Zooplankton feeding ecology: clearance and ingestion rates of the salps Thalia democratica, Cyclosalpa affinis and Salpa cylindrica on naturally occurring particles in the Mid-Atlantic Bight

Aggregate stages of the salps Thalia democratica, Cyclosalpaaffinis and Salpa cylindrica collected by SCUBA diving in theMid-Atlantic Bight were fed with naturally occurring food assemblages.This is one of the few studies where salps have been fed withnatural food assemblages. The estimated clearance rate for allspecies based on disappearance of chlorophyll varied from 82to 444 mL individual^–1 day^–1. Cell counts showedthat T. democratica mostly ingested carbon from autotrophicnanoflagellates and autotrophic dinoflagellates. Ingestion byS. cylindrica was primarily on larger prey, such as dinoflagellates,while C. affinis ingested auto- and heterotrophic nanoflagellates.All main prey types ingested by salps corresponded to thosethat contributed most to biomass at each experimental station.Thus, salps fed on naturally occurring particles primarily inproportion to prey biomass and to their mechanical capacityto be retained and ingested. Feeding by salps on dinoflagellatesand ciliates implies that they may act not only as potentialcompetitors with microzooplankton, but also as consumers ofthem.     

Prey Abundance, Intraguild Predators, Ants and the Optimal Egg-laying Strategy of a Furtive Predator

Larval performance can have a great influence on female oviposition choice, especially in insects where the newly hatched offspring are unable to move any great distance to find an appropriate food source. For furtive predators, like the predatory midge Aphidoletes aphidimyza which preys on aphids while simultaneously residing and remaining undetected within their colonies, oviposition behaviour is crucial because these slow moving offspring are restrained to their natal colony. Here we develop a new model for predicting the optimum number of eggs that a furtive predator should lay in an aphid colony, based on: (1) the number of available prey (aphids); (2) the protection from predation conferred by “hiding” in the colony and (3) the effects of interspecific and intraspecific competition. We also explore the effect of aphid attendance by ants on oviposition behavior. We compare model predictions with empirical field observations of the clutch sizes of A. aphidimyza in apple orchards. The simplest of the four models best fits the observed data and provides the first field evidence that a furtive predator adjusts its clutch size as a function of prey density. The slope of the relationship between clutch size and aphid number is quite close to that predicted by our models suggesting that intra-clutch competition is the main factor governing furtive aphid midge oviposition choice.