Unnecessary roughness? Testing the hypothesis that predators destined for molecular gut‐content analysis must be hand‐collected to avoid cross‐contamination

Molecular gut‐content analysis enables detection of arthropod predation with minimal disruption of ecosystem processes. Mass‐collection methods, such as sweep‐netting, vacuum sampling and foliage beating, could lead to regurgitation or rupturing of predators along with uneaten prey, thereby contaminating specimens and compromising resultant gut‐content data. Proponents of this ‘cross‐contamination hypothesis’ advocate hand‐collection as the best way to avoid cross‐contamination. However, hand‐collection is inefficient when large samples are needed, as with most ecological research. We tested the cross‐contamination hypothesis by setting out onto potato plants immature Coleomegilla maculata and Podisus maculiventris that had been fed larvae of either Leptinotarsa decemlineata or Leptinotarsa juncta, or unfed individuals of these predator species along with L. decemlineata larvae. The animals were then immediately re‐collected, either by knocking them vigorously off the plants onto a beat cloth and capturing them en masse with an aspirator (‘rough’ treatment) or by hand‐searching and collection with a brush (‘best practice’). Collected predators were transferred in the field to individual vials of chilled ethanol and subsequently assayed by PCR for fragments of cytochrome oxidase I of L. decemlineata and L. juncta. Ten to 39 per cent of re‐collected fed predators tested positive by PCR for DNA of both Leptinotarsa species, and 14–38% of re‐collected unfed predators contained L. decemlineata DNA. Overall levels of cross‐contamination in the rough (31%) and best‐practice (11%) samples were statistically different and supported the cross‐contamination hypothesis. A pilot study on eliminating external DNA contamination with bleach prior to DNA extraction and amplification gave promising results.     

Maximizing collection and minimizing risk: does vacuum suction sampling increase the likelihood for misinterpretation of food web connections?

Molecular tools that characterize the structure of complex food webs and identify trophic connectedness in the field have become widely adopted in recent years. However, characterizing the intensity of predator-prey interactions can be prone to error. Maximizing collection success of small, fast-moving predators with vacuum suction samplers has the potential to increase the likelihood of prey DNA detection either through surface-level contamination with damaged prey or direct consumption within the sampling device. In this study, we used PCR to test the hypothesis that vacuum suction sampling will not cause an erroneous increase in the detection of 'predation', thereby incorrectly assigning trophic linkages when evaluating food web structure. We utilized general (1) Aphidoidea and (2) Collembola primers to measure the predation rates of Glenognatha foxi (Araneae: Tetragnathidae) on these prey collected by hand versus those sampled with a vacuum suction device. With both primer pairs, there was no significant increase in predators screening positive for prey DNA when sampled by vacuum suction versus those predators collected, in parallel, by hand. These results clearly validate the application of vacuum suction sampling during molecular gut-content analysis of predator-prey feeding linkages in the field. Furthermore, we found no evidence that predation was occurring inside the suction sampler because specimens collected were never observed to be feeding nor did they screen positive at greater frequencies than hand-collected individuals. Therefore, it can be concluded that the use of vacuum suction sampling devices (in this case a Modified CDC Backpack Aspirator Model 1412) is suitable for molecular gut-content analysis.     

A tale of two fluids: does storing specimens together in liquid preservative cause DNA cross‐contamination in molecular gut‐content studies?

The study of food webs and trophic interactions increasingly relies on PCR‐based molecular gut‐content analysis. However, this technique may be prone to error from contamination of minute quantities of DNA; i.e., simply storing specimens together in a liquid medium may lead to cross‐contamination. In this study, we used PCR to determine the contamination rate when (1) specimens were stored together in 95% ethanol for various time periods, and (2) predators fall into ethylene glycol‐filled pitfall traps where the dying predator may inadvertently consume prey DNA‐contaminated liquid. We designed experiments and PCR primers to quantify the risk of contamination for both situations and found no contamination by storing specimens together in 95% ethanol. Furthermore, zero predators contained prey DNA in their gut contents from imbibing prey DNA‐contaminated ethylene glycol. These results support the use of mass sampling techniques, like wet pitfall traps, for molecular gut‐content analysis.     

Feeding mode and prey detectability half-lives in molecular gut-content analysis: an example with two predators of the Colorado potato beetle

The time during which prey remains are detectable in the gut of a predator is an important consideration in the interpretation of molecular gut-content data, because predators with longer detectability times may appear on the basis of unweighted data to be disproportionately important agents of prey population suppression. The rate of decay in detectability, typically expressed as the half-life, depends on many variables; one that has not been explicitly examined is the manner in which the predator processes prey items. The influence of differences in feeding mode and digestive physiology on the half-life of DNA for a single prey species, the Colorado potato beetle Leptinotarsa decemlineata (Say), is examined in two predators that differ dramatically in these attributes: the pink ladybeetle, Coleomegilla maculata (DeGeer), which feeds by chewing and then ingesting the macerated material into the gut for digestion; and the spined soldier bug, Podisus maculiventris (Say), which physically and enzymatically processes the prey extra-orally before ingestion and further digestion in the gut. In order to standardize the amount of DNA consumed per predator, a single L. decemlineata egg was used as the prey item; all predators were third instars. The PCR assay yields estimated prey DNA half-lives, for animals maintained under field temperatures, of 7.0 h in C. maculata and 50.9 h in P. maculiventris. The difference in the prey DNA half-lives from these two predators underscores the need to determine detectabilities from assemblages of predators differing in feeding mode and digestive physiology, in order to weight positives properly, and hence determine the predators' relative impacts on prey population suppression.     

Linking habitat complexity with predation of pests through molecular gut-content analyses

Improving the diversity of farm systems or landscapes can lead to more effective biological control by providing refuge and alternative resources for colonising natural enemies. Within an experimental cabbage agroecosystem, we examined the effects of habitat management (i.e. herbicide use and cover crops) on pest populations and predator community structure, and report one of the first studies on the trophic links in this system using molecular gut-content analysis. In response to herbicide and cover crop management treatments designed to create different levels of habitat diversity, we quantified the abundance of two pests, Plutella xylostella Linnaeus (Lepidoptera: Plutellidae) and Pieris rapae Linnaeus (Lepidoptera: Pieridae), and predators. We designed species-specific primers to detect prey DNA in predators' guts. Pieris rapae were significantly more abundant in plots where cover crops were killed early in the season, and habitat management generated unique predator communities in response to weed management treatments. Thirty-five per cent of predators tested positive for prey DNA, and habitat management had interactive effects on predation of P. xylostella. Combined we found that habitat management has variable effects on natural enemy–pest interactions.     

An order-specific monoclonal antibody to Diptera reveals the impact of alternative prey on spider feeding behavior in a complex food web

Generalist predators have the capacity to restrict pest population growth, especially early in the season before densities increase. However, their polyphagous feeding habits sometimes translate into reduced pest consumption when they target alternative prey. An order-specific monoclonal antibody was developed to examine the strength of trophic connections between Diptera, a major category of non-pest prey, and linyphiid spiders in alfalfa. We report the development and characterization of a monoclonal antibody with order-level specificity to Diptera. This antibody elicited strong absorbance to 22 Diptera from 13 families, no false-positive reactivity to non-dipteran invertebrates, and antigen detection periods following prey consumption that were comparable between spiders. Over 900 field-collected females of the linyphiid spiders Erigone autumnalis and Bathyphantes pallidus were screened for Diptera antigen. Significantly more B. pallidus screened positive for Diptera (40%) compared to E. autumnalis (16%), indicating differential reliance on these prey. In parallel with the collection of spiders for gut-content analysis, prey availability was estimated at web sites. The two spiders exhibited different feeding responses to prey availability. Consumption of Diptera by B. pallidus was strongly correlated with Diptera abundance whilst the availability of other potential prey did not influence predation rates. Conversely, E. autumnalis did not prey upon Diptera in proportion to availability, but increased Collembola activity-density reduced dipteran consumption. Integration of molecular gut-content analysis with precise sampling of prey demonstrated how two closely related linyphiid spiders exhibit different feeding responses to the availability of prey under natural field conditions. Elucidating the feeding preferences of natural enemies is critical to effective incorporation of biological control by generalist predators in the management of agricultural pests.     

Molecular gut-content analysis of a predator assemblage reveals the effect of habitat manipulation on biological control in the field

Habitat manipulation in agroecosystems can influence predator–prey interactions. In this study, we collected foliar predators from field potato plots with different mulch treatments and assayed them for DNA of the target prey, Leptinotarsa decemlineata (Say), using species-specific primers. Concurrently, L. decemlineata larval abundance and plant damage were recorded from the same plots. Predator species abundance and diversity were not influenced by habitat manipulation, while prey density was highest in plots without mulch. Gut-content analysis revealed that the highest incidence of predators positive for L. decemlineata DNA was in plots without mulch, where target prey abundance was highest. Therefore, the lower prey abundance in mulched plots was not due to predation. The most abundant species in the predator assemblage was Coleomegilla maculata, which had the lowest proportion of L. decemlineata DNA in the gut. Podisus maculiventris, Perillus bioculatus, and Lebia grandis were less abundant but had a higher incidence of target prey DNA in the gut. DNA detectability half-lives were used to adjust for inter-specific variation in DNA digestive rates of the four predator species. Using this information to adjust actual number of positives for prey DNA, we compared proportions positive for L. decemlineata and found that P. maculiventris is the most effective predator species in the complex.     

The influence of time and temperature on molecular gut content analysis： Adalia bipunctata fed with Rhopalosiphum padi

Gut content analysis is a useful tool when studying arthropod predator-prey interactions. We used polymerase chain reaction （PCR） technique to examine how detection of prey DNA in the gut content of predators was influenced by digestion time and temperature. Such knowledge is critical before applying PCR-based gut content analysis to field collected predators. Larvae of the two-spotted ladybeetle （Adalia bipunctata L.） were fed with the bird cherry-oat aphid （Rhopalosiphum padi L.） at either 21℃ or 14℃. After consuming one aphid, the predators were allowed to digest the prey for a range of time periods up to 24 hours. The influence of temperature on A. bipunctata feeding behavior was also recorded. From the fed larvae, total DNA was extracted and PCR reactions with R. padi specific primers were run. The number ofA. bipunctata that tested positive for R. padi DNA was negatively related to the length of digestion time. Temperature influenced larval feeding behavior but did not have a significant effect on R. padi DNA detection. After pooling the data from both temperature treatments we estimated the time point when R. padi DNA could be amplified from 50% of the fed A. bipunctata by PCR to be 4.87 hours. With such a rapid decrease in prey DNA detection success, positive PCR reactions will most likely be the result of predation events occurring shortly before capture. If a defined digestion temperature range has proven not to influence prey detection, PCR data obtained from predators collected within that particular range can be interpreted in the same way.     

Tracking the role of alternative prey in soybean aphid predation by Orius insidiosus: a molecular approach

The soybean aphid, Aphis glycines (Hemiptera: Aphididae), is a pest of soybeans in Asia, and in recent years has caused extensive damage to soybeans in North America. Within these agroecosystems, generalist predators form an important component of the assemblage of natural enemies, and can exert significant pressure on prey populations. These food webs are complex and molecular gut-content analyses offer nondisruptive approaches for examining trophic linkages in the field. We describe the development of a molecular detection system to examine the feeding behaviour of Orius insidiosus (Hemiptera: Anthocoridae) upon soybean aphids, an alternative prey item, Neohydatothrips variabilis (Thysanoptera: Thripidae), and an intraguild prey species, Harmonia axyridis (Coleoptera: Coccinellidae). Specific primer pairs were designed to target prey and were used to examine key trophic connections within this soybean food web. In total, 32% of O. insidiosus were found to have preyed upon A. glycines, but disproportionately high consumption occurred early in the season, when aphid densities were low. The intensity of early season predation indicates that O. insidiosus are important biological control agents of A. glycines, although data suggest that N. variabilis constitute a significant proportion of the diet of these generalist predators. No Orius were found to contain DNA of H. axyridis, suggesting intraguild predation upon these important late-season predators during 2005 was low. In their entirety, these results implicate O. insidiosus as a valuable natural enemy of A. glycines in this soybean agroecosystem.     

Web orientation and prey resources for web-building spiders in eastern hemlock

We examined the arthropod community on eastern hemlock, Tsuga canadensis (L.) Carr, in the context of its role in providing potential prey items for hemlock-associated web-weaving spiders. Using sticky traps simulating spider webs, we evaluated what prey items are available to web-weaving spiders in eastern hemlock based on web orientation (horizontal versus vertical) and cardinal direction. We found that the overwhelming majority (>70%) of prey items available to spiders in hemlock canopies were Diptera. Psocoptera, Hymenoptera, and Hemiptera comprised most of the remaining potential prey. A significant direction × orientation interaction, and greater trap capture in some direction-orientation combinations, suggests that spiders might locate their webs in eastern hemlock canopies for thermoregulatory purposes, ultimately optimizing prey capture. We also evaluated these findings in the context of hemlock infestation by the invasive hemlock woolly adelgid, Adelges tsugae Annand. The adelgid is a sedentary insect with a mobile crawler stage that provides a readily available, easily obtained food source for predators in hemlock canopies. However, an abundance of alternative prey will affect within canopy spider distribution and the potential intensity with which spiders consume these prey. Understanding the response of spiders to potential prey availability is essential to understanding the trophic interactions involving these predators and their potential for influencing herbivore populations.     

Molecular identification of predation by carabid beetles on exotic and native slugs in a strawberry agroecosystem

Generalist predators have the capacity to exert significant pressure on prey populations. However, integrating them into biological control programs relies on a detailed understanding of their foraging behavior and the levels of trophic connectedness with pest species. Carabid beetles are important predators of slugs, pests of agricultural, floricultural and horticultural crops worldwide, but these interactions have been rarely studied outside the Western Palearctic ecozone. Diagnostic molecular gut-content analysis was used to examine the strength of trophic pathways between a community of carabid beetles and two slug species, the exotic Deroceras reticulatum and native Deroceras laeve, in strawberry agroecosystems. Strawberries were grown according to standard horticultural practices for central Kentucky, following traditional bare ground planting or with the addition of detrital subsidies, to quantify the impact of habitat management on the abundance of pests and the strength of these trophic pathways. Following laboratory characterization of species-specific molecular markers targeting both Deroceras species, carabid beetles collected from a strawberry agroecosystem were screened for slug DNA. Field collections revealed important food web pathways existed between Harpalus pensylvanicus and D. reticulatum, with 7.2% screening positive for these prey yet none screening positive for D. laeve. In contrast, Chlaenius tricolor was found to feed on both slug species in the field, with 16% screening positive for both Deroceras. Despite below normal rainfall limiting slug densities in the field, the results presented here reveal the potential importance of carabid beetles in slug population dynamics in the Nearctic.     

Can multiple-copy sequences of prey DNA be detected amongst the gut contents of invertebrate predators?

The first experiments to clearly demonstrate that DNA techniques might be used to detect predator-prey interactions between arthropods are reported. The accurate modelling of such interactions has depended until now upon a mixture of laboratory experiments, population monitoring and biochemical tests. The latter involve gut-content analyses, and have most recently depended upon the development of prey-specific monoclonal antibodies. Although these are excellent for detecting predation on a target prey, they are impractical for analysing the prey range of a particular predator. Molecular detection depends upon the ability of DNA to resist digestion in the predator gut and of the polymerase chain reaction (PCR) to amplify prey-specific DNA from semidigested material. As a first step, experiments using carabid beetles, Pterostichus cupreus L., as predators and mosquitoes as prey are reported. The target sequences were fully characterized multiple-copy esterase genes from two laboratory strains of Culex quinquefasciatus Say. Although DNA was extracted from homogenates of whole beetles (minus appendages), a 146 bp product could be amplified from both mosquito strains digested in the beetle gut for 28 h. The larger, 263 bp product was detectable for 28 h in one mosquito strain, but could not be amplified after 5 h from the other. Whether the beetles had eaten one mosquito or six, digested for zero or 28 h, the prey were equally detectable. Having demonstrated that shorter, multiple-copy sequences survive digestion for a considerable period in the gut of a predator, the opportunity exists to develop new detection systems for studying predation in the field.     

Short-term responses to elevated predator densities: noncompetitive intraguild interactions and behavior

We investigated the short-term response of an arthropod assemblage to elevated generalist predator densities by introducing Chinese mantids (Tenodera sinensis) to field plots in a replicated, controlled experiment. Abundances of carnivorous arthropods were reduced by mantids to a greater extent than herbivores, and cursorial spiders emigrated from treatment plots in greater numbers than from controls. Initially, this emigration consisted only of small spiders that were demonstrated in the laboratory to be prey for mantids. Thus, the initial response of an arthropod assemblage to increased predators, densities was increased interactions among predators, which caused decline in predator population densities in a shorter time than competition for prey would require. Predator avoidance behavior must be considered together with intraguild predation and competition when interpreting the outcome of predator manipulations. Shortterm experiments may be more valuable than longer term studies in detecting this effect.     

Infiltration by alien predators into invertebrate food webs in Hawaii: a molecular approach

Abstract Alien invertebrate predators have been introduced to Hawaii to control pests, particularly in lowland areas where most crops are grown. We developed techniques for assessing the impact of these predators on native food webs in relatively pristine upland areas where, it was hypothesized, few lowland predators might be found. Predator densities were assessed along transects within the Alakaii Swamp on Kaua'i. The most numerous alien biocontrol agents found were Halmus chalybeus (Coccinellidae), a species known to feed on Lepidoptera eggs. Laboratory experiments were conducted using two genera of endemic Lepidoptera, Scotorythra and Eupithecia (Geometridae), that are of considerable conservation value, the former because of its recent speciation across Hawaii, the latter for its unique predatory larvae. Techniques were developed for detecting Lepidoptera DNA within the guts of alien predators using prey-specific PCR primers. General primers amplified fragments of the mitochondrial cytochrome oxidase I gene from beetles and Lepidoptera. The sequences were aligned and used successfully to design target-specific primers for general detection of the remains of Geometridae and for particular species, including Scotorythra rara and Eupithecia monticolans. DNA fragments amplified were short [140-170 base pairs (bp)], optimizing detection periods following prey ingestion. Trials using the introduced biocontrol agent Curinus coeruleus (Coccinellidae) demonstrated detection of Lepidoptera DNA fragments = 151 bp in 85-100% of beetles after 24 h digestion of an early instar larva. This study provides a framework for future use of molecular gut analysis in arthropod conservation ecology and food web research with considerable potential for quantifying threats to endemic species in Hawaii and elsewhere.     

Trophic structure of the spider community of a Mediterranean citrus grove: A stable isotope analysis

Spiders are dominant terrestrial predators that consume a large variety of prey and engage in intraguild predation. Although the feeding habits of certain species are well known, the trophic structure of spider assemblages still needs to be investigated. Stable isotope analysis enables characterisation of trophic relationships between organisms because it tracks the energy flow in food webs and indicates the average number of trophic transfers between a given species and the base of the web, thus being a useful tool to estimate the magnitude of intraguild predation in food webs. Using this technique, we studied the trophic groups of spiders and their links within the arthropod food web of a Mediterranean organic citrus grove. We assessed the trophic positions of the 25 most common spider species relative to other arthropod predators and potential prey in the four seasons of the year, both in the canopy and on the ground. The analyses showed great seasonal variation in the isotopic signatures of some arthropod species, as well as the existence of various trophic groups and a wide range of trophic levels among spiders, even in species belonging to the same family. Differences in δ¹⁵N between spiders and the most abundant prey in the grove usually spanned two trophic levels or more. Our findings provide field evidence of widespread intraguild predation in the food web and caution against using spider families or guilds instead of individual species when studying spider trophic interactions.     

Taxon-specific PCR for DNA barcoding arthropod prey in bat faeces

The application of DNA barcoding to dietary studies allows prey taxa to be identified in the absence of morphological evidence and permits a greater resolution of prey identity than is possible through direct examination of faecal material. For insectivorous bats, which typically eat a great diversity of prey and which chew and digest their prey thoroughly, DNA-based approaches to diet analysis may provide the only means of assessing the range and diversity of prey within faeces. Here, we investigated the effectiveness of DNA barcoding in determining the diets of bat species that specialize in eating different taxa of arthropod prey. We designed and tested a novel taxon-specific primer set and examined the performance of short barcode sequences in resolving prey species. We recovered prey DNA from all faecal samples and subsequent cloning and sequencing of PCR products, followed by a comparison of sequences to a reference database, provided species-level identifications for 149/207 (72%) clones. We detected a phylogenetically broad range of prey while completely avoiding detection of nontarget groups. In total, 37 unique prey taxa were identified from 15 faecal samples. A comparison of DNA data with parallel morphological analyses revealed a close correlation between the two methods. However, the sensitivity and taxonomic resolution of the DNA method were far superior. The methodology developed here provides new opportunities for the study of bat diets and will be of great benefit to the conservation of these ecologically important predators.     

Species‐level predation network uncovers high prey specificity in a Neotropical army ant community

Army ants are among the top arthropod predators and considered keystone species in tropical ecosystems. During daily mass raids with many thousand workers, army ants hunt live prey, likely exerting strong top‐down control on prey species. Many tropical sites exhibit a high army ant species diversity (>20 species), suggesting that sympatric species partition the available prey niches. However, whether and to what extent this is achieved has not been intensively studied yet. We therefore conducted a large‐scale diet survey of a community of surface‐raiding army ants at La Selva Biological Station in Costa Rica. We systematically collected 3,262 prey items from eleven army ant species (genera Eciton, Nomamyrmex and Neivamyrmex). Prey items were classified as ant prey or non‐ant prey. The prey nearly exclusively consisted of other ants (98%), and most booty was ant brood (87%). Using morphological characters and DNA barcoding, we identified a total of 1,103 ant prey specimens to the species level. One hundred twenty‐nine ant species were detected among the army ant prey, representing about 30% of the known local ant diversity. Using weighted bipartite network analyses, we show that prey specialization in army ants is unexpectedly high and prey niche overlap very small. Besides food niche differentiation, we uncovered a spatiotemporal niche differentiation in army ant raid activity. We discuss competition‐driven multidimensional niche differentiation and predator–prey arms races as possible mechanisms underlying prey specialization in army ants. By combining systematic prey sampling with species‐level prey identification and network analyses, our integrative approach can guide future research by portraying how predator–prey interactions in complex communities can be reliably studied, even in cases where morphological prey identification is infeasible.     

Effects of Management Intensity and Season on Arboreal Ant Diversity and Abundance in Coffee Agroecosystems

Agricultural intensification decreases arthropod predator diversity, abundance and population stability, and may affect interactions between top predators and their arthropod prey – ultimately affecting ecosystem services. Coffee management intensification (reduction or removal of shade trees) reduces diversity of arthropod predators (ground-foraging ants). Because ants provide ecosystem services by controlling pests, influences of intensification on arboreal, coffee-foraging ant diversity and abundance are important. We here address how coffee intensification affects: (1) coffee-foraging ant diversity and abundance and (2) seasonal fluctuations in ant abundance. In each of four coffee sites of varying management intensity in Chiapas, Mexico, we sampled vegetation and using two methods, sampled ant diversity and abundance over two years. Sites significantly differed in vegetation and management intensity. Coffee-foraging ant diversity generally decreased with increasing management intensity (16–26% fewer species observed in the most intensively-managed site). Ant abundance was higher in the wet season. Management intensity, however, did not influence ant abundance or seasonal fluctuations in abundance. Our results highlight the importance of diverse agricultural systems in maintaining arthropod predator diversity, and point to one model system in which we may effectively test how diversity per se affects ecosystem services.     

Predation pressure on the arboreal epiphytic herbivores of larch trees in southern England

Abstract. 1. Quantitative collections were made of arthropod predators from larch trees at Ranmore Common, Surrey, from August 1977 to September 1978.
2. Spiders were numerically the most important predators accounting for 64% of the total.
3. Antisera of the main larch dwelling, epiphyte herbivore orders, Collembola and Psocoptera, and the other common arboreal herbivore group at Ranmore, the Honioptera (principally Psyllidae) were prepared in rabbits.
4. Serological determinations of predator diets indicated that, with a few exceptions, the predators were taking approximately equal proportions of the three prey groups.
5. This apparent non-specificity of the predators is analysed in relation to both the numerical abundance and biomass of the three prey types during the study period.
6. These arboreal arthropod predators are selecting for large but rare prey items in preference to small and common ones.
7. These findings are discussed in relation to the population dynamics of arboreal epiphytic herbivores.     

Testing the validity of functional response models using molecular gut content analysis for prey choice in soil predators

Analysis of predator–prey interactions is a core concept of animal ecology, explaining structure and dynamics of animal food webs. Measuring the functional response, i.e. the intake rate of a consumer as a function of prey density, is a powerful method to predict the strength of trophic links and assess motives of prey choice, particularly in arthropod communities. However, due to their reductionist set‐up, functional responses, which are based on laboratory feeding experiments, may not display field conditions, possibly leading to skewed results. Here, we tested the validity of functional responses of centipede predators and their prey by comparing them with empirical gut content data from field‐collected predators. Our predator–prey system included lithobiid and geophilomorph centipedes, abundant and widespread predators of forest soils and their soil‐dwelling prey. First, we calculated the body size‐dependent functional responses of centipedes using a published functional response model in which we included natural prey abundances and animal body masses. This allowed us to calculate relative proportions of specific prey taxa in the centipede diet. In a second step, we screened field‐collected centipedes for DNA of eight abundant soil‐living prey taxa and estimated their body size‐dependent proportion of feeding events. We subsequently compared empirical data for each of the eight prey taxa, on proportional feeding events with functional response‐derived data on prey proportions expected in the gut, showing that both approaches significantly correlate in five out of eight predator–prey links for lithobiid centipedes but only in one case for geophilomorph centipedes. Our findings suggest that purely allometric functional response models, which are based on predator–prey body size ratios are too simple to explain predator–prey interactions in a complex system such as soil. We therefore stress that specific prey traits, such as defence mechanisms, must be considered for accurate predictions.