The “habitat heterogeneity hypothesis” predicts positive effects of structural complexity on species coexistence. Increasing habitat heterogeneity can change the diversity (number of species, abundances) and the functional roles of communities. The latter, however, is not well understood as species and individuals may respond very differently and dynamically to a changing environment.Here, we experimentally test how habitat heterogeneity affects generalist arthropod predators, including epigaeic spiders, carabid and staphylinid beetles, under natural conditions by assessing their diversity and directly measuring their trophic interactions (which provide a proxy for their functional roles). The experiment was conducted in spring barley fields in Southern Sweden where habitat heterogeneity was manipulated by increasing within‐field plant diversity.Increased habitat heterogeneity triggered rapid changes in the feeding behaviour of generalist predators characterized by lower trophic specialization at both network (H2’, degree of interaction specialization in the entire network) and species level (d’, degree of interaction specialization at the species level). We presume that this is because spatial separation resulted in relaxed competition and allowed an increased overlap in resources used among predator species. Predators collected from heterogenous habitats also showed greater individual‐level dietary variability which might be ascribed to relaxed intraspecific competition.Our results provide conclusive evidence that habitat heterogeneity can induce rapid behavioural responses independent of changes in diversity, potentially promoting the stability of ecosystem functions. A plain language summary is available for this article.

The rapidly growing field of molecular diet analysis is becoming increasingly popular among ecologists, especially when investigating methodologically challenging groups, such as invertebrate generalist predators. Prey DNA detection success is known to be affected by multiple factors; however, the type of dietary sample has rarely been considered. Here, we address this knowledge gap by comparing prey DNA detection success from three types of dietary samples. In a controlled feeding experiment, using the carabid beetle Pterostichus melanarius as a model predator, we collected regurgitates, faeces and whole consumers (including their gut contents) at different time points postfeeding. All dietary samples were analysed using multiplex PCR, targeting three different length DNA fragments (128, 332 and 612 bp). Our results show that both the type of dietary sample and the size of the DNA fragment contribute to a significant part of the variation found in the detectability of prey DNA. Specifically, we observed that in both regurgitates and whole consumers, prey DNA was detectable significantly longer for all fragment sizes than for faeces. Based on these observations, we conclude that prey DNA detected from regurgitates and whole consumers DNA extracts are comparable, whereas prey DNA detected from faeces, though still sufficiently reliable for ecological studies, will not be directly comparable to the former. Therefore, regurgitates and faeces constitute a useful, nonlethal source for dietary information that could be applied to field studies in situations when invertebrate predators should not be killed.

The nucleotide variation in the cytochrome c oxidase subunit I (COI) gene makes it ideal for assigning sequences to species. However, this variability also makes it difficult to design truly universal primers. Here, we present the forward primer “Sauron‐S878,” specifically designed to facilitate library preparation for metabarcoding. This primer is modified to improve the coverage of terrestrial species compared to the primer mCOIintF, optimized for aquatic systems, which raised the in silico coverage from 74.4% to 98.3% of available NCBI sequences (perfect match in 3′ region, up to three mismatches in remaining primer). When paired with the reverse primer “jgHCO2198” (fragment length ~313 bp), these primers amplified 98.4% of 255 tested DNA extracts from various taxa, which are better than many other common COI barcoding primers. Furthermore, a single‐tube protocol was developed, wherein these primers amplify the target gene, and attach MIDs and Illumina sequencing adapters in one reaction. This eliminates the need for re‐amplification or enzymatic ligation during library preparation while keeping the flexibility to modularly combine primers and MIDs. Using the single‐tube approach, three replicates of three mock samples were sequenced on a MiSeq platform with no adverse effects compared to commercial Nextera indexing kits. From this run, 75% of all included taxa could be recovered, with no considerable bias among taxonomic groups. Despite the fact that 98.4% of the extracts were confirmed to amplify in vitro, this number was lower than expected. A reason for this discrepancy was a clear link between the relative concentration of a specific DNA type in the template and the number of returned reads for this DNA. We would argue that such a bias may be especially problematic in metabarcoding where samples usually contain trace DNA in unknown amounts. However, how this affects the completeness of metabarcoding results has yet been poorly investigated.

Facultative bacterial endosymbionts can protect their aphid hosts from natural enemies such as hymenopteran parasitoids. As such, they have the capability to modulate interactions between aphids, parasitoids and hyperparasitoids. However, the magnitude of these effects in natural aphid populations and their associated parasitoid communities is currently unknown. Moreover, environmental factors such as plant fertilization and landscape complexity are known to affect aphid–parasitoid interactions but it remains unclear how such environmental factors affect the interplay between aphids, parasitoids and endosymbionts.Here, we tested whether facultative endosymbionts confer protection to parasitoids in natural populations of the English grain aphid, Sitobion avenae, and if this is affected by plant fertilization and landscape complexity. Furthermore, we examined whether the effects of facultative endosymbionts can cascade up to the hyperparasitoid level and increase primary‐hyperparasitoid food web specialization.Living aphids and mummies were collected in fertilized and unfertilized plots within 13 wheat fields in Central Germany. We assessed the occurrence of primary parasitoid, hyperparasitoid and endosymbiont species in aphids and mummies using a newly established molecular approach.Facultative endosymbiont infection rates were high across fields (~80%), independent of whether aphids were parasitized or unparasitized. Aphid mummies exhibited a significantly lower share of facultative endosymbiont infection (~38%). These findings suggest that facultative endosymbionts do not affect parasitoid oviposition behaviour, but decrease parasitoid survival in the host. Facultative endosymbiont infection rates were lower in mummies collected from fertilized compared to unfertilized plants, indicating that plant fertilization boosts the facultative endosymbiont protective effect. Furthermore, we found strong evidence for species‐specific and negative cascading effects of facultative endosymbionts on primary and hyperparasitoids, respectively. Facultative endosymbionts impacted parasitoid assemblages and increased the specialization of primary‐hyperparasitoid food webs: these effects were independent from and much stronger than other environmental factors.The current findings strongly suggest that facultative endosymbionts act as a driving force in aphid–parasitoid–hyperparasitoid networks: they shape insect community composition at different trophic levels and modulate, directly and indirectly, the interactions between aphids, parasitoids and their environment.

Next‐generation sequencing (NGS) is increasingly used for diet analyses; however, it may not always describe diet samples well. A reason for this is that diet samples contain mixtures of food DNA in different amounts as well as consumer DNA which can reduce the food DNA characterized. Because of this, detections will depend on the relative amount and identity of each type of DNA. For such samples, diagnostic PCR will most likely give more reliable results, as detection probability is only marginally dependent on other copresent DNA. We investigated the reliability of each method to test (a) whether predatory beetle regurgitates, supposed to be low in consumer DNA, allow to retrieve prey sequences using general barcoding primers that co‐amplify the consumer DNA, and (b) to assess the sequencing depth or replication needed for NGS and diagnostic PCR to give stable results. When consumer DNA is co‐amplified, NGS is better suited to discover the range of possible prey, than for comparing co‐occurrences of diet species between samples, as retested samples were repeatedly different in prey detections with this approach. This shows that samples were incompletely described, as prey detected by diagnostic PCR frequently were missed by NGS. As the sequencing depth needed to reliably describe the diet in such samples becomes very high, the cost‐efficiency and reliability of diagnostic PCR make diagnostic PCR better suited for testing large sample‐sets. Especially if the targeted prey taxa are thought to be of ecological importance, as diagnostic PCR gave more nested and consistent results in repeated testing of the same sample.