Herbivorous arthropods cause immense damage in crop production annually. Consumption of these pests by insectivorous animals is of significant importance to counteract their adverse effects. Insectivorous bats are considered amongst the most voracious predators of arthropods, some of which are known crop pests. In vineyard-dominated Mediterranean agroecosystems, several crops are damaged by the attack of insect pests. In this study we aimed 1) to explore the diet and pest consumption of the lesser horseshoe bat Rhinolophus hipposideros and 2) analyse whether the composition of pest species in its diet changes throughout the season. We employed a dual-primer DNA metabarcoding analysis of DNA extracted from faeces collected in three bat colonies of a wine region in Southwestern Europe during the whole active period of most pest species. Overall, 395 arthropod prey species belonging to 11 orders were detected; lepidopterans and dipterans were the most diverse orders in terms of species. Altogether, 55 pest species were identified, 25 of which are known to cause significant agricultural damage and 8 are regarded as pests affecting grapevines. The composition of pest species in faeces changed significantly with the season, thus suggesting several periods should be sampled to assess the pest consumption by bats. As a whole, the results imply that R . hipposideros acts as a suppressor of a wide array of agricultural pests in Mediterranean agroecosystems. Therefore, management measures favouring the growth of R . hipposideros populations should be considered.

Knowledge on the trophic interactions among predators and their prey is important in order to understand ecology and behaviour of animals. Traditionally studies on the diet composition of insectivorous bats have been based on the morphological identification of prey remains, but the accuracy of the results has been hampered due to methodological limitations. Lately, the DNA metabarcoding and High Throughput Sequencing (HTS) techniques have changed the scene since they allows prey identification to the species level, ultimately giving more precision to the results. Nevertheless, the use of one single primer set to amplify faecal DNA produces biases in the assessed dietary composition. Three horseshoe bats overlap extensively in their distribution range in Europe: Rhinolophus euryale , R . hipposideros and R . ferrumequinum . In order to achieve the deepest insight on their prey list we combined two different primers. Results showed that the used primers were complementary at the order and species levels, only 22 out of 135 prey species being amplified by both. The most frequent prey of R . hipposideros belonged to Diptera and Lepidoptera, to Lepidoptera in R . euryale , and Lepidoptera, Diptera and Coleoptera in R . ferrumequinum . The three bats show significant resource partitioning, since their trophic niche overlap is not higher than 34%. Our results confirm the importance of combining complementary primers to describe the diet of generalist insectivorous bats with amplicon metabarcoding techniques. Overall, each primer set showed a subset of the prey composition, with a small portion of the total prey being identified by both of them. Therefore, each primer presented a different picture of the niche overlap among the three horseshoe bats due to their taxonomic affinity.

No abstract

Targeted molecular methods such as conventional PCR (cPCR) and quantitative PCR (qPCR), combined with species-specific primers and probes, are widely applied for pest species detection. Besides, the potential of qPCR to quantify DNA in samples makes it an invaluable molecular tool to infer the predation levels on specific prey by analysing predators’ stools. Nevertheless, studies on the diet of bats failed to find any empirical relationship, and it remains to be evaluated. Thus, we developed and evaluated two species-specific PCR assays to detect and quantify DNA of a major forest pest, the pine processionary, Thaumetopoea pityocampa, in bats’ faeces. Further, we empirically compared a range of different known DNA concentrations (input) of the target species mixed with mocks and bat faecal samples against DNA abundances yielded by qPCR (output) for a quantitative assessment. Overall, cPCR showed a lower detection rate than qPCR, but augmenting the replicate effort from one to three replicates led to a greater increase in the detection rate of the cPCR (from 57 to 80%) than the qPCR (from 90 to 99%). The quantitative experiment results showed a highly significant correlation between the input and output DNA concentrations (t = 10.84, p < 0.001) with a mean slope value of 1.05, indicating the accuracy of our qPCR assay to estimate DNA abundance of T. pityocampa in bat faeces. The framework of this study can be taken as a model to design similar assays applicable to other species of interest, such as agricultural pests or insects of public health concern.