Ecological specialization and niche overlap of subterranean rodents inferred from DNA metabarcoding diet analysis.

Knowledge of how animal species use food resources available in the environment can increase our understanding of many ecological processes. However, obtaining this information using traditional methods is difficult for species feeding on a large variety of food items in highly diverse environments. We amplified the DNA of plants for 306 scat and 40 soil samples, and applied an environmental DNA metabarcoding approach to investigate food preferences, degree of diet specialization and diet overlap of seven herbivore rodent species of the genus Ctenomys distributed in southern and midwestern Brazil. The metabarcoding approach revealed that these species consume more than 60% of the plant families recovered in soil samples, indicating generalist feeding habits of ctenomyids. The family Poaceae was the most common food resource retrieved in scats of all species as well in soil samples. Niche overlap analysis indicated high overlap in the plant families and molecular operational taxonomic units consumed, mainly among the southern species. Interspecific differences in diet composition were influenced, among other factors, by the availability of resources in the environment. In addition, our results provide support for the hypothesis that the allopatric distributions of ctenomyids allow them to exploit the same range of resources when available, possibly because of the absence of interspecific competition.

Body size determines soil community assembly in a tropical forest.

Tropical forests shelter an unparalleled biological diversity. The relative influence of environmental selection (i.e., abiotic conditions, biotic interactions) and stochastic-distance-dependent neutral processes (i.e., demography, dispersal) in shaping communities has been extensively studied for various organisms, but has rarely been explored across a large range of body sizes, in particular in soil environments. We built a detailed census of the whole soil biota in a 12-ha tropical forest plot using soil DNA metabarcoding. We show that the distribution of 19 taxonomic groups (ranging from microbes to mesofauna) is primarily stochastic, suggesting that neutral processes are prominent drivers of the assembly of these communities at this scale. We also identify aluminium, topography and plant species identity as weak, yet significant drivers of soil richness and community composition of bacteria, protists and to a lesser extent fungi. Finally, we show that body size, which determines the scale at which an organism perceives its environment, predicted the community assembly across taxonomic groups, with soil mesofauna assemblages being more stochastic than microbial ones. These results suggest that the relative contribution of neutral processes and environmental selection to community assembly directly depends on body size. Body size is hence an important determinant of community assembly rules at the scale of the ecological community in tropical soils and should be accounted for in spatial models of tropical soil food webs.

Evaluating mortality rates with a novel integrated framework for nonmonogamous species.

The conservation of wildlife requires management based on quantitative evidence, and especially for large carnivores, unraveling cause-specific mortalities and understanding their impact on population dynamics is crucial. Acquiring this knowledge is challenging because it is difficult to obtain robust long-term data sets on endangered populations and, usually, data are collected through diverse sampling strategies. Integrated population models (IPMs) offer a way to integrate data generated through different processes. However, IPMs are female-based models that cannot account for mate availability, and this feature limits their applicability to monogamous species only. We extended classical IPMs to a two-sex framework that allows investigation of population dynamics and quantification of cause-specific mortality rates in nonmonogamous species. We illustrated our approach by simultaneously modeling different types of data from a reintroduced, unhunted brown bear (Ursus arctos) population living in an area with a dense human population. In a population mainly driven by adult survival, we estimated that on average 11% of cubs and 61% of adults died from human-related causes. Although the population is currently not at risk, adult survival and thus population dynamics are driven by anthropogenic mortality. Given the recent increase of human-bear conflicts in the area, removal of individuals for management purposes and through poaching may increase, reversing the positive population growth rate. Our approach can be generalized to other species affected by cause-specific mortality and will be useful to inform conservation decisions for other nonmonogamous species, such as most large carnivores, for which data are scarce and diverse and thus data integration is highly desirable.

Individual genotypes from environmental DNA: Fingerprinting snow tracks of three large carnivore species.

Continued advancements in environmental DNA (eDNA) research have made it possible to access intraspecific variation from eDNA samples, opening new opportunities to expand non-invasive genetic studies of wildlife populations. However, the use of eDNA samples for individual genotyping, as typically performed in non-invasive genetics, still remains elusive. We present successful individual genotyping of eDNA obtained from snow tracks of three large carnivores: brown bear (Ursus arctos), European lynx (Lynx lynx) and wolf (Canis lupus). DNA was extracted using a protocol for isolating water eDNA and genotyped using amplicon sequencing of short tandem repeats (STR), and for brown bear a sex marker, on a high-throughput sequencing platform. Individual genotypes were obtained for all species, but genotyping performance differed among samples and species. The proportion of samples genotyped to individuals was higher for brown bear (5/7) and wolf (7/10) than for lynx (4/9), and locus genotyping success was greater for brown bear (0.88). The sex marker was typed in six out of seven brown bear samples. Results for three species show that reliable individual genotyping, including sex identification, is now possible from eDNA in snow tracks, underlining its vast potential to complement the non-invasive genetic methods used for wildlife. To fully leverage the application of snow track eDNA, improved understanding of the ideal species- and site-specific sampling conditions, as well as laboratory methods promoting genotyping success, is needed. This will also inform efforts to retrieve and type nuclear DNA from other eDNA samples, thereby advancing eDNA-based individual and population-level studies.

eDNA metabarcoding for biodiversity assessment, generalist predators as sampling assistants.

With an accelerating negative impact of anthropogenic actions on natural ecosystems, non-invasive biodiversity assessments are becoming increasingly crucial. As a consequence, the interest in the application of environmental DNA (eDNA) survey techniques has increased. The use of eDNA extracted from faeces from generalist predators, have recently been described as "biodiversity capsules" and suggested as a complementary tool for improving current biodiversity assessments. In this study, using faecal samples from two generalist omnivore species, the Eurasian badger and the red fox, we evaluated the applicability of eDNA metabarcoding in determining dietary composition, compared to macroscopic diet identification techniques. Subsequently, we used the dietary information obtained to assess its contribution to biodiversity assessments. Compared to classic macroscopic techniques, we found that eDNA metabarcoding detected more taxa, at higher taxonomic resolution, and proved to be an important technique to verify the species identification of the predator from field collected faeces. Furthermore, we showed how dietary analyses complemented field observations in describing biodiversity by identifying consumed flora and fauna that went unnoticed during field observations. While diet analysis approaches could not substitute field observations entirely, we suggest that their integration with other methods might overcome intrinsic limitations of single techniques in future biodiversity surveys.

Replication levels, false presences and the estimation of the presence/absence from eDNA metabarcoding data.

Environmental DNA (eDNA) metabarcoding is increasingly used to study the present and past biodiversity. eDNA analyses often rely on amplification of very small quantities or degraded DNA. To avoid missing detection of taxa that are actually present (false negatives), multiple extractions and amplifications of the same samples are often performed. However, the level of replication needed for reliable estimates of the presence/absence patterns remains an unaddressed topic. Furthermore, degraded DNA and PCR/sequencing errors might produce false positives. We used simulations and empirical data to evaluate the level of replication required for accurate detection of targeted taxa in different contexts and to assess the performance of methods used to reduce the risk of false detections. Furthermore, we evaluated whether statistical approaches developed to estimate occupancy in the presence of observational errors can successfully estimate true prevalence, detection probability and false-positive rates. Replications reduced the rate of false negatives; the optimal level of replication was strongly dependent on the detection probability of taxa. Occupancy models successfully estimated true prevalence, detection probability and false-positive rates, but their performance increased with the number of replicates. At least eight PCR replicates should be performed if detection probability is not high, such as in ancient DNA studies. Multiple DNA extractions from the same sample yielded consistent results; in some cases, collecting multiple samples from the same locality allowed detecting more species. The optimal level of replication for accurate species detection strongly varies among studies and could be explicitly estimated to improve the reliability of results.