Metabarcoding insights into the diet and trophic diversity of six declining farmland birds

Abstract

Knowledge of feeding ecology of declining species, such as farmland birds, is essential to address their conservation requirements, especially when their habitats are suffering important reductions of trophic resources. In this study, we apply a metabarcoding approach to describe the diet composition of six of the most significant farmland birds inhabiting European cereal pseudo-steppes: little bustard, great bustard, pin-tailed sandgrouse, black-bellied sandgrouse, red-legged partridge, and common quail. We further studied seasonal diet variations (autumn to spring) in all species but the common quail, whose diet was studied during spring and summer. We show that study species´ diets mostly consisted of plants, although in the case of little bustard and great bustard arthropods are also highly relevant. Among arthropods, we found high proportions of thrips, arachnids, and springtails, which were previously unreported in their diet, and some taxa that could be used as antiparasitic food. Moreover, we report that little bustard's diet is the least rich of that of all studied species, and that diet of all these species is less diverse in winter than in autumn and spring. Diet composition of these declining species supports the importance of natural and semi-natural vegetation and landscape mosaics that can provide a wide variety of arthropods, plants, and seeds all year-round.

Figure 1

Graphical illustration of the results of the NMDS analysis of the studied birds’ diet (OTUs). Larger dots represent the centroid of the flocks of each species, while smaller dots represent the individual flock data of each species.

Figure 2

Mean percentage of reads (a measure of abundance) and OTUs (a measure of richness) of the Chloroplastida, Arthropoda, Mollusca, and Annelida taxa present on the diet of studied farmland bird species. Percentages > 10% are specified numerically and in the case of reads, SD is indicated in brackets. In the OTUs graphs, OTUs richness is also indicated. Note that samples of great bustard, little bustard, pin-tailed sandgrouse, black-bellied sandgrouse, and red-legged partridges were collected during the non-breeding season, whereas samples of common quail were collected during the breeding season.

Figure 3

OTUs richness in the diet of studied species. The whiskers represent the 95% confidence intervals. The statistical significance of the Tukey post-hoc test is provided by * symbols. ***P < 0.001. **P < 0.01. *P < 0.05. The colour of the * corresponds to the colour of the species with which the comparison is made. For those comparisons that were not statistically significant, no significance data are provided. For detailed statistical results of Tukey post-hoc tests, see Supplementary Table S2 online.

Figure 4

Mean percentage of reads (a measure of abundance) and OTUs (a measure of richness) of each phylum of Arthropoda present on the diet of studied farmland bird species, with phylum Insecta data broken down into orders. Percentages > 10% are detailed numerically. Among Insect’s orders, those with a percentage > 2% are identified. Note that samples of great bustard, little bustard, pin-tailed sandgrouse, black-bellied sandgrouse, and red-legged partridges were collected during the non-breeding season, whereas samples of common quail were collected during the breeding season. The average number of OTUs per species and diet taxa is provided in Supplementary Table S3 online.

Figure 5

Graphical illustration of the NMDS base on diet OTUs of little bustard, great bustard, pin-tailed sandgrouse, and red-legged partridge represented by season. Large symbols represent the centroid of the flocks of each species in each season, while small dots represent data of each different flock per species and season. The circles group data of each species.

Figure 6

Mean percentage of reads (a measure of abundance) and OTUs (a measure of richness) of the Chloroplastida, Arthropoda, Mollusca, and Annelida taxa present on the diet of great bustard, little bustard, pin-tailed sandgrouse, and red-legged partridge in autumn, winter and spring. Percentages > 10% are detailed numerically and in the case of reads, SD is indicated in brackets. In the OTUs graphs, OTUs richness is also indicated. For great Bustard and red-legged Partridge, only one flock was sampled in winter so no SDs are provided.

Figure 7

Mean percentage of reads (a measure of abundance) and OTUs (a measure of richness) of each phylum of Arthropoda present on the diet of great bustard, little bustard, pin-tailed sandgrouse, and red-legged partridge in autumn, winter, and spring, with phylum Insecta data broken down into orders. Percentages > 10% are detailed numerically. Among Insect’s orders, those with a percentage > 2% are identified. For great bustard and red-legged partridge, only one flock was sampled in winter so no SDs are provided.

Figure 8

Map of Spanish regions with the spatial distribution of each species samples. The number of flocks per species sampled in each region is shown in brackets. This map was generated using QGIS v3.10.