The influence of season, hunting mode, and habitat specialization on riparian spiders as key predators in the aquatic-terrestrial linkage

Abstract

Freshwater ecosystems subsidize riparian zones with high-quality nutrients via the emergence of aquatic insects. Spiders are dominant consumers of these insect subsidies. However, little is known about the variation of aquatic insect consumption across spiders of different hunting modes, habitat specializations, seasons, and systems. To explore this, we assembled a large stable isotope dataset (n > 1000) of aquatic versus terrestrial sources and six spider species over four points in time adjacent to a lotic and a lentic system. The spiders represent three hunting modes each consisting of a wetland specialist and a habitat generalist. We expected that specialists would feed more on aquatic prey than their generalist counterparts. Mixing models showed that spiders' diet consisted of 17-99% of aquatic sources, with no clear effect of habitat specialization. Averaged over the whole study period, web builders (WB) showed the highest proportions (78%) followed by ground hunters (GH, 42%) and vegetation hunters (VH, 31%). Consumption of aquatic prey was highest in June and August, which is most pronounced in GH and WBs, with the latter feeding almost entirely on aquatic sources during this period. Additionally, the elevated importance of high-quality lipids from aquatic origin during fall is indicated by elemental analyses pointing to an accumulation of lipids in October, which represent critical energy reserves during winter. Consequently, this study underlines the importance of aquatic prey irrespective of the habitat specialization of spiders. Furthermore, it suggests that energy flows vary substantially between spider hunting modes and seasons.

Figure 1

Areal perspective of the sampling site (A, not to scale) together with the location within Germany (B). Aquatic and terrestrial sampling sites are presented in dark blue and red, respectively. A compass and measure are included in the legend to provide cardinal direction and an approximate scale, respectively. Panel A was generated in Affinity Photo (1.10.6) and panel B used the R packages “ggplot2” (3.4.4) and “raster” (3.6.26).

Figure 2

50% (light), 75% (medium), and 95% (dark) highest density intervals of the posterior distribution of aquatic proportion to spiders’ diet. Black lines show the maximum a posteriori probability and white dots the median of the posterior distribution. The seasonal data is shown for ground hunters, web builders, and vegetation hunters in the riparian system adjacent to the pond and the stream. For each feeding type, one species is considered a wetland specialist (i.e., Tetragnatha extensa, Pirata piraticus, and Dolomedes fimbriatus) while the other is considered a habitat generalist (i.e., Tetragnatha montana, Pardosa amentata, and Pisaura mirabilis). If the model could not be fitted, no crossbar is shown.

Figure 3

Stable isotope signatures of spiders’ opisthosomata (points with solid hull) and prey corrected for trophic enrichment (areas). Data is separated for each month (columns) and system (rows). To improve readability, Collembola are indicated by a red arrow if the data is hardly visible.

Figure 4

Bootstrapped means with 95% confidence intervals of C:N ratios of spiders’ opisthosomata in April (circles), June (squares), August (diamonds), and October (triangles).