Introduction
Arthropod species living in wetlands often face more or less regular periods of flooding, during which the habitat is temporally submerged, compromising the life of terrestrial organisms (Batzer and Wu 2020, Plum 2005). Terrestrial species inhabiting these shifting habitats have developed a range of adaptations that allow them to persist despite the regular disturbances (Kolesnikov, et al. 2012, Lafage, et al. 2015, Marx, et al. 2012, Ramey and Richardson 2017, Rothenbücher and Schaefer 2006). Some species simply withstand the floods until the water recedes, which may require special adaptations to avoid drowning. Other species adapt their life cycle to account for regular flooding regimes or retreat to drier grounds, returning when the water is again gone. Among arthropods, previous studies from large flood plains suggest that plant- and leafhoppers (Auchenorrhyncha) often tolerate prolonged submersion and therefore remain in the wetland during floods, whereas most spiders and ground beetles instead tend to move in and out of wetlands in response to fluctuating water levels (Lafage, Sibelle, Secondi, Canard and Petillon 2015, Rothenbücher and Schaefer 2006). However, even among beetles and spiders, some species are well able to survive at least short-term flood events (Kolesnikov, Karamyan and Hoback 2012).
The occurrence of topographical heterogeneities including vegetation that create local dry ground refuges during floods likely influences species survival and recolonization ability. For instance, wetlands often contain a mixture of hummocks and hollows, with the former providing arthropods with safe ground during high-water events. These small-scale heterogeneities have seldom been considered in studies of local arthropod communities and, are in fact rarely quantified even in hydrological studies (but see Diamond, et al. 2021). A study by Datry et al. (2014) demonstrated that heterogeneities in flood disturbance at a larger scale within a flood plain can significantly affect arthropod communities by creating spatial differences in plant successional patterns. However, the two only studies known to us that combined a quantification of small scale heterogeneities in a large wetland and arthropod communities are Åhlén et al. (2023), which quantified inundation frequencies that reflect local wetness-dryness at a 2x2 m2 resolution and used SLAM (sea, land, air – Malaise) traps for arthropod quantification, and Åhlén et al. (2024). While other studies have incorporated flood dynamics, they generally do so at a coarse scale (but see O’Callaghan, et al. 2013 for an example from the riparian zone of a large river). These studies have documented both positive and negative relations between the inundation frequency and arthropod densities. For instance, Åhlén et al. (2023) found that groups such as leafhoppers were more abundant in frequently flooded areas, whereas dance flies were more prevalent in drier parts of the wetland. This study was however biased by the choice of sampling method. SLAM-traps, like all Malaise traps, primarily capture flying insects and are therefore less suited for quantifying the densities of cursorial species such as spiders and beetles, that are less able to recolonize after disturbances.
In wetlands, spiders and predatory beetles commonly occur at high densities, utilizing the high prey production in these habitats (Ramey and Richardson 2017). Many previous studies have also explored mechanisms underlying community responses of both spiders and predatory beetles, showing the prime importance of the hydrological regime at larger spatial scales (Bonn, et al. 2002, Cartron, et al. 2003, Gerisch, et al. 2006, Jachertz, et al. 2019, Sienkiewicz and Zmihorski 2012, Uetz, et al. 1979). In fact, it has been suggested that the flooding regime is more important than habitat management for both spiders and carabid beetles (Lafage and Petillon 2016). There are also potential differences in responses between spiders and beetles, where these communities are filtered by different environmental factors in wetlands (Åhlén, Jarsjö, et al. 2024). For instance, Lambeets et al. (2008) showed that spider densities decreased strongly with all types of flooding whereas carabid beetles seemingly benefit from intermediate flooding regimes. Their study however used pitfall traps which bias spider catches towards wolf spiders, and carabid beetles, and provide poor density estimates of most other spider groups (Amiar, et al. 2023), and of small staphylinid beetles. Nevertheless, other studies similarly show that carabid beetles are often good indicators of local flooding regimes (Gerisch, Schanowski, Figura, Gerken, Dziock and Henle 2006, Moran, et al. 2012), whereas spider communities seem to be more determined by habitat structure (Bonn, Hagen and Wohlgemuth-Von Reiche 2002, Åhlén, Jarsjö, Jonsell, Klatt, Schneider, Strand and Hambäck 2024), which may directly or indirectly be affected by flooding regimes.
In this study, we first identified sites with varying inundation patterns in a large wetland based on the quantification of small-scale inundation patterns from Åhlén et al. (2023). We then estimated the density and species diversity of beetles and spiders in these sites using suction sampling, which reflect total communities better than pitfall traps, to connect arthropod communities and inundation patterns. We selected sites along two gradients, with wet or dry local (2x2 m2, centered on the arthropod collection point) conditions and wet or dry conditions at a larger spatial scale (6x6 m2). To further identify relevant spatial scale of responses, we calculated inundation patterns also at larger distances from the arthropod collection sites. In addition, we measured local habitat characteristic such as vegetation height and soil carbon content that are known to affect arthropod communities in wetlands (Ye, et al. 2024). For an overview of the methods see Figure 1. Our prediction was that arthropod densities in locally very wet sites depend on the presence of dry sites nearby, whereas the opposite pattern would be less important.