Vegetation structure influences predation rates of early nests in subarctic breeding waders

Ground‐nesting species are vulnerable to a wide range of predators and often experience very high levels of nest predation. Strategies to reduce nest vulnerability can include concealing nests in vegetation and/or nesting in locations in which nests and eggs are camouflaged and less easy for predators to locate. These strategies could have important implications for the distribution of ground‐nesting species and the success rates of nests in areas with differing vegetation structure. However, the factors influencing the success of nest concealment and camouflage strategies in ground‐nesting species are complex. Here we explore the effects of local vegetation structure and extent of nest concealment on nest predation rates in a range of ground‐nesting, sympatric wader species with differing nest concealment strategies (open‐nest species: Oystercatcher Haematopus ostralegus, Golden Plover Pluvialis apricaria and Whimbrel Numenius phaeopus; concealed‐nest species: Black‐tailed Godwit Limosa limosa, Redshank Tringa totanus and Snipe Gallinago gallinago) in south Iceland, in landscapes that comprise substantial variability in vegetation structure at a range of scales. We monitored 469 nests of these six wader species in 2015 and 2016 and ~40% of these nests were predated. Nest predation rates were similar for open‐nest and concealed‐nest species and did not vary with vegetation structure in the surrounding landscape, but nest‐concealing species were ~10% more likely to have nests predated when they were poorly concealed, and the frequency of poorly concealed nests was higher in colder conditions at the start of the breeding season. For concealed‐nest species, the reduced capacity to hide nests in colder conditions is likely to reflect low rates of vegetation growth in such conditions. The ongoing trend for warmer springs at subarctic latitudes could result in more rapid vegetation growth, with consequent increases in the success rates of early nests of concealed‐nest species. Temperature‐related effects on nest concealment from predators could thus be an important mechanism through which climate change affecting vegetation could have population‐level impacts on breeding birds at higher latitudes.

) to determine the predator species 130 active on these nests. Cameras were attached to poles ~10 cm above ground level and 2 m from nests.

131
The cameras were programmed to take ten pictures when triggered with no interval between trigger 132 events and on the highest sensitivity level. 133

134
When each nest was first located, the PERCENTAGE OF EGGS VISIBLE from directly above the nest 135 (observer standing with a leg on either side of the nest and looking down towards the nest cup) was 136 estimated by eye in the field (i.e. the eggs of open-nesting species were predominantly 100% visible).

137
The habitat surrounding each nest was assessed in the field at three spatial scales: the nest cup, the 5 x 5 138 m and the 50 x 50 m area surrounding each nest. The NEST HABITAT of the nest cup was identified (Table   139  1   between nesting type and habitat heterogeneity were included (Table 2). Non-significant (P > 0.05) 159 variables were sequentially removed from these models (although their estimates and associated 160 probabilities in initial maximal models are also reported, for completeness). All models were carried out in 161 R (v 3.4.1) using the lme4 package, with model goodness-of-fit evaluated by inspecting deviance residuals.

162
Daily predation rates (DPR) predicted from these models were then transformed to predation  (Table 3; model ii, Fig. 2 concealed-nesting species were predated throughout the season and at all times of day, and both 177 mammalian and avian predators were captured on camera (Fig. 3, Table S1). Within concealed-nest 178 species, the visibility of nests was significantly greater in 2015 than 2016, and visibility decreased 179 significantly as the season progressed (Table 3;  182 Daily nest predation rates did not vary significantly in relation to the habitat heterogeneity or the extent 183 to which the dominant habitat covered the area surrounding the nest, at either 5 x 5 or 50 x 50 m scales 184 (Table 4). In addition, the dissimilarity between the habitat at the nest cup and in the surrounding area did 185 not influence daily nest predation rates for open-or concealed-nest species (Table 4). Most nests were 186 laid in habitats that were the same as the surroundings (Fig. S4e-h).  we cannot explore any effect of cameras with these data.

215
While predator avoidance appears difficult to achieve for ground-nesting species  (Table 4; models xi-xiv). However, the great majority of nests were 226 laid in habitats that were the same as the surroundings (Fig. S4e-h) Accepted Article predated (Table 3). Our metric of nest concealment is related to visibility from above, but permeability of 233 the surrounding vegetation may also influence predation risk, particularly in relation to mammalian 234 predators. Egg visibility declined through the season in both years, and was consistently higher in the 235 colder year (Fig. 4). This suggests that the onset and rate of vegetation growth could potentially constrain 236 the availability of suitable nesting locations for these species, and influence nest success, particularly

Accepted Article
This article is protected by copyright. All rights reserved   Table 2 for model details).

Accepted Article
This article is protected by copyright. All rights reserved

Accepted Article
This article is protected by copyright. All rights reserved   Table 3; model iii).

Table S1
Outcome of open nesting species with nest cameras   Table 1 for details).

Accepted Article
This article is protected by copyright. All rights reserved