Still no evidence for disruption of global patterns of nest predation in shorebirds

Many shorebird species are rapidly declining (Piersma et al. 2016; Munro 2017; Studds et al. 2017), but it is not always clear why. Deteriorating and disappearing habitat, e.g. due to intensive agriculture (Donal et al. 2001; Kentie et al. 2013; Kentie et al. 2018), river regulation (Nebel et al. 2008) or mudflat reclamation (Ma et al. 2014; Larson 2017), and hunting (Reed et al. 2018; Gallo-Cajiao et al. 2020) are some of the documented causes. A recent study suggests yet another possible cause of shorebird decline: a global increase in nest predation (Kubelka et al. 2018). The authors compiled an impressive dataset on patterns of nest predation in shorebirds and their analyses suggest that global patterns of nest predation have been disrupted by climate change, particularly in the Arctic. They go as far as to conclude that the Arctic might have become an ecological trap (Kubelka et al. 2018). Because these findings might have far-reaching consequences for conservation and related political decisions, we scrutinized the study and concluded that the main conclusions of Kubelka et al. (2018) are invalid (Bulla et al. 2019a). The authors then responded by reaffirming their conclusions (Kubelka et al. 2019b). Here, we evaluate some of Kubelka et al.s (2019b) responses, including their recent erratum (2020), and show that the main concerns about the original study still hold. Specifically, (1) we reaffirm that Kubelka et al.s (2018) original findings are confounded by study site. Hence, their conclusions are over-confident because of pseudo-replication. (2) We reiterate that there is no statistical support for the assertion that predation rate has changed in a different way in the Arctic compared to other regions. The relevant test is an interaction between a measure of time (year or period) and a measure of geography (e.g., Arctic vs the rest of the world). The effect of such an interaction is weak, uncertain and statistically non-significant, which undermines Kubelka et al.s (2018) key conclusion. (3) We further confirm that the suggested general increase in predation rates over time is at best a weak and uncertain trend. The most parsimonious hypothesis for the described results is that the temporal changes in predation rate are an artefact of temporal changes in methodology and data quality. Using only high-quality data, i.e. directly calculated predation rates, reveals no overall temporal trend in predation rate. Below we elaborate in detail on each of these points. We conclude that (i) there is no evidence whatsoever that the pattern in the Arctic is different from that in the rest of the world and (ii) there is no solid evidence for an increase in predation rate over time. While we commend Kubelka et al. for compiling and exploring the data, we posit that the data underlying their study, and perhaps all currently available data, are not sufficient (or of sufficient quality) to test their main hypotheses. We call for standardized and consistent data collection protocols and experimental validation of current methods for estimating nesting success.