Understanding the factors affecting the dynamics of spatially‐structured populations (SSP) is a central topic of conservation and landscape ecology. Invasive alien species are increasingly important drivers of the dynamics of native species. However, the impacts of invasives are often assessed at the patch scale, while their effects on SSP dynamics are rarely considered. We used long‐term abundance data to test whether the impact of invasive crayfish on subpopulations can also affect the whole SSP dynamics, through their influence on source populations. From 2010 to 2018, we surveyed a network of 58 ponds and recorded the abundance of Italian agile frog clutches, the occurrence of an invasive crayfish, and environmental features. Using Bayesian hierarchical models, we assessed relationhips between frog abundance in ponds and a) environmental features; b) connectivity within the SSP; c) occurrence of invasive species at both the patch‐ and the SSP‐levels. If spatial relationships between ponds were overlooked, we did not detect effects of crayfish presence on frog abundance or trends. When we jointly considered habitat, subpopulation and SSP features, processes acting at all these levels affected frog abundance. At the subpopulation scale, frog abundance in a year was related to habitat features, but was unrelated to crayfish occurrence at that site during the previous year. However, when we considered the SSP level, we found a strong negative relationship between frog abundance in a given site and crayfish frequency in surrounding wetlands during the previous year. Hence, SSP‐level analyses can identify effects that would remain unnoticed when focussing on single patches. Invasive species can affect population dynamics even in not invaded patches, through the degradation of subpopulation networks. Patch‐scale assessments of the impact of invasive species can thus be insufficient: predicting the long‐term interplay between invasive and native populations requires landscape‐level approaches accounting for the complexity of spatial interactions.

Amphibians are an exemplary case of the current biodiversity crisis, being among the vertebrates suffering the fastest decline. Population dynamics of amphibians can result from processes acting at different scales. Both the local characteristics of breeding wetlands and the features of the surrounding landscape can strongly affect the temporal dynamics of amphibian populations. European newts are particularly threatened by land‐use change and invasive alien species. While it is known that newts are declining across Europe, few studies have performed broad‐scale assessments of their decline, either because abundance dynamics are more complex to analyse than presence/absence data or because they require a high sampling effort and long‐term monitoring. In this study, we show that long‐term distribution data can be combined with demographic models to quantify the decline in abundance of newt species at the regional scale, and to assess the importance of multiple factors in determining abundance dynamics. We performed multiple surveys between 1996 and 2020 and used N‐mixture models in a Bayesian framework. We then calculated abundance changes between the first and the last sampling season, which were performed with an average timespan of 13 years across all wetlands. Both Italian crested newts and smooth newts showed large declines, with an average estimated abundance loss between the first and last sampling season of 57% and 63%, respectively. Local characteristics of the wetlands were the main determinants of abundance dynamics: the abundance of both species showed a positive relationship with the area and the permanence of the wetland and a negative relationship with the presence of fish. Additionally, the abundance of Italian crested newts was negatively related to the presence of invasive crayfish. No relationship was detected between abundance and terrestrial habitat availability or connectivity measures. Despite uncertainties in the absolute values of estimated abundance, the striking regional‐scale decline of newts is evident. Among the major determinants of population dynamics, fish and crayfish presence increased their prevalence in the study area, while other factors remained more stable. Management actions aimed at eradicating or controlling invasive fish and crayfish might halt abundance loss and even revert this declining trend. The application of N‐mixture models to long‐term data from representative sites permits the analysis of temporal trends of species at the regional scale even when data come from complex monitoring schemes. We found large declines in abundance of two newt species, suggesting that European newts may be more threatened than previously thought.

No abstract

No abstract