A new study based on 12 years of standardized monitoring in the native forests of Terceira Island shows that the absence of obvious species loss does not necessarily mean that forest communities are stable. Using long-term data from the SLAM project, researchers found that while native and non-native arthropod assemblages did not show systematic declines in species richness, the abundance structure of native communities changed through time, with many previously abundant native species becoming progressively rarer.

Published in Ecology, the study analyzed arthropods sampled in ten 50 × 50 m plots distributed across four fragments of Terceira’s native forest, using SLAM (Sea, Land and Air Malaise) traps operated seasonally from 2013 to 2024. The work focused separately on three groups: native endemic species, native non-endemic species, and non-native species.

The researchers found that presence–absence data alone suggested relative stability across the study period, with no systematic species losses in any of the three groups. However, when abundance patterns were analyzed, a different picture emerged: native endemic and native non-endemic assemblages showed increasing Bray–Curtis dissimilarity through time, indicating that community structure was shifting even though total species numbers remained broadly similar.

In contrast, non-native arthropod assemblages followed a different temporal pathway. Their patterns were more consistent with neutral-model expectations, suggesting that stochastic colonization, local extinction, and repeated re-colonization from surrounding human-modified habitats are important drivers of their dynamics. The paper interprets this as evidence for source–sink mass-effect dynamics linking native forest fragments with the wider disturbed landscape around them.

The study also has an important conservation message: biodiversity change on islands may occur as a gradual reorganization of communities before clear species losses become visible. For this reason, the authors argue that monitoring programs should not rely only on species counts, but should also track abundance structure and long-term temporal dynamics.

Terceira Island is an especially important place to detect these patterns. It contains the largest remaining area of native pristine forest in the Azores, yet the archipelago has already lost about 95% of its original native forest cover since human settlement. This makes the remaining fragments both highly valuable and highly vulnerable to long-term pressures associated with land-use change, biological invasions, and other environmental shifts.

This new paper builds on earlier SLAM-based work showing that non-native arthropods in Azorean native forest often have higher temporal turnover than natives, that exotic diversity has increased in the system, and that long-term monitoring is essential for detecting subtle but ecologically important changes in island communities.

The authors emphasize that protecting island biodiversity will require continued, standardized, long-term monitoring capable of detecting these hidden changes early enough to guide conservation action. The SLAM project continues to provide one of the most important long-term datasets for understanding how Azorean arthropod communities respond to environmental change.

Article
Zhang, Y., Borges, P.A.V., Lhoumeau, S., Matthews, T. & Liao, J. (2026). Divergent temporal dynamics of native and non-native insular arthropods in fragmented forests. Ecology, 107: e70363. https://doi.org/10.1002/ecy.70363