Introduction
Islands are dynamic entities with continuously evolving geographical settings influencing the distribution and genetic evolution of organisms they host. At deep time scales (Ma), an island’s ontogeny is controlled by geological processes (Whittaker et al. 2008), while on shorter time scales climatic processes dominate. Climate drives major sea-level fluctuations causing islands to shrink and expand, fragment and merge, or even disappear and emerge (e.g. Simaiakis et al. 2017). Late-Quaternary sea-level changes have left their imprint on insular species diversity patterns (Ali and Aitchison 2014, Rijsdijk et al. 2014, Weigelt et al. 2016, Ávila et al. 2018, 2019, Norder et al. 2019). All of these did however focus on volcanic oceanic islands, or did not make a distinction between oceanic volcanic and continental insular systems and lumped them in their analyses (Weigelt et al. 2016, Veron et al. 2019). By disregarding the specifics of continental archipelagos, relevant biogeographical processes are potentially overlooked (Ali 2017, Simaiakis et al. 2017, Whittaker et al. 2017).
Continental shelf islands (sensu Ali 2017) are characterized by sitting on a continental shelf and may include “land-bridge” islands, that were connected to the mainland in the past when sea levels were lower. Another characteristic is the proximity of the mainland often surrounding continental islands (Weigelt and Kreft 2013). The geo-spatial effects of sea-level dynamics on continental island biota have been studied extensively (e.g. Diamond 1972, Cardillo et al. 2008, Itescu et al. 2020). The Aegean archipelago in the Mediterranean Sea is located between the Greek and the Anatolian peninsulas, and is one of the largest archipelagos on Earth (Blondel et al. 2010). Its complex geological history and high environmental heterogeneity contribute to its high biodiversity and endemism, thus rendering an ideal stage for biogeographical studies (Strid 2016). The Aegean islands form a typical continental shelf archipelago with a high geological heterogeneity and numerous land-bridges due to sea-level change. Unsurprisingly, its biogeography has been studied intensively (e.g. Triantis et al. 2018). Although providing crucial biogeographical insights, most studies remain limited in scope by focusing on a single explanatory factor of interest: i.e., by not distinguishing between native and endemic species (Itescu et al. 2020), by considering a single species group or a limited part of the archipelago (e.g. Panitsa et al. 2010). The combined effect of area change, fragmentation and connectivity driven by sea-level changes on insular species diversity has never been investigated for the Aegean archipelago; while we think that such a combined analysis is crucial to interpret the relevant biogeographic processes correctly (see Kougioumoutzis and Tiniakou 2014, Norder et al. 2019). Moreover, the influence of continental land-bridge islands on biodiversity patterns in different chorotypes as reflected in multivariate models and Island Species-Area Relationships (ISARs) has never been quantitatively assessed (e.g. Triantis et al. 2008, 2012). Since we have recently quantified the paleogeographic change of islands in the Aegean Sea (Simaiakis et al. 2017) and its biota are well studied, this setting represents an ideal study system to test our hypotheses.
Our aim is to investigate the combined impact on continental insular species diversity of current as well as past island area and connectivity for the Aegean archipelago for two well-studied taxonomic groups, angiosperms and centipedes. Our first hypothesis (H1) is that more native species occur on land-bridge islands than on true islands, reflecting the higher establishing rates of native species on land-bridge islands (Simaiakis et al. 2017). Our second hypothesis (H2) is that endemism is negatively influenced by land-bridge islands, as allopatric speciation is suppressed by repetitive gene-flow from the continents. Our third hypothesis (H3) concerns the effect of the duration of the archipelagic configurations as a result of sea-level drop affecting richness patterns. On oceanic islands, the median archipelago configuration could largely explain the endemism richness patterns than the extreme and short-lasting Late Glacial Maximum (LGM) configuration (Norder et al. 2019). Thus on land-bridge islands, the long-lasting median glacial configuration may have largely influenced the species richness patterns, rather than the short lasting LGM extreme state, assuming that short-lasting spatial configurations do not provide the time needed for speciation to occur. To test our hypotheses, we analyse the effects of land-bridge islands on species richness of different endemism levels. We investigate H3 by analysing species richness patterns in relation to two paleogeographical settings: 1) during LGM, and 2) the median sea level during the last glacial-interglacial cycle (Figure 1).