The Macaronesian region: a history of its plant diversity through time

Oceanic islands are important, dynamic ecosystems with high endemism rates, making them natural laboratories for understanding how diversification, dispersal, and disturbance shape biotas. Within this global pattern, the North Atlantic region of Macaronesia, comprised of the archipelagos of the Azores, Madeira, Canary Islands, and Cape Verde, is a perfect model system to investigate how island properties, natural climatic variations, and recent human impacts interact to shape insular floras. Yet, sharp contrasts exist among major land-plant lineages: while Macaronesian spermatophytes exhibit high rates of endemism, bryophytes display the lowest insular endemism rates reported across land plants. This thesis aims at identifying the ecological and evolutionary mechanisms underlaying such strikingly different patterns of endemism and to evaluate how climate and humans have influenced the demographic trajectories of endemic insular plants.
Two main questions guide the work: (Q1) which mechanisms explain the contrasted endemicity patterns between bryophytes and spermatophytes in Macaronesia, and what do they reveal about islands as refugia or stepping-stones for continental diversity? (Q2) to what extent have human disturbance and natural climatic cycles shaped demographic histories of Macaronesian endemic plant species? These questions are approached through (i) a synthesis and re-analysis of published phylogeographic and biogeographic evidence across land-plant lineages, (ii) targeted phylogeographic analyses of broadly distributed (amphi-Atlantic) bryophytes to evaluate colonization routes and the timing of island-continent exchanges, and (iii) population genomic inferences of demographic change in endemic Canarian and Macaronesian vascular plants from the two historically most disturbed ecosystems, namely the thermo-sclerophyllous woodland and laurel forest.
To address the question of the lower endemism rates of bryophytes as compared to spermatophytes, the thesis evaluates three non-exclusive processes: (1) intrinsically low evolutionary capacity, (2) reduced island speciation due to high connectivity and/or differences in speciation modes, and (3) the loss of endemic status via extinctions and/or subsequent (back-) colonization of mainland areas. The evidence assembled here challenges the simplest “high dispersal prevents speciation” view. Our results show that island and mainland bryophyte populations often retain measurable genetic structure and, in a substantial subset of species, a significant phylogeographic signal consistent with long-term isolation and incipient divergence. Similarly, coalescent-based assessments of migration scenarios do not systematically favor high migration rates, implying that chance long-distance dispersal potential does not necessarily translate into frequent successful migration. In contrast to spermatophytes, where a limited number of iconic radiations generated a large proportion of the island endemic flora, bryophyte insular endemics are found to arise mostly through allopatric speciation, with little evidence for subsequent in situ diversification. This difference suggests that endemism patterns reflect not only dispersal and isolation, but also differences in how often islands generate radiations versus single, geographically structured lineages.
A key outcome of this thesis is that bryophyte endemism rates may substantially underestimate island-driven diversification because bryophyte lineages that originate (or diversify) on islands can more easily expand on other territories. Evidence consistent with Macaronesian origins was found for different bryophyte species distributed across islands and mainland areas, and accounting for such cases would raise Macaronesian bryophyte endemism rate well above the known insular endemism percentage. This supports a dynamic view of islands as contributors to continental biodiversity through back-colonization and de novo colonization, potentially facilitated by shifts in wind regimes during glacial periods and by temporary stepping-stones such as seamount chains known as Palaeo-Macaronesia.
Regarding demographic history, population effective sample size reconstructions of endemic plants reveal strong temporal heterogeneity but generally place major expansions or contractions well before the onset of human occupation in the Canary Islands (~2000 years ago), suggesting limited detectability of recent anthropogenic impacts in extant genomic variation. The clearest exception is a localized population decline globally concomitant with the earliest human settlement of people from Berber/Amazigh origin. Across species, signals more often align with pre-human climatic dynamics, including post Last Glacial Maximum changes and mid-Holocene variability (notably the African Humid Period), while differences among taxa are plausibly mediated by life-history traits and dispersal ecology. These results also highlight the hypothesis of a genetic extinction debt, whereby recent habitat disruption does not always translate into an immediate genetic signature, even in insular settings.
Overall, this thesis concludes that Macaronesia is a system shaped by both diversification and migrations. Spermatophyte endemism largely reflects island radiations and the retention of narrow-range lineages, whereas bryophyte endemicity is reduced not only by dispersal, but by the combination of limited in situ radiations and comparatively frequent post-speciation (back-) colonization of continents. Demographic histories of Macaronesian and Canarian endemic plants largely track Quaternary climate dynamics, while recent anthropic perturbations may have weaker, more localized, or genetically delayed impacts. Together, these findings caution against interpreting island endemism as a universal proxy for speciation and strengthen a dynamic view of islands as both refugia and sources of biodiversity at broader spatial scales.