Long distance dispersal by migratory birds : from processes to patterns of biodiversity in aquatic ecosystems

Programa de Doctorado en Estudios MedioambientalesBiological species are distributed across the earth's surface according to their ability to disperse and the set of environmental conditions that fulfil their ecological niche, which limit their potential dispersal range.For species lacking the ability to move by themselves, such as plants, dispersal must be mediated by a vector that transports them. In inland aquatic ecosystems many taxa that produce propagules, such as macrophytes and invertebrates, rely on a variety of dispersal vectors for overland dispersal, i.e. between isolated waterbodies. Sorne of these vectors may disperse propagules over long distances (long distance dispersal,LOO), depending on their nature or movement behaviour.In particular, migratory waterbirds have been shown to disperse large amounts of propagules, but their potential to disperse propagules during active migration has been poorly studied. Furthermore, realistic estimates of long distance propagule dispersal had never been produced, because, besides flight speed, waterbird movement behaviour had never been incorporated into dispersal models. In Chapter 1 we developed mechanistic models of passive dispersal of macrophytes and zooplankton that integrated (i) movement data (based on recoveries of ringed individuals) of two waterfowlspecies, collected for over a century during the fall migration season,(ii) propagule retention time in the birds¿ gut, and (iii) bird flight speed. 8oth waterfowl species had the potential to disperse propagules over hundreds of kilometres, but the frequency and distance of dispersal events, namely those of LOO events (i.e. those produced by migratory movements), differed between migratory strategies and regions (Europe and North America}, as well as between different vectored aquatic species, in particular between plant seeds and zooplankton cysts. Although dispersal distances were shorter than those based only on the vector flight speed, they suggest that the propagules of aquatic organisms may be dispersed regularly over tens of kilometres and occasionally over hundreds of kilometres (i.e. at continental scale). Different migratory vectors might then contribute differently to the dispersal of a given aquatic organism (because of, among other factors, different migratory behaviours; Chapter 1), thus its full dispersal potential has to be estimated according to the full array of vectors.In Chapter 2,we estimated propagule dispersal by migratory birds by using a wide range of both waterbird and vectored aquatic species, and determined key features of bird movement and propagule retention time involved in the dispersal process (by means of a sensitivity analysis).Bird size scaled to these key features (related to migratory flight distance and propagule retention time), providing a simple analytical tool to investigate the potential of each migratory bird species as vector of long distance propagule dispersal. Moreover, the inclusion of several waterbird species and the examination of some traditional model assumptions allowed to estimating the range of LOO frequency (0-3.5% of dispersal events longer than 100 km). Nevertheless, empirical and direct evidence on propagule dispersal by migratory vectors during active migration had never been obtained. We investigated whether migratory birds flying southward, between Europe and Africa, during autumn migration were carrying propagules in their digestive tract (Chapter 3). This study was conducted in a small islet of the Canary lslands (Alegranza) where Eleonora's falcons prey upon migratory birds along their migration route (though blown off their usual course by trade winds). All migratory birds were caught while in migratory flight over the sea, in an area between Alegranza and the African coast, as proven by the analysis of locations provided by GPS-tagged falcons. We sampled the digestive tracts of 408 migratory birds stored around the falcons' nests and found that 1.2% of these birds (two passerine and one galliform species) were transporting seeds (57 seeds of five different families) while in migratory flight. Most seeds belonged to plant families currently absent from the Canary lslands, or at least from those nearby Alegranza (the Eastem group). We therefore conclude that migrating birds are actually able to ensure long-distance dispersal of ingested propagules,even across continents or into oceanic islands. Moreover, the absence of plant species dispersed by migratory birds from either the whole Canary lslands or the closest (and driest) ones indicates that the local environment (abiotic and/or biotic) probably hamper their successful establishment. We then investigated whether the high dispersal potential of organisms dispersed by migratory birds is reflected in their communities by quantifying the effect of spatial (i.e. regional) processes,representing dispersallimitation, on diversity patterns. Using variation partitioning analyses, we assessed the relative importance of spatial and niche processes in driving species richness (¿-diversity;Chapter 4) and variation in species composition (-diversity; Chapter 5) in natural communities of aquatic plants (aquatic angiosperms) and cladocerans (crustacean zooplankton) at local, regional (within-region: up to 300 km) and continental (among-region) scales.In Chapter 4 the lack of spatial correlation among the species richness of the sampled lakes after partialling out the effect of environmental variation suggested weak dispersal limitation within regions. Although some connectivity-related variables (amount of nearby habitat) had a significant effect, the environment had a much stronger effect on species richness, particularly the total phosphorus concentration (a surrogate of productivity in many aquatic systems). Region was the most important factor in explaining variation in species richness, which indicates an important effect of biogeographic factors such as the number of unique species found in each region (i.e. the levelof "endemicity" in the study). Variation in community composition was mostly explained by species turnover rather than by gradual species loss (i.e. nestedness), both within and among regions {Chapter 5).In addition, the spatial structure of lakes within regions, including the distance among them, did not influence the variation in community composition, meaning that these communities are not generally limitad by dispersal. Rather, both biotic and abiotic variables, and to a lesser extent connectivity variables, explained the majority of the variation in community composition. Much of this environmental variation was regionally structured,in particular climatic variables (temperatura and precipitation), which resultad in high variation in community composition among regions. The pure effect of region was also significant, providing evidence for biogeographic patterns ¿ which might be reinforced by dispersal limitation at continental scale. Although spatial and niche processes could explain many diversity patterns, we observed an extremely high species turnover at short spatial scales.even among nearby lakes where environmental conditions are more similar and dispersal limitation is almost inexistent. In Chapter 6 we investigated experimentally whether the resident community of either macrophytes or zooplankton could prevent immigration, through priority effects (i.e. whether the time of arrival influenced immigration success) and diversity resistance (i.e. whether increased resident diversity caused less immigration); and whether these effects could ultimately determine the species richness and composition of such communities. Strong priority effects were observed in the assemblages of both taxa. In macrophyte communities, increasing resident diversity along the growth season reinforced priority effects (i.e. increased diversity resistance), whereas strong priority effects exerted by the founder organisms (since the very beginning of community assembly) constrained the invasibility of zooplankton communities. Overall, 1 conclude that migratory movements of birds can mediata long distance propagule dispersal (over distances of more than 100 km) and influence the community ecology and biogeography of vectored organisms. Different migratory bird species, acting as propagule vectors, can provide different but complementary dispersal services depending on their body size and in turn on their migratory strategy. Waterbirds, in particular,have the potential to disperse propagules regularly over tens of kilometres and occasionally (up to 3.5% of the dispersal events) over more than 100 km. In fact, vectored aquatic organisms are not generally limited by dispersal, although connectivity-related factors might affect species diversity in some regions. lnstead,niche processes largely determine ¿- and -diversity through abiotic and biotic filtering. Dispersal limitation might only be detectable at continental scale and thus reinforce compositional heterogeneity at broad scale (i.e. promote biogeographic patterns).Finally, priority effects (which, depending on the taxon,might be reinforced by diversity resistance) might explain the high species turnover observed at short spatial scale.In combination with long distance propagule dispersal,priority effects might explain the "mosaic" distributional patterns of aquatic communities. This thesis provides a general framework that can be used for hypotheses testing relating LOO mediated by migratory birds to its hypothesised ecological and biogeographical consequences.Universidad Pablo de Olavide. Centro de Estudios de Postgrad