• Citer

Propagules and patterns: the population genetic structure of mangroves in the Western Indian Ocean as a result of dispersal and its constraints

par De Ryck, Dennis
Président du jury Hardy, Olivier J.
Promoteur Dahdouh-Guebas, Farid ;Triest, Ludwig
Co-Promoteur Koedam, Nico
Publication Non publié, 2017-09-04

Thèse de doctorat

• Accès en ligne
• Détails
• Contenu
• Statistiques

Résumé :

In this thesis, I have looked at the process of dispersal, its resulting patterns and the consequences or effects on the genetic population structure in mangroves.The intertidal mangrove forest ecosystem with its emblematic stilted trees is known for its many ecosystem services, both ecological and economical. Local livelihoods in particular depend on the benefits of a healthy mangrove ecosystem for food (e.g. fish, shrimps and crabs), firewood and timber. Mangroves are also important for their biodiversity, coastal protection and carbon storage. Mangroves are however declining worldwide mainly as a result of coastal development, aquaculture expansion and overharvesting, resulting in increased fragmentation. Mangroves have a naturally fragmented coastal distribution; further area reduction and fragmentation both have an important effect on connectivity between populations, hence their genetic diversity and structure. Since mangrove trees do not reproduce vegetatively, they solely depend on dispersal of their seeds or seedlings (propagules) by water to regenerate and (re)colonize areas. Knowledge about the dispersal properties of the mangrove propagules is therefore crucial to understand the regeneration process of these forests, to interpret genetic patterns and to draw up suitable conservation plans. First, I focused on the propagule dispersal of the widespread Ceriops tagal and Rhizophora mucronata mangrove species within the mangrove forest through in situ propagule tracking, predation and root-growth experiments. This revealed C. tagal’s strategy of releasing high numbers of propagules and fast dispersal within the dense forest (slender, low density and smaller size). In contrast, R. mucronata has adopted a dispersal strategy of survival where fewer propagules are released. However, they are more resistant to predators due to their larger size and they can anchor themselves faster due to quicker root-growth. C. tagal propagules have thus a theoretical advantage to disperse over longer distances over the thicker, longer and denser R. mucronata propagules. C. tagal propagules have, however, lower establishment chances due to slower root-growth, desiccation sensitivity and smaller size. Overall, the propagule characteristics of both species result in different and alternative dispersal strategies, eventually leading to a similar capacity for long-distance dispersal.A consecutive field study and flume (water-racetrack) experiments were performed to simulate the dispersal process, using mangrove species (R. mucronata, C. tagal, Heritiera littoralis and Xylocarpus granatum) that cover a broad range of propagule types. These experiments indicated that dispersal distances and trajectories of hydrochorous mangrove propagules are, apart from species-specific aspects, not only determined by prevailing hydrodynamic conditions but also by dominant wind forces. The degree to which wind determines a propagule's dispersal path depends on a combination of the propagule's density, floating orientation and morphology. The low density Heritiera littoralis propagules are easily steered by wind forces acting on their dorsal ‘sail’, while they float on the water surface. Quite in contrast, the impact of wind is limited on the deeply submerged Xylocarpus granatum fruits. For more elongated propagules (such as the ones of C. tagal and R. mucronata), the floating orientation turned out to be even more important for dispersal, as vertically floating propagules dispersed further in our release-recapture experiment than the horizontally floating propagules which were blown towards the shore. An infraspecific range of floating orientations allows a species to disperse divergingly depending on the prevailing conditions.The wide distribution of most mangrove species indicates their potential for long-distance dispersal and (re)colonization of suitable habitat. The range can however neither describe the process, its direction, nor the frequency of dispersal events. Resulting patterns in actual populations can however be analyzed. I studied Avicennia marina in much detail, a typical colonizer of new habitats, to understand dispersal as reflected in the genetic structure. I found high connectivity with indirect evidence for historical long-distance dispersal (LDD) events over hundreds, sometimes thousands of kilometers in the central regions of its Western Indian Ocean (WIO) range. If (long-distance) dispersal occurs frequently, homogenization of populations can take place when several environmental conditions are met and ecological hurdles are taken. Niche unfilling because of time constraints, local adaptation, founder effects, asymmetrical dispersal vectors, physical barriers and combinations of these different influencing factors cause populations to show various degrees of differentiation. One clear example was found at the southern range edge of A. marina (South Africa) where a combined influence of factors may explain the pattern of high genetic differentiation with low allelic richness and niche unfilling. Namely, historical sporadic dispersal events (founder effects) with subsequent inbreeding due to high selfing rates and/or sustained dispersal limitation between populations in deep and less accessible river inlets, as well as range-edge effects.Furthermore, a strong genetic break was found between the Seychelles and East Madagascar populations on the one hand and all the other East African populations on the other hand. We hypothesize that Avicennia in the Seychelles and East Madagascar could either be a new species for the WIO (A. germinans or A. sp.), or be an introgressed hybrid based on the genetic patterns I found with the two chloroplast markers (matK and trnH-psbA) and on the observation that the two strongly diverged groups are able to form individuals with more than one copy of the chloroplast markers (double peaks). Experimental checks are however needed to determine the cause of this pattern: biparental chloroplast inheritance or the existence of nuclear copies of plasmid DNA (NUPTs). For example, I could sequence the whole chloroplast genome of the individuals with more than one copy of the chloroplast marker and make use of methylation-sensitive restriction enzymes to extract nuclear DNA and sequence NUPTs if they occur. In conclusion, the novelty of this work lies in the insights it provides into the differential influence of environmental and biological drivers on propagule dispersal as well as species-specific propagule characteristics on the dispersal potential for a range of common mangrove species. Furthermore, an in-depth study on A. marina in the WIO showed it had lower allelic, genetic and genotypic diversity than other regions within its distribution range. The complex signal of divergence on the nuclear and chloroplast level shows the importance of considering all possible inheritance patterns and gene transfers when interpreting phylogenetic results and thus to delineate species correctly. Correct species delineation and range description is paramount for the conservation of species, their genetic diversity and thus the evolutionary potential they harbour to respond to environmental change.