Population genetic structure and phylogeography of two important Neotropical tree species: Vochysia ferruginea Mart. & Cedrela odorata L.

Abstract

Tropical forest is being cleared and fragmented at a rapid pace. Unsustainable forestry practices are eroding natural resources where they are badly needed and contributing to the decline of global biodiversity. Increasingly, efforts are being made to address the problem, through conservation of remaining forest and restoration of areas of degraded land. However, for policies to be successful in the long term, it will be necessary to understand the level and distribution of natural variation within forest species and the effect that human impacts may have on its maintenance. This thesis aims to use molecular methods to investigate the genetic diversity and population structure of two commercially important neotropical tree species in a human-influenced Central American landscape.

Red Yemeri, Vochysia ferruginea Mart., is recommended as a potentially highly useful species for forestry reclamation of degraded land. Populations from throughout its range in Costa Rica were analysed for both organellar (chloroplast specific PCR-RFLP) and total genomic (AFLP) variation. All populations were fixed for one of two chloroplast DNA (cpDNA) haplotypes identified. One haplotype was distributed exclusively throughout the Atlantic watershed, whilst the second was found only on the Pacific slope. AFLP diversity was partitioned primarily within populations (80.5%) and small but significant between-population differentiation was detected (®st = 0.195). The relationships between populations fitted an isolation by distance model when an ecological distance metric linking populations through suitable habitat was used, and indicated gene flow around a central mountain range. In combination with the cpDNA data, these data suggest that pollen flow around the mountain range is maintained whilst seed flow appears to be more restricted and cpDNA structure is more probably a consequence of historical population processes.

A Central American rangewide study of cpDNA variation in C. odorata identified a total of five haplotypes. These formed three groups reflecting geographic distribution, representing Northern-, Central- and Southern-Central American populations. However, the genetic relationships between groups did not follow the geographic pattern. The Northern group was most strongly differentiated from the other two, although it is morphologically similar, geographically proximate and occurs in similar habitat to the Central type. This divergence most likely reflects a very ancient colonisation event. In the south, the Central (dry) type and the Southern (wet) group were differentiated along environmental boundaries. Given the distributions of the latter two cpDNA types and taking AFLP diversity into account, it seems most probable that this pattern reflects colonisation and extinction processes that would have accompanied the vegetation and climate changes in the region during the glacial cycles of the last million years.

The data obtained on these two important species are discussed in terms of application to their conservation and sustainable use. In particular, the identification of Evolutionarily Significant Units in C. odorata will help to shape an effective policy for its long term survival.