Weak latitudinal trends in reproductive traits of Afromontane forest trees

Is the increase in species diversity patterns towards lower latitudes linked to reproductive traits? Plant reproductive organs influence reproductive isolation and hence species divergence. Abiotic differences between temperate and tropical regions can also directly impact on plant reproductive traits. Here we provide a novel overview of southern hemisphere, Afromontane forest tree taxonomical patterns and ask whether reproductive traits relate to latitude, while accounting for environmental (tree height) and evolutionary (biogeographical affinity) selective forces.

We compiled a novel dataset with (1) flower colour, size and pollination syndrome and (2) fruit colour, size and dispersal syndrome for 331 tree species found in six Afromontane forest regions. We categorized each species into latitudinal distribution using these six regions, spanning the southern Cape (34º S) to Mount Kenya (0º S). Additionally, we gathered maximum tree height (m) for each species and determined the global distribution of all 196 tree genera (Afrotropical, Palaeotropical or Pantropical).

Species, genera and families showed a general decrease in richness away from tropical and subtropical forests towards warm temperate forests. Southern Afrotemperate forests (the furthest south) had the highest tree endemism. There was no relationship between latitude and the reproductive traits tested here. Biogeographical affinity related to fruit colour and dispersal syndrome, with palaeotropical genera showing relative increases in black-purple fruit colour compared with pantropical genera, and palaeotropical genera showing relative increases in biotic seed dispersal compared with Afrotropical genera, which showed higher relative abiotic seed dispersal. Taller trees had a higher chance to be wind or insect pollinated (compared with bird pollinated) and had larger fruits.

Latitude explained patterns in Afromontane tree taxonomic diversity; however, tree reproductive traits did not relate to latitude. We suggest that phylogenetic conservatism or convergence, or both, explain the reported patterns.