Animal seed dispersal and its consequences for plant recruitment
Thesis DisciplineBiological Sciences
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
Seed dispersal in New Zealand relies on a different guild from the one present before the arrival of humans into the islands. Post-dispersal biotic and abiotic filters have also changed due to the introduction of seed predators and herbivores. The aim of this thesis was to evaluate the actual contribution of seed dispersers to the patterns of seed fall and to understand the effects of biotic (seed density, predators, distance from conspecifics) and abiotic (pH, canopy openess, etc) factors on seed and seedling survival. I expected that tree species with different fruit sizes would be selected by different assemblages of dispersers, therefore being deposited and recruiting differentially on the forest floor.
I analyzed the changes in small-scale seed diversity with the use of seed traps during two seasons in two 0.36 ha plots in a mixed broadleaf-podocarp temperate forest. The diversity of seeds was lower for in-fruit seeds (1.40 species per grid-cell) than for bird-dispersed seeds (2.62). The dominant species D. cupressinum was dispersed towards con- and heterospecific canopies (wind-dispersed trees Weinmannia racemosa and Metrosideros umbellata), giving a probable mechanism for the observed increase in finescale seed diversity. Then, using a 12-year dataset on seed fall I analyzed the changes in ground-level seed heterogeneity generated by bird-dispersed seed rain (propagule-scale) relative to the canopy composition (adult-scale) on six tree species. A Non Metric Multidimensional Scaling analysis demonstrated that the differences in species composition between sampling points (seed traps) were larger for tree canopies than for dispersed seeds. Different tree species acted as seed importers and/or exporters, although these patterns were not explained by fruit size or the richness of the disperser assemblages.
I also investigated the effects of seed and seedling density on seed predation, seedling emergence and seedling survival to 12 months of five tree species. Seed predation increased with seed size, and it was higher at high density only for the largest-seeded species Beilschmiedia tawa and Prumnopitys ferruginea. Access of seed and seedling predators to the propagules was the main factor affecting seedling emergence and seedling survival. Abiotic variables were important for germination and seedling survival of the two small-seeded species, but only biotic predictors (mammalian exclosure and density) were important for medium and large-seeded species. Finally, using a split-plot experimental design I measured the effects of (a) the presence of conspecific canopy, (b) high conspecific seed densities and (c) introduced (or invasive) mammal access on seedling emergence and seedling survival (Janzen-Connell effects). Distance effects were stronger than density effects, and mediated by native and exotic predators. Final survival under conspecifics was strongly reduced in both species, with zero survival of B. tawa under conspecifics, and near zero (2/896) under D. dacrydioides.
Seed dispersal service in New Zealand is still enough to produce changes in the spatial pattern of seed deposition in the forest. Post-dispersal processes such as seed predation and herbivore are having strong impacts on the survival and potential recruitment of plant species. Large-seeded trees are the most affected in terms of the impact of introduced mammals on seed predation, but seedling survival is affected independent of seed size. However, it is unknown how different these ecological processes are from the original situation without pests in New Zealand. The results of my research can contribute with valuable information for future research trying to understand the consequences of seed dispersal for recruitment, as well as inform complex models that try to predict the long-term impacts of the exotic mammals introduced in New Zealand.