Disrupting Dispersal? Consequences of Seed Harvesting for Natural Regeneration of an Economically Important Tree in the East Usambara Mountains

19 Jan 2011 Amani Nature Forest Reserve, Tanzania, Africa Forests

Carrie Seltzer

Allanblackia stuhlmanni, a rainforest canopy tree, has seeds that are harvested for vegetable oil. The goal of this project is to examine how varying seed density affects seed dispersal and the probability of seedling establishment by experimentally manipulating seeds and following seed fate, with the aim to develop guidelines for sustainably managing seed harvest.

Washed and dried Allanblackia seeds.

Washed and dried Allanblackia seeds.

Seeds of the rainforest canopy tree Allanblackia stuhlmannii (Clusiaceae) have a growing international market for vegetable oil, but intensive harvesting of the seeds by humans may reduce the natural regeneration of Allanblackia over time. When Allanblackia seeds are not collected by humans, they are eaten, cached, or hoarded by squirrels, giant pouched rats, and smaller rodents. Cached and hoarded seeds are intended to be eaten later but forgotten seeds may germinate and subsequently establish. Heavy seed harvesting may alter rodents’ seed dispersal behaviour, which subsequently affects Allanblackia regeneration.

This research will measure how seed abundance of Allanblackia stuhlmannii impacts seed dispersal and regeneration. At the end of the experiment, we will be able to compare seed dispersal and seedling establishment under three levels of seed availability (low, medium, and high densities), the lowest of which mimics heavy harvesting by humans. If low seed availability significantly depresses establishment, then policies will be recommended to prevent over-harvesting within protected forests.

Allanblackia seed density will be experimentally manipulated in designated plots by removal or addition of seeds. There will be three different treatment levels: control seed density (no fruits added or removed), low seed density (all fruits removed regularly), and high seed density (fruits removed from low density plot will be added to this one). To compare rodent seed dispersal, arrays of seeds tagged with thread and flagging tape will be monitored by cameras record the identity of seed dispersers. When seeds are moved, the new location of the seed will be determined by following the thread. Seed fate will be determined by regularly checking tagged seeds and recording seed fate as rotten, cached, eaten, missing, germinated, or established.

The main outcomes will be: 1) Quantifying the relationship between Allanblackia seed density and probability of seedling establishment. 2) Recommendations to local managers for sustainable management of Allanblackia and for further research if necessary. 3) Establishment of test plots that can be used for future Allanblackia regeneration research. 4) Support of a Tanzanian Masters student in Botany to collaborate on part of the project.

The results of this research can be useful not only in Tanzania, but in many other places where seed extraction is an important non-timber forest resource. Non-timber forest products are attractive alternatives to destructive harvesting, but the effects of non-destructive practices on forest ecology and biodiversity must also be considered.

Project Updates

Download Reports

PROJECT UPDATES

EVALUATION REPORTS