Background Documenting the entire lifetime of long-lived organisms requires splicing together short-term observations. Matrix demography provides a tool to calculate lifetime statistics, but large samples from juvenile to adult are needed, and few such studies have been done in tropical trees because high species diversity limits sample sizes. The 50-ha plot at Barro Colorado in Panama was designed to provide large samples, and with 30 years of censuses, accurate population matrices can be constructed.
Methods In 31 abundant species, I divided all individuals ≥1 cm dbh into 4 or 5 size class in each of seven censuses. Movements of stems between size classes over two censuses are termed transitions, and I constructed complete transition matrices for each species. From the matrices, I derived analytic solutions for lifetime demographic statistics. Expected adult lifespan from the sapling stage was the key statistic.
Results Expected adult lifespan from the sapling stage varied 100-fold over the 31 species, from 0.5 to 50 years, and maturation time varied from 19 to nearly 200 years. Species with the highest growth rates also had high death rates, and theoretical calculations of reproductive lifespan show that the fast-growing pioneer species have short expected adult lifespans relative to the average slow-growing, shade-tolerant species. Within the slow-growth category, however, there was high variation in expected adult lifespan, and several shade-tolerant species under-performed the pioneers in terms of adult lifespan.
Conclusions Analytical solutions from population matrices allow theoretical analyses that integrate short-term growth records into lifespans of tropical trees. The analyses suggest that pioneer species must reproduce more successfully than shade-tolerant species to persist in the Barro Colorado forest. My next goal is to incorporate seed production and germination into lifetime matrix demography to test this hypothesis.
DownLoad: