Progression and stability analysis of rain forest tree growth under environmental stochasticity

Abstract

Changes in relative stem growth rate (rgr) across periods of multi-censused plot data in tropical forest allow response indices (RIs) to be related to environmental (climatic) variation over time. Among small trees (12.5 to <50 cm girth at breast height [gbh]) in two permanent 4-ha plots at Danum, Sabah, Borneo (lowland dipterocarp formation), growth responses were followed before, during, and after a major ENSO drying event in 1998 (censuses of 1986, 1996, 1999, 2001, and 2007 in subplots). Overall, RI was negative in the interval with the drought, increased or reflected overcompensation in rgr immediately afterwards, and then settled back to pre-event levels. However, among the various tree species, responses and trajectories were very different: Some species showed positive or negative trends, while others showed stabilizing ones, and several had stable and unstable oscillations. Response index was graphed for pairs of periods in two ways: "consecutive" and "constitutive," allowing, respectively, progression (fundamental, implicit and explicit) and stability (fundamental, intended, and extended) to be examined. Strongest RIs were found among the understory species. Variation in RI was, nevertheless, very high between individuals within species and masked most average species' differences. Environmental stochasticity appears to lead to strong mixing effects of species composition over time, not necessarily randomly, but in a highly complicated manner depending on tree size, topographic location, neighborhood, and over-understory status, which would compound demographic stochasticity in recruitment to the small-tree class. Fluctuating responses cascade over one another in a highly unpredictable manner. These pluralistic responses may form the basis to a new understanding of tree population dynamics within the forest ecosystem, and to an improved theory about the maintenance of high species richness. Plasticity in allocation of resources to roots vs. shoots under water limitation offers a feasible hypothesis for high individual variation in growth. Instead of tropical rain forests being viewed as "complex," an empirically more sound way is to regard them as being highly "complicated" within a mechanistic and functional approach. Multi-species tree population dynamics are accordingly highly unpredictable and quasi-indeterminate.