Centre for Ecosystem Studies, Wageningen University, 6708 PB Wageningen, THE NETHERLANDS.
The succession of tropical secondary forests on abandoned agricultural fields has been studied since long, most often by comparing stands of different age since abandonment. These so-called chronosequence studies have yielded much insight in general patterns of succession and the constraints and conditions that affect the course of succession (shortly reviewed in chapter 1). Successional dynamics, however, are inferred rather than directly monitored in such studies; i.e. direction and rates of change of community structure and composition are deduced indirectly on the assumption that similar successional processes take place in any one stand. Only very few studies followed successional dynamics through repeated measurements in permanent sample plots, hence our current understanding of the variation in successional dynamics across secondary forest stands, the underlying processes of growth, mortality and recruitment, and the processes and mechanism that steer them, is still poor.
This thesis presents data from one of the very few studies on secondary tropical forest succession in multiple permanent sample plots. The focus of this work is on the first years of secondary forest succession, and on the role of competition on the level of individuals, populations and communities. The main aims of this work were: 1) To evaluate the dynamics of young secondary forests on abandoned agricultural fields in terms of rates and direction of community changes, how these dynamics vary across sites and how they change with time since abandonment; 2) To evaluate the dynamics of young secondary forests in terms of growth, mortality and recruitment, and how community changes are related to these processes; and 3) To examine the degree to which competition for light can account for variation in growth and mortality of saplings within a population and for variation in population dynamics across distinct sites.
The study was carried out in three villages in the Marqués de Comillas (MdC) region, part of the Selva Lacandona in eastern Chiapas, Mexico. The climate of the MdC region is humid tropical, with an average annual temperature of 24 .C, mean annual rainfall of ~3000 mm and a short dry period in February-April. The original vegetation consists mainly of tropical evergreen and semi-deciduous forests. Colonization of the region started in the late 1960s, and since then a large part of the landscape has been converted to a mosaic of temporary and permanent grazing pastures, cultivated land, young secondary forest and old-growth remnants. The prevailing land use types in the three villages are slash-and-burn shifting cultivation practices (mainly maize) with short fallow periods and semi-permanent cattle ranging.
Data for this study were collected in secondary forests on abandoned agricultural fields and pastures with time since abandonment (age) ranging from 0.5 to 18 years. A first dataset consisted of a chronosequence of 72 stands, widely distributed over an area of roughly 60 km2 and encompassing different geomorphological units (low hills vs. alluvial planes) and former land-use types (abandoned agricultural fields vs. pastures). In each stand, three 2 x 25 m transects were randomly located in which all woody stems with diameter at 1.3-m above ground (dbh) ≥ 1 cm identified to lowest possible taxon and their dbh and height were measured. A second dataset consists of permanent sample plots in twelve secondary forests with similar former land-use (maize fields) and geomorphology (low hills with sandy acidic and unfertile soils). Initial stand age ranged from 1-17 y. In each stand a 10 x 50-m permanent sample plot is established, and all trees with height ≥ 1.5-m are recorded, and tagged. Location of each tree was determined using a 1 x 1-m grid system. Every year dead trees and new recruits are recorded and dbh and height of all living trees is (re-)measured. The first dataset is used in chapter 2, and in chapters 3-7 data are used from the first three years of permanent plot study.
Chapter 2 demonstrates that patterns of species diversity, frequency and dominance are highly variable across secondary forests of an agricultural landscape, even when they are of similar age. Diversity varied widely between the plots, which could only partly be explained by differences in time since abandonment and geomorphology while the influence of former land use was even more ambiguous. Species richness changed with age, but dominance levels remained similar over the first 18 year of succession. Similarly, species richness, but not dominance, differed between secondary forests on low hills and alluvial plains. Across-stand frequency and abundance patterns varied strongly among tree species, and most among pioneer species. Despite the high species richness (almost 200 species identified across all plots), only a very small subset of the regional species pool dominated secondary forests in terms of frequency of occurrence and abundance, all of them pioneer species. Most pioneer species, however, and nearly all shade tolerant species, occurred in less than 15% of the plots and always in low densities. Only two species (Cecropia peltata and Trichospermum mexicanum) occurred in ≥ 50% of the plots and they were also the only species that were relatively dominant (> 10 stems / > 10% of stand basal area) in ≥ 20% of the plots. Other species were dominant (> 10% basal area) in the plots were they occurred, but never abundant (e.g. Ochroma pyramidale); frequent but never abundant (e.g. Vernonia patens); or locally abundant but with low relative basal area (e.g. Acalypha diversifolia).
The “Initial Floristic Composition” hypothesis predicts that, on abandoned agricultural fields with light previous land-use and close to seed sources, both pioneer and shade-tolerant species colonize a site directly after abandonment and that the recruitment of pioneers sharply declines while the recruitment of shade-tolerant species continues as soon as the canopy closes. It also predicts higher mortality among pioneers. Consequently, recruited and dead trees are expected to differ in species composition, with highest species richness for the recruits. To evaluate these predictions, recruitment and mortality was examined in eight permanent sample plots with initial fallow age of 1 to 5 y (chapter 3). Shade-tolerant species established in the first years of succession, albeit in low numbers. As predicted, recruited and dead trees differed in species richness and composition, and in shade-tolerant frequency. In contrast to expected, over 50% of recruits were from pioneer species. Results suggested that high stand-level mortality opened new opportunities for continued pioneer colonization. Results demonstrate that species turnover starts very early in succession but is not always a gradual and continuous process, complicating prevailing succession models.
In chapter 4, stand structure dynamics (i.e, rate and direction of stand structure changes) were related to mortality, growth, and recruitment rates over a two-year period, and the dependence of these demographic processes on fallow age and initial stand structure attributes was evaluated. Changes in stand structure were especially fast in the first 5 y of succession, and decreased rapidly afterwards, which resulted from similar stand-level changes in relative mortality, growth, and recruitment rates. Changes in most stand structure attributes (e.g. basal area, canopy height) were closely correlated, and more linked with tree growth rates than with net changes in tree density due to recruitment and mortality. Demographic processes were negatively related with initial stand basal area, but independent of initial tree density. Results suggest that asymmetric competition and resulting patterns of size and density dependent mortality and growth are major driving forces determining secondary forest successional pathways. Sudden mass mortality among dominant species in some stands showed that early secondary forest succession is not always a gradual and unidirectional process.
The strong spatial and temporal variability of successional dynamics emphasizes the need to monitor these dynamics in permanent plots across a range of initial stand ages, with multiple plots in a given age class. In chapter 5, case studies in Costa Rica and Mexico show that the actual rates of change in tree communities often deviate from chronosequence trends. With respect to tree species composition, sites of different ages differ more than a single site followed over time through the same age range. While changes in stem density have little to do with chronosequence trends, changes in basal area within stands generally followed chronosequence trends. Stem turnover rates were poor predictors of species turnover rates, particularly at longer time-intervals. Effects of the surrounding landscape on tree community dynamics within individual plots were not specifically included in this or in any of the other chapters, but results suggest that they are likely to be important determinants of species accumulation rates and relative abundance patterns.
Competition for light is often considered to be one of the main mechanisms that regulate successional vegetation dynamics. Competition for light is largely asymmetric, as larger trees shade smaller trees and not the other way around. Consequently, within site variation in growth and survival of tree saplings can be expected to relate to variation in the degree of neighborhood competition, especially from the larger neighbors. In chapter 6, these hypotheses were tested for Cecropia peltata and Trichospermum mexicanum saplings in two adjacent 1-2 year old secondary forest stands. The degree of competition was estimated by neighborhood basal area (NBA). Both diameter and height growth were strongly negatively related to NBA of the larger neighbors. Only a small proportion of variation in survival, on the other hand, could be accounted for by variation in NBA, suggesting that other factors are more important. The effect of NBA of smaller neighbors on growth and survival was very weak or non-significant, indicating that competition between neighbors, indeed, is largely asymmetric.
Population dynamics of a species vary much from one secondary forest to another. In chapter 7 we examined if such between-site differences in demographic behavior can be related to variation in asymmetric competition that result from differences in stand and population size structures. Specifically, we tested if differences in mean growth and mortality of the biggest and of the smallest trees of a population (growth and survival asymmetry) were related to differences in the average level of asymmetric neighborhood competition in both size groups (competition asymmetry). We used 24 populations of four pioneer tree species in nine early (1-5 y) secondary forest stands. For each population, we compared average neighborhood basal area, growth and survival of the twenty percent smallest and twenty percent largest trees. The between-plot variation in competition asymmetry explained a significant part of the between-plot variation in growth asymmetry, with smallest trees growing faster in populations with low competition asymmetry, and largest trees growing fastest in populations with high competition asymmetry. Survival was size-related, with lowest survival rates among the smallest plants, as expected. However, survival asymmetry was significantly related to competition asymmetry in just one species. The differences in the effects of asymmetric competition on the mortality and survival of the four species show that susceptibility to competition for, presumably, light, varies considerably between early successional pioneer species. Results of chapter 6 and 7 support the prediction that in young secondary forests, asymmetric competition for light is one of the main mechanisms that govern secondary forest dynamics.