Tropical forests are non-equilibrium ecosystems governed by interspecific competition based on universal 1/6 niche width

11Citations
Citations of this article
70Readers
Mendeley users who have this article in their library.

Abstract

Tropical forests are mega-diverse ecosystems that display complex and non-equilibrium dynamics. However, theoretical approaches have largely focused on explaining steady-state behaviour and fitting snapshots of data. Here we show that local and niche interspecific competition can realistically and parsimoniously explain the observed non-equilibrium regime of permanent plots of nine tropical forests, in eight different countries. Our spatially-explicit model, besides predicting with accuracy the main biodiversity metrics for these plots, can also reproduce their dynamics. A central finding is that tropical tree species have a universal niche width of approximately 1/6 of the niche axis that echoes the observed widespread convergence in their functional traits enabling them to exploit similar resources and to coexist despite of having large niche overlap. This niche width yields an average ratio of 0.25 between interspecific and intraspecific competition that corresponds to an intermediate value between the extreme claims of the neutral model and the classical niche-based model of community assembly (where interspecific competition is dominant). In addition, our model can explain and yield observed spatial patterns that classical niche-based and neutral theories cannot. © 2013 Fort, Inchausti.

Figures

  • Table 1. Observed (bold) and predicted species richness for all trees with diameter at breast height (dbh)$1 cm for the first census in nine tropical forest plots.
  • Figure 1. Observed (bold) and predicted distribution of relative species abundances (RSA) for all trees with diameter at breast height (dbh)$1 cm for the first census of nine in nine tropical forest plots. Data from Center for Tropical Forest Science [11]. The predicted (grey) are averages 6 std of 100 model simulations for the best estimates of model parameters of each forest (see Information S1 for comparisons with other censuses). The calculation of the RSA is explained in Information S1. doi:10.1371/journal.pone.0082768.g001
  • Figure 2. Species-area curves (SAR) and spatial patterns of tree species richness for selected censuses of tropical forests. Predicted curves correspond to averages over 100 simulations for the best estimates of model parameters (Fig. 2a), and the error bars correspond to one std. a. Observed and predicted (grey line) number of tree species with dbh$1 cm for sampling areas of different sizes at Barro Colorado (1990, triangles) and Pasoh (1987, crosses). Estimated curves for Barro Colorado and Pasoh were calculated using data from the Center for Tropical Forest Science [11] and dividing the entire plots into non-overlapping quadrats [29]. The calculation of the SAR is explained in Information S1. b. The estimated (triangles) and predicted probability F(r) that two randomly selected trees of dbh$10 cm located r meters apart for the 1990 census at Barro Colorado plot are conspecific. The curves are shown only for 10#r#100 m, a range of distances for which the NTB fails to reproduce the estimated F(r) [6]. doi:10.1371/journal.pone.0082768.g002
  • Figure 3. Relation between species rarity and spatial aggregation for Spachea membranacea at Barro Colorado in the 1995 and model species # 253. Both species have the highest spatial aggregation, as measured by the aggregation index V0R10 and the same abundance of 14 individuals. a. The clumpy distribution of relative abundance of species in the finite niche axis, showing that species # 253 (niche position x = 0.8179) lies in a gap between clumps of coexisting species. b. The observed distribution of Spachea membranacea at Barro Colorado in the 1995 census [11,24] yielding V0R10 = 689.3. c. Spatial distribution of a model species # 253 in grey. The selected 969 sublattice shows the species identity of individuals in the immediate neighbourhood containing all 14 individuals of model species # 253, yielding V0R10 = 1031.7. d. The set of fitnesses in the same sublattice containing all 14 individuals of model species # 253. Individuals of the rare species # 253 (in grey) are poorer competitors because they have lower fitnesses than most of their immediate neighbours. doi:10.1371/journal.pone.0082768.g003
  • Figure 4. Observed and predicted compositional changes in tree communities between censuses at Pasoh and Barro Colorado forests. Compositional changes are measured by the coefficient of determination R2 of the regression of the log-transformed, time-lagged population abundance of all species between censuses for all individuals with dbh$1 cm and species with two or more individuals in first census of each forest [25,26]. At a time lag of zero, no change in community composition can yet have occurred, and thus R2 is by definition equal to unity. As time elapses between censuses, the progressive compositional changes are reflected by the decay in R2: empirical results for Barro Colorado (triangles) and Pasoh (crosses), and predicted values, corresponding to averages over 100 simulations with error bars equal to one standard deviation for the best parameters estimates for each forest. The model predicts a nearly perfectly linear decay in average values of R2 that is indistinguishable from the prediction from NTB. doi:10.1371/journal.pone.0082768.g004
  • Figure 5. Sequential procedure used to estimate the model parameters for each forest illustrated with a hypothetical grid of 363 whose nodes are occupied by different species. a. In the first stage, forest dynamics starting from 100 random initial conditions (central niche positions m(s) for the s species and spatial positions for all L individuals dynamics, both drawn from uniform distributions), dynamics consisted on the sequence of ti (i: 1…100) individual replacements (see rules in the main text) until reaching a configuration whose standardised Shannon diversity was similar to the one observed for the first census. We constrained the species richness to be equal to the one observed in the first census. We then searched for the values of m and s yielding the maximum value of the coefficient of determination R2 obtained by regressing the relative species abundance (RSA, averaged across the 100 simulations) distribution on the RSA of the first census (provided that R2$0.95). b. For the fitted values of m and s, the second stage estimated the value of T for those forests having more than one complete census. Starting for the 100 configurations comparable to the first census, we restarted the sequence of t9i (i: 1…100) individual replacements (now letting species to become extinct) until each configuration had a standardised Shannon diversity or species richness (whichever came first) similar to the one observed for the second census. As in the first step, we searched for the value of T yielding an average RSA similar to the one observed in the subsequent censuses, as judged by the R2 criterion as in step (a). doi:10.1371/journal.pone.0082768.g005

References Powered by Scopus

Time-dependent statistics of the Ising model

3098Citations
N/AReaders
Get full text

Niche tradeoffs, neutrality, and community structure: A stochastic theory of resource competition, invasion, and community assembly

1252Citations
N/AReaders
Get full text

Light-gap disturbances, recruitment limitation, and tree diversity in a neotropical forest

1207Citations
N/AReaders
Get full text

Cited by Powered by Scopus

Local and regional determinants of vascular epiphyte mortality in the Andean mountains of Colombia

25Citations
N/AReaders
Get full text

Competition across diverse taxa: quantitative integration of theory and empirical research using global indices of competition

14Citations
N/AReaders
Get full text

Community diversity and total abundance: Quantitative predictions from competition niche theory

10Citations
N/AReaders
Get full text

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Cite

CITATION STYLE

APA

Fort, H., & Inchausti, P. (2013). Tropical forests are non-equilibrium ecosystems governed by interspecific competition based on universal 1/6 niche width. PLoS ONE, 8(12). https://doi.org/10.1371/journal.pone.0082768

Readers over time

‘14‘15‘16‘17‘18‘19‘20‘21‘22‘23‘24‘25036912

Readers' Seniority

Tooltip

PhD / Post grad / Masters / Doc 23

52%

Researcher 16

36%

Professor / Associate Prof. 5

11%

Readers' Discipline

Tooltip

Agricultural and Biological Sciences 34

65%

Environmental Science 15

29%

Biochemistry, Genetics and Molecular Bi... 2

4%

Computer Science 1

2%

Save time finding and organizing research with Mendeley

Sign up for free
0