Carbon pools recovery through succession in a tropical dry forest: Explanatory and predictive ability of fallow age, forest structure and functional traits

Background/Question/Methods

Although a plethora of studies have reported forest biomass or carbon accumulation during succession in tropical forests using a chronosequence approach, few of them have tested which are the attributes of the tree community associated to recovery. Previous work suggest that fallow age, structural attributes such as basal area or mean height, or functional attributes such as wood density and maximum size of the species are associated to carbon stocks. However, neither study has tested if carbon recovery inferred from chronosequences matches real-time change in carbon stocks, and if those forest attributes associated to carbon stocks are the same predicting real-time changes. In this study we answer this question using a system of twelve permanent plot sites in different successional stages from a tropical dry forest located on the Mexican pacific coast. For each site we estimated carbon in aboveground living biomass (ABC) for three repeated censuses (2004, 2007, and 2010). We used a chronosequence approach to test recovery of carbon through time and assessed its predictive ability by comparing chronosequence based trajectories with observed trajectories from the repeated measurements. Then we tested if structural (basal area, mean tree height, tree density) or functional attributes (wood density, specific leaf area, maximum height, leaf nitrogen content, and leaf delta C¹³) of the tree community explained observed trajectories using structural equation modeling.

Results/Conclusions

ABC showed a positive relation with fallow age, but the predictive ability of the chronosequence model was low, with a consistent overestimation of real changes in carbon stock. We also found that although basal area and specific leaf area explain differences in stocks, neither of the structural or functional attributes tested explain observed changes. Our results show that explaining carbon stocks is not the same as explaining carbon accumulation during succession. Also, results suggest that there is significant uncertainty in explaining and predicting real changes in C pools through succession in these tropical dry forests.