Non-stable natural edges and functional groups in gallery forests in Brazil

Background/Question/Methods

Gallery forests are a rare example of natural edges. This study aimed to address the following questions: How are the dynamic rates in natural edges? Do different functional groups have contrasting survival probability in the edges? In order to answer these questions, we studied natural edges of ten gallery forests, using three plots of 15 × 20 m in each one. We measured and mapped all individuals with DBH ≥ 1.0 cm, recording their DBH and height. One year after the first survey, we remeasured all surviving individuals, and recorded the dead ones. We calculated the annual rates of mortality (M) and recruitment (R) for individual trees, annual rates of basal area loss (L) and gain (G), and turnover rates (T_n for individuals and T_BA for basal area). We modeled the survival probability, in function of the edge distance, using generalized linear models, binomial distribution and logit link. We tested the models for nine abundant species distributed into three functional groups: 1) Pioneer species: Vismia guianensis, Myrsine guianensis, Miconia paulensis; 2) Light-demanding climax species:  Protium widgrenii, Protium spruceanum, Clethra scabra; 3) Shade-tolerant climax species: Calyptranthes brasiliensis, Calyptranthes clusiifolia, Casearia decandra.

Results/Conclusions

We followed the outcome of 7191 individuals. We found the recruitment rate higher than the mortality (R= 9.16%, M=2.44%), resulting in a net change rate of 7.41% and indicating a significant positive increase of the number of individuals. We found high turnover rates for the number of individuals (T_n=5.81%) and for the basal area (T_BA=6.63%). The gain in basal area was superior to the loss (G=9.94%; L=3.31%) resulting in positive rate of change (Ch_BA = 8.13%). The pioneer species presented a higher survival probability when they are closer to the edge, while light-demanding and shade-tolerant climax species behaved in the opposite way. We found the expected high dynamic rates in the edges only for the recruitment but not for the mortality. The increase of number of individuals and basal area from one year to another indicates a possible expansion of the edges, although possible cyclic fluctuations of the rates associated to large scale climatic variation demand further investigation at larger time scale. We found that the contrasting functional groups can be affected differentially by the strong environment gradient present in the edges even in very short distances (10m).