Ordination Analysis of Components of Resilience of Quercus Coccifera Garrigue

The trajectories or vectors of sites through the statistical space of ordinations were used to quantify components of resilience for data on garrigue in southern France, from six experimental fire treatments and a control, through 13 yr of annual records. Four types of data were ordinated: foliar cover, point contacts, point contacts within horizontal strata, and presence—absence. An additional data set was created by adding the data on foliar cover to the cover data of 156 releves sampled by Braun—Blanquet and his associates earlier in the century, and which we took to represent the surrounding vegetation. Only the data of point contacts in strata consistently showed directional movement that could feasibly be used to quantify components of resilience, indicating the processes by which the vegetation, and especially the dominant Quercus coccifera, recovered. Postfire resprouting was so prevalent that measures of foliar cover or presence—absence could not distinguish well between prefire and postfire conditions, although they obviously were different. The ordinations showed that garrigue has a very high initial elasticity, which slows later; it is a function of growth. High elasticity was maintained in all treatments throughout the experiment and was the best indicator of potential resilience. The amplitude of the community was never exceeded, even by very frequent fires. Vectors of recovery toward the predisturbance condition were always evident; when the garrigue was examined in the context of the surrounding vegetation it never actually moved into the areas in ordination space that defined other communities. While the experiments disturbed the community too often to compare a steady postdisturbance location in ordination space with the predisturbance location, the garrigue was already recovering to the point where malleability was obviously low, and damping was high. The vegetation recovered in such a way that little opportunity was afforded for invasion from other communities. Inertia, seen indirectly as the distance from the predisturbance point to the point representing the sample the following spring, was also quite strong. The factors that did make a difference among the components of resilience are those previously identified as important factors in the recovery of garrigue: season of fire first, frequency second. The season of fire was important, because for autumn burns the plants had little time and poor conditions for regrowth before the census the following spring. This difference is seen in the initial vectors from the origin, and although the subsequent vectors show strong recovery, the initial deficit was never overcome. The fire intensity of controlled burns is shown to have little effect on resilience.