Vegetation Dynamics: Recruitment and Regeneration in Two-Phase Mosaics

Abstract

Noy-Meir (1979/80) discussed the biology of desert vegetation in terms of two competing hypotheses that he named autecological and ecosystem. The former hypothesis suggests that "the dynamics of each population are determined by its independent reaction to the environment," whereas the latter holds "that nature, in general, consists of integrated ecosystems, in which all populations and many environmental factors are (directly or indirectly) linked and regulated by biological interactions and feedbacks." Clearly, the complex interrelationships between vegetation and various parts of the physical environment described in this review support the dominance of the ecosystem hypothesis. Explicit integration and cross-linking of knowledge about biota, resources, and physical conditions is essential to understand banded vegetation function (chapter 2, this volume). A range of abiotic processes (Shmida, Evenari, and Noy-Meir 1985) as well as biotic processes (Niering, Whittaker, and Lowe 1963; Noy-Meir 1985; Callaway 1995) contributes to the structure and functioning of banded landscapes. Also, spatial and temporal variability in the development of the successional process (Yarranton and Morrison 1974; Robinson et al. 1992), in turn, may arise from the variability in the outcome of biological interactions (Bronstein 1994 ). Banded vegetation patterns can be considered as a nested organizational hierarchy (Urban, O'Neill and Shugart 1987; Mauchamp 1992). At the coarsest or landscape organization level, banded landscapes are a mosaic composed of vegetation bands and bare areas, linked by the dynamic redistribution of rainwater by sheetflow into an alternating runoff/runon pattern. At a second finer-scale level, as soil water availability increases from the upslope edge to somewhere within the vegetation band, the vegetation cover and biomass increases to a maximum (Galle, Ehrmann, and Peugeot 1999) and then decreases to the downslope bare zone. If the availability of light changes in concert with the gradient in water and biomass, zonation of vegetation species composition may occur, as it does in Niger. Without light restriction, only the plant population size distribution is affected, as in Mexico and Australia. Few of the models have explicitly incorporated the effects of biological interactions between vegetation elements at the within-band level (Mauchamp, Rambal, and Lepart 1994; Thiery, d'Herbès, and Valentin 1995; Dunkerley 1997a,b; Lejeune, Couteron, and Lefever 1999). A third and least-studied level considers the individual plants within the vegetation array. The elements linking them are demographic processes, biotic interactions (e.g., facilitation and inhibition/competition), and fine-scale abiotic factors, (e.g., microtopography, soil crusts). The present review reveals that vegetation dynamics are the outcome of complex interactions at this level. Models predicting the behavior of individual system elements at this scale are inevitably complex, and none yet exist. The outcomes may be analyzed in terms of both succession models and of more process-based conceptual frameworks involving feedback loops (chapter 2, this volume). In the future, manipulative experiments in natural communities and careful analysis techniques (Gurevitch and Collins 1994) will be essential to verify those models. Demographic processes such as dispersal and establishment need much more elucidation to build on the work of Mauchamp and co-workers (1993), López-Portillo, Montaña, and Ezcurra (1996), and López-Portillo and Montaña (1999). This becomes important as human-use patterns become threatening processes to the survival of the vegetation system. The differential survival of mature plants in the different zones of the pattern implies spatially linked physiological limitations for perennial species in relation to water availability and other factors. These relationships need further attention to shed light on the functional role of the observed vegetation structures in the provision of ecosystem services (Mauchamp et al. 1993; Montaña, Cavagnaro, and Briones 1995; Seghieri and Galle 1999).