Plant Competition And Facilitation And Their Role In Driying Population Dynamics Across Environmental Gradients: A Study Based On Zone-of-influence (ZOI) Models

The relationship between plant-plant interactions and environmental conditions is central to ecological research, as represented by the classic "Tilman-Grime" debate about the role of competition in plant communities and the recently proposed "stress-gradient hypothesis" concerning the balance between negative (competition) and positive (facilitation) interactions. One of the fundamental principles of complex systems science is that processes at the individual level is the driving force of system-level properties, therefore clarifying how plant interactions vary along environmental gradients is critical to understanding the variations in population dynamics, community structure and ecosystem functions over space and time. Competition, as an important community organizer, has been extensively studied. However, despite increased attention and interest, few studies have explicitly examined the role of facilitation at higher levels of biological organization. It is time for plant ecologists to move beyond merely investigating where and when facilitation might occur, and to make more effort to integrate facilitation into the main framework of ecological theory. This study focuses on the individual and population levels. By developing one-and two-layer (shoot and root layers) zone-of-influence models and then carrying out simulation experiments, this study comprehensively explores how different components and facets of plant interactions change over environmental gradients, and further tests how competition and facilitation interplays to determine some of the most important population dynamics across environmental gradients. Simulated population dynamics were compared with field patterns.Results showed that the intensity and importance of plant interactions, as well as the interplay between above-and belowground components of the net interaction, varied along simulated abiotic stress gradients. In the absence of facilitation both competition intensity and importance decreased with increasing stress, with aboveground competition intensity decreasing at a lower rate than belowground competition intensity. In the presence of facilitation, the net interaction shifted from negative under benign conditions to positive in stressful environments. Interaction intensity varied generally in parallel among populations of different densities, indicating the lack of stress x density interaction. The higher the density, the more negative the interaction. Interaction intensity and importance varied along stress gradients in different manners:the shift of the former was characterized by a concave curve, while the latter exhibited a convex curve; the former was density-dependent, whereas the latter depended little on density. These differences may account for the complex and variable relationship between the two measures of competition observed over natural productivity gradients. A strong negative interaction between above-and belowground competition (one form reduces the other) was detected under benign conditions. As stress increases, above-and belowground competition became independent on each other. In the presence of facilitation, however, a negative interaction between above-and belowground components of the net interaction held over the entire stress gradient.The interaction among facilitation, competition and abiotic stress determined the spatial patterning of populations during density-dependent mortality. Started with a clustered pattern, asymmetric competition consistently led to immediate and spatially non-random mortality towards regularity, thus rapidly decayed the initially clustered pattern to final patterns of small-scale regularity and large-scale randomness. The role of symmetric competition in decaying the clustered pattern increased with abiotic stress because the stress-induced reduction in plants'growth rates can make individuals in high-density clusters more likely to die even from symmetric competition. Facilitation played a clear role in counteracting the effect of stress, thus tended to maintain the degree of clustering of the pattern during density-dependent mortality. This is because the amelioration of harsh conditions by neighboring plants relieved the reduction in plant growth due to competition, thus slowed down and reduced the mortality inside clusters (relative to that outside clusters). Moreover, the effect of facilitation appeared to increase with abiotic stress. Field investigations showed that in the Gobi Desert of Northwest China the population spatial pattern of dominant shrub species changed from random to clustered as the degree of drought increased. Simulation results revealed that facilitation among neighboring plants is partially responsible for clustered population spatial patterns observed in stressful environments, even though its contribution relative to other factors (e.g. local dispersal and environmental heterogeneity) remains to be evaluated.Both the speed and form of plant self-thinning varied across simulated abiotic stress gradients. Plant height-crown radius allometry and competitive and facilitative interactions were controlling the dynamics of density-dependent mortality and the self-thinning line across stress gradients. The one-layer model was used with different settings of competitive size-symmetry, facilitation and stress to test the robustness of the conclusions of a recent publication using a similar model. One-layer simulation results were then compared to those from the two-layer version, in which the overall size-symmetry of competition is regulated by the adaptive allocation to root vs. shoot growth, thus by the relative strength of root vs. shoot competition. One-layer model simulations revealed that increasingly asymmetric competition accelerated thinning, and markedly steepened (slope ranged from about-1to-4/3) and lowered self-thinning lines. Stress accelerated thinning slightly under completely symmetric competition, whereas slowed it down considerably under more asymmetric competition modes. Stress had no effect on the self-thinning slope, but acted to decrease the intercept of self-thinning lines. Facilitation simply counteracted stress effects. Allometric growth (lateral expansion being faster than vertical growth) in stressful environments flattened the self-thinning line. In the two-layer model, both stress and facilitation affected mortality dynamics in the same way as in the one-layer version when competition was not completely symmetric. Different from the one-layer model results, the two-layer model predicted that high stress dramatically flattened self-thinning lines, whereas facilitation again counteracted the stress, thus steepened thinning lines. The influences of both stress and facilitation on the self-thinning slope were mediated by their effects on the size-symmetry of competition. Although based on current knowledge about plant-plant interactions our two-layer model seems to better represent the real ecological scenario, and its predictions matched field patterns more closely, elaborately designed experiments need to be conducted to examine how competition and facilitation interact to determine plant population dynamics in nature.