Ph.D Thesis

Ph.D StudentGao Junjing
SubjectA Global Meta-Analysis of Grazing Effects on Plant Richness
DepartmentDepartment of Civil and Environmental Engineering
Supervisor PROF. Yohay Carmel
Full Thesis textFull thesis text - English Version


Plant diversity in rangelands is a major component of world biodiversity. Yet, the determinants of the diverse responses of species richness to grazing are still not fully understood. Here, I used a combination of theoretical enquiries and empirical analyses of existing studies, in order to gain new insights into this old question.

The intermediate disturbance hypothesis (IDH), the dynamic equilibrium model (DEM) and the Milchunas-Sala-Lauthroth (MSL) model, are three widely-used theories to explain grazing-diversity relations. So far very few studies investigated the relations between them. My overview revealed overlaps in explanatory variables, mechanisms and predicted grazing-diversity patterns among these models. Just two patterns of grazing / species-richness, a hump-shaped curve and a monotonic decrease, can be summarized from these three models.

 Thus, the goal of next part of my thesis was testing how grazing-richness relations vary with aridity through evaluating the IDH and the MSL models using a global meta-analysis. Unfortunately, only few studies reported data variability. To include more studies, a vote-counting analysis was conducted as well. Both analyses revealed that the climate-specific MSL model better predicts the impact of grazing on diversity than the simplistic IDH in rangelands. Besides grazing intensity, ecosystem type (e.g. grasslands vs. shrublands) and livestock type are also important factors in affecting grazing-richness relations.

In the next part of my thesis, I aimed to quantitatively evaluate how other factors (i.e. evolutionary history of grazing, aridity, ecosystem type, stocking rate, duration of no-grazing, etc.) affect plant richness, and to determine how much variation of plant richness can be explained by these factors. Towards this end, I conducted an additional global meta-analysis with meta-regression analyses. This work has not been done before. I found that globally, grazing significantly increased species richness compared to no-grazing, with the strongest effects recorded in wet grasslands. Evolutionary history of grazing insignificantly affected vegetation responses to grazing, which cannot support the major perception behind the MSL model. Aridity, ecosystem type, and stocking rate were three important variables that together explained 41% of the global variation in species richness. Stocking rate had a stronger effect in wet areas and the duration of no-grazing had a stronger effect in dry areas.

 To enrich the existing models conceptually, I proposed a conceptual model with additional variables, e.g. livestock type, spatial and temporal scale, etc. This may distinguish between direct- and indirect effects of grazing and other variables, and it may also help us know how the mechanisms work from external variables or internal variables. Whether plant richness will increase, decrease, or remain the same depends on the sum of all positive- and negative effects caused by different variables.

In summary, this research revealed a full picture of grazing effects on plant richness at a global scale. Hopefully it may make contribution to our understanding of grazing-richness relations and provide a support to rangeland managers for devising proper grazing strategies to maintain plant diversity according to the site conditions in grazing management.