Estimation of above-ground biomass is vital for understanding ecological processes. Since direct measurement of above-ground biomass is destructive, time consuming and labor intensive, canopy cover can be considered a...Estimation of above-ground biomass is vital for understanding ecological processes. Since direct measurement of above-ground biomass is destructive, time consuming and labor intensive, canopy cover can be considered as a predictor if a significant correlation between the two variables exists. In this study, relationship between canopy cover and above-ground biomass was investigated by a general linear regression model. To do so, canopy cover and above-ground biomass were measured at 5 sub-life forms（defined as life forms grouped in the same height classes） using 380 quadrats, which is systematic-randomly laid out along a 10-km transect, during four sampling periods（May, June, August, and September） in an arid rangeland of Marjan, Iran. To reveal whether obtained canopy cover and above-ground biomass of different sampling periods can be lumped together or not, we applied a general linear model（GLM）. In this model, above-ground biomass was considered as a dependent or response variable, canopy cover as an independent covariate or predictor factor and sub-life forms as well as sampling periods as fixed factors. Moreover, we compared the estimated above-ground biomass derived from remotely sensed images of Landsat-8 using NDVI（normalized difference vegetation index）, after finding the best regression line between predictor（measured canopy cover in the field） and response variable（above-ground biomass） to test the robustness of the induced model. Results show that above-ground biomass（response variable） of all vegetative forms and periods can be accurately predicted by canopy cover（predictor）, although sub-life forms and sampling periods significantly affect the results. The best regression fit was found for short forbs in September and shrubs in May, June and August with R^2 values of 0.96, 0.93 and 0.91, respectively, whilst the least significant was found for short grasses in June, tall grasses in August and tall forbs in June with R^2 values of 0.71, 0.73 and 0.75, respectively. Even though the estimated above-ground biomass by NDVI is also convincing（R^2=0.57）, the canopy cover is a more reliable predictor of above-ground biomass due to the higher R^2 values（from 0.75 to 0.96）. We conclude that canopy cover can be regarded as a reliable predictor of above-ground biomass if sub-life forms and sampling periods（during growing season） are taken into account. Since,（1） plant canopy cover is not distinguishable by remotely sensed images at the sub-life form level, especially in sparse vegetation of arid and semi-arid regions, and（2） remotely sensed-based prediction of above-ground biomass shows a less significant relationship（R^2=0.57） than that of canopy cover（R^2 ranging from 0.75 to 0.96）, which suggests estimating of plant biomass by canopy cover instead of cut and weighting method is highly recommended. Furthermore, this fast, nondestructive and robust method that does not endanger rare species, gives a trustworthy prediction of above-ground biomass in arid rangelands.展开更多
文摘Estimation of above-ground biomass is vital for understanding ecological processes. Since direct measurement of above-ground biomass is destructive, time consuming and labor intensive, canopy cover can be considered as a predictor if a significant correlation between the two variables exists. In this study, relationship between canopy cover and above-ground biomass was investigated by a general linear regression model. To do so, canopy cover and above-ground biomass were measured at 5 sub-life forms（defined as life forms grouped in the same height classes） using 380 quadrats, which is systematic-randomly laid out along a 10-km transect, during four sampling periods（May, June, August, and September） in an arid rangeland of Marjan, Iran. To reveal whether obtained canopy cover and above-ground biomass of different sampling periods can be lumped together or not, we applied a general linear model（GLM）. In this model, above-ground biomass was considered as a dependent or response variable, canopy cover as an independent covariate or predictor factor and sub-life forms as well as sampling periods as fixed factors. Moreover, we compared the estimated above-ground biomass derived from remotely sensed images of Landsat-8 using NDVI（normalized difference vegetation index）, after finding the best regression line between predictor（measured canopy cover in the field） and response variable（above-ground biomass） to test the robustness of the induced model. Results show that above-ground biomass（response variable） of all vegetative forms and periods can be accurately predicted by canopy cover（predictor）, although sub-life forms and sampling periods significantly affect the results. The best regression fit was found for short forbs in September and shrubs in May, June and August with R^2 values of 0.96, 0.93 and 0.91, respectively, whilst the least significant was found for short grasses in June, tall grasses in August and tall forbs in June with R^2 values of 0.71, 0.73 and 0.75, respectively. Even though the estimated above-ground biomass by NDVI is also convincing（R^2=0.57）, the canopy cover is a more reliable predictor of above-ground biomass due to the higher R^2 values（from 0.75 to 0.96）. We conclude that canopy cover can be regarded as a reliable predictor of above-ground biomass if sub-life forms and sampling periods（during growing season） are taken into account. Since,（1） plant canopy cover is not distinguishable by remotely sensed images at the sub-life form level, especially in sparse vegetation of arid and semi-arid regions, and（2） remotely sensed-based prediction of above-ground biomass shows a less significant relationship（R^2=0.57） than that of canopy cover（R^2 ranging from 0.75 to 0.96）, which suggests estimating of plant biomass by canopy cover instead of cut and weighting method is highly recommended. Furthermore, this fast, nondestructive and robust method that does not endanger rare species, gives a trustworthy prediction of above-ground biomass in arid rangelands.
