Spatial pattern of Trees in the Riparian Forests (Case Study: Wildlife Refuge of Karkhe)

Analysis of the spatial pattern is an important element for the optimum management and having knowledge about ecosystem function. This study was conducted to analyze the spatial pattern of trees and shrub in the riparian forest landscape of Karkhe, also for assessment of their spatial variability regarding the distance from the river. The number of trees and shrubs were sampled using 117 plot (20m × 20 m) along parallel transects (perpendicular to the river). The distance between transects was 500 m. The variograms of Populus euphratica and Lycium shawii were spherical. Tamarix sp. and total of them were exponential. All revealed the presence of spatial autocorrelation. The range of influence was 186 m for Tamarix sp., 362 m for Populus euphratica, 749 m for Lycium shawii and 208 for total. The kriging maps showed spatial variability of them. The spatial pattern of species total and Tamarix sp. are similar that was shown in correlation. So Tamarix sp. is a major part of this forest. Populus euphratica is replacing with Lycium shawii. Also the higher distribution is nearby river for all trees and shrub.
Extended Abstract
1-Introduction 
In the last 15-20 years, riparian forests have become recognized as important components of landscape and serve as a vital link between the aquatic environment and upland ecosystems. Riparian ecosystems are aquatic-terrestrial ecotones with unique biotic, biophysical and landscape characteristics. Accordingly, sustainability and maintenance of riparian vegetation or restoring of degraded sites is critical to sustain inherent ecosystem function and values. Many scientists believe promoting sustainability is the overarching goal of landscape (and regional) planning. A current challenge in ecology is underestimating how patterns and processes vary with scale and searching for general principles (assembly rules) which determine the species composition of communities. Also a fundamental and unanswered research question in landscape ecology centers on how the spatial arrangement of the ecosystems influences the distribution and abundance of species. Description of patterns in species assemblages and diversity is an essential step before generating hypotheses in functional ecology. Information about the spatial pattern of tree species in riparian forest ecosystems is limited though required, e.g. for understanding spatial distribution effects of trees on ecosystem processes. The practical consequences of these findings are useful for sustainable management of forest ecosystems and in monitoring of forest evolution. One main reason responsible for the absence of information about spatial pattern researche is the lack of adequate methods for analyzing data. Geostatistics provide descriptive tools such as variogram to characterize the spatial pattern of continuous and categorical forest attributes. This method allows assessment of consistency of spatial patterns as well as the scale at which they are expressed. This study was conducted to analyze spatial patterns of tree species in the riparian forest landscape of Karkhe.
2- Materials and Methods
The study was carried out in Wildlife Refugee of Karkhe in the riparian forest of the southwestern Iran (31^o  57^/- 32 ^o 05^/ N and 48 ^o 13^/- 48 ^o 16^/ E). The climate of the study area is semi-arid. Average yearly rainfall is about 325.8 mm with a mean temperature of 24^oc. Plant cover, mainly comprises Populus euphratica and Tamarix sp. The both sides of river are similar, so we sampled on one of the two sides. The number of trees and shrubs were sampled using 117 plot (20m × 20 m) along parallel transects (perpendicular to the river). The maximum distance between plots was 0.5 km. The sampling procedure was hierarchically, we considered maximum distance between samples as 0.5 km, but the samples was taken at 250m, 100m, 50m at different location of sampling.We began with an exploratory analysis of patterns in the data. So, classical statistical parameters, i.e. mean, standard deviation, coefficient of variation, minimum and maximum, were calculated using SPSS17 software. Next, we applied geostatistics methods to examine spatial structure in tree distribution.To determining the spatial structure, we calculated the semivariances. Semivariance quantifies the spatial dependence of spatially ordered variable values. The semivariance essentially expresses the average variance of pairs of points at a given distance. Empirical variograms are plots of the semivariances, averaged over distance classes (called lags), against the lag distance. To describe spatial autocorrelation of a variable quantitatively, a (theoretical) variogram model can be fitted to the empirical variogram in order to obtain the model parameters nugget, sill and range.   
3- Results and Discussion
The variograms of Populus euphratica and Lycium shawii were spherical. Tamarix sp. and total of them were exponential. All revealed the presence of spatial autocorrelation. The range of influence was 186 m for Tamarix sp., 362 m for Populus euphratica, 749 m for Lycium shawii and 208 for total. The kriging maps showed spatial variability of them. Typically these structures constitute one source of the nugget variance of the variograms (Rossi, 2003). However, the variograms reported here featured a somewhat high ratio of nugget variance to sill. This result showed that there was the small-scale variability and important proportion of unexplained variance. Spatial distribution of tree speciec may be influenced by factors like gradients in soil organic matter (quantity and or quality) and texture. These factors together with intrinsic population processes constitute proximate controlling factors of population structure.
4- Conclusion
The spatial pattern of total species and Tamarix sp. are similar that was shown in correlation. So Tamarix sp. is a major part of this forest. Populus euphratica is replacing with Lycium shawii. Also the higher distribution is nearby river for all trees and shrub.