Effects of different sampling modes on the results of vegetation ordination analysis

The relationships between plant and environment are an important topic in community ecology. To study these relationships, quadrate is often used to sample vegetation and environmental data, and ordination techniques are used to analyze the data. However, the size and shape of quadrate considerably affect the ordination results. To understand this effect remains a research item for various vegetation and environmental data settings. This paper studied the effects of different sampling quadrate (size and shape) on the results obtained from three common ordination techniques (CA, DCA and CCA). We did this by sampling the same transect in forest form map six times, using six different quadrates (0. 5 km × 0. 5 km, 0.5 km × 2 km, 2 km × 0.5 km, 1 km × 1 km, 1 km × 4 km, 2 km × 2 km). The results showed that large rectangle quadrate could capture greater percent variance of species data and more information about rare and unique species than small square quadrate. All sizes and shapes of quadrate used in this study had little effect on the dominant species. The captured soil information was sensitive both to the size and to the shape of quadrate, and the information of slope, longitude and latitude was sensitive to the change of quadrate size. Slope position, altitude, temperature and precipitation were sensitive to the change of quadrate shape. Large quadrate reduced the importance of altitude, temperature and precipitation while increased the importance of exposure, but these environmental factors appeared to be important in small quadrate sampling.