Chin J Plan Ecolo ›› 2016, Vol. 40 ›› Issue (1): 48-59.DOI: 10.17521/cjpe.2015.0246

• Orginal Article • Previous Articles     Next Articles

Geostatistical analysis of spatial variations in leaf traits of woody plants in Tiantong, Zhejiang Province

XU Ming-Shan1,2, ZHAO Yan-Tao1,2, YANG Xiao-Dong3, SHI Qing-Ru4, ZHOU Liu-Li1,2, ZHANG Qing-Qing1,2, Ali ARSHAD1,2,5, YAN En-Rong1,2,*   

  1. 1School of Ecological and Environmental Sciences, East China Normal University, Shanghai 200241, China
    2Tiantong National Forest Ecosystem Observation and Research Station, Ningbo, Zhejiang 315114, China
    3Institute of Resources and Environment Science, Xinjiang University, Ürümqi 830046, China
    4Youth Science and Technology Guide Station of Baoshan District, Shanghai 200904, China
    5Department of Environmental Sciences, Abdul Wali Khan University, Mardan 23200, Pakistan
  • Online:2016-01-31 Published:2016-01-28
  • Contact: En-Rong YAN
  • About author:

    # Co-first authors

Abstract: AimsExploring spatial variations in leaf traits and their relationships with environmental properties is crucial for understanding plant adaptation strategies and community assembly. This study aimed to reveal how leaf traits varied spatially and the role of environmental factors.MethodsThe study was conducted in a 5-hm2 forest plot in Tiantong, Zhejiang Province. Three leaf traits, including individual leaf area (ILA), specific leaf area (SLA), and leaf dry matter content (LDMC) were measured for 20253 individual trees with diameter at breast height (DBH) ≥1 cm. Soil properties measured included contents of soil total nitrogen, soil total phosphorus, soil total carbon, soil pH value, soil volumetric water content, bulk density, and humus depth. Topographic variables measured included elevation, slope and convexity. We used geostatistical analysis to reveal spatial variations of the three leaf traits. Relationships between leaf variability and environmental factors were analyzed using principal component analysis (PCA) and Pearson’s correlation.Important findings Spatial variability followed the order of ILA > SLA > LDMC. Spatial autocorrelation of three leaf traits was weak within a distance of 5.16 m. The optimal model of the semi-variogram function was Gaussian model for ILA, and exponential model for SLA and LDMC. ILA showed the largest variability at the direction of northeast-southwest, and smallest variability at the direction of northwest-southeast. In contrast, SLA and LDMC had the highest variability at the direction of northwest-southeast and least variability at the direction of northeast-southwest. There were significantly negative relationships between ILA and topographic factors (r = -0.12, p < 0.0001), and between SLA and soil nutrients (r = -0.16, p < 0.0001). In contrast, LDMC was positively correlated with soil nutrients (r = 0.13, p < 0.0001). Relative to soil nutrients, topographic factors affected much more variations in ILA, SLA and LDMC at the direction of northeast-southwest. Distinctly, at the direction of northwest-southeast, variability of ILA was affected mainly by topographic factors, while soil nutrients resulted in the most variability of SLA and LDMC. In conclusion, leaf traits varied considerably with spatial direction in the studied forest plot. Associations between leaf traits and topographic factors and soil nutrients indirectly indicated effects of environmental filtering on community assembly.

Key words: leaf traits, spatial variation, soil nutrients, topographic factors, geostatistical analysis