Influences of stand, soil and space factors on spatial heterogeneity of leaf area index in a spruce-fir valley forest in Xiao Hinggan Ling, China

YANG Huan-Ying, SONG Jian-Da, ZHOU Tao, JIN Guang-Ze, JIANG Feng, LIU Zhi-Li

Chin J Plant Ecol 2019, 43 (4): 342-351. DOI: 10.17521/cjpe.2018.0310

Abstract （1230）

HTML （76）

PDF（pc）（2590KB）（691）

AimsSpatial heterogeneity of leaf area index (LAI) is very important for exploring the growth and spatial distributions of plants, as well as response strategy of plants to climate changes. Many previous studies have shown that biotic and abiotic factors had significant influences on spatial heterogeneity of LAI. However, few studies have been conducted to show the relative contributions of different influencing factors to the total variations of LAI. Our aim was to quantify the relative contributions of stand, soil and space factors to the total spatial variations of LAI in a spruce-fir valley forest in northeast China.
MethodsWe relied on a 9.12 hm ² (380 m × 240 m) spruce-fir valley forest plot in Xiao Hinggan Ling, China, which was divided into 228 subplots (20 m × 20 m). First, we measured LAI for each subplot by using the LAI-2200 plant canopy analyzer and then analyzed the spatial heterogeneity of LAI using geo-statistic methods (semivariogram and Kriging interpolation). Second, we measured 28 stand factors and 10 soil factors for each subplot, and quantified space factors using principal coordinates of neighbor matrices (PCNM). Finally, we quantified the relative contributions of stand, soil and space factors to the total spatial variations of LAI using the variance partitioning method.
Important findings The results showed that strong spatial autocorrelations of LAI values within 37 m distances in the spruce-fir valley forest, and the LAI presented different spatial patterns along distinct directions. The stand, soil and space factors totally explained 50.4% of the total spatial variations of LAI in the forest plot. The space factors explained greater spatial variations of LAI in relative to stand and soil factors, and solely explained 25.5% of the total spatial variations. The density of middle tree group (5 cm < diameter at breast height ≤ 10 cm) and basal area of major tree groups (including Abies nephrolepis and Picea spp.) were both significantly and positively correlated with LAI; and soil mass moisture content was significantly and negatively correlated with LAI. These results generally suggest that the spatial autocorrelation is more important than stand factor and soil factor for determining spatial heterogeneity of LAI of the spruce-fir valley forest in Xiao Hinggan Ling, China.

式中, λ(h)为变异函数, N(h)为间距为向量h的点对总数, Z(x_i)为系统某属性Z在空间位置x_i处的值, 即LAI值, Z(x_i + h)为Z在(x_i + h)处的值。通过变异函数及曲线图可得到3个重要参数: 基台值(sill, C₀ + C)、块金值(nugget, C₀)和变程(range, A)。基台值是当采样点间的距离h增大时, 变异函数λ(h)从初始的块金值达到一个相对稳定的常数, 基台值越大, 表明LAI总的空间异质性程度越高; 块金值表示随机因素产生的空间异质性, 较大的块金方差表明较小尺度上的某种过程不可忽视, 其主要源于LAI在空间尺度上存在的差异或测量误差; 变程反映空间变异的尺度, 在变程内, LAI存在空间自相关, 反之不存在(李哈滨等, 1998; 王政权, 1999)。此外, 空间结构比(C₀/ (C₀ + C))常用来反映自相关引起的空间变异在总空间异质性中的比例(王政权等, 2000)。当空间结构比>0.75时, LAI的空间自相关性很弱; 空间结构比在0.25-0.75之间时, 空间自相关为中等强度; 当空间结构比<0.25时, 空间自相关性很强(Li & Reynolds, 1995)。本研究采用的变异函数理论模型包括球状模型、高斯模型和指数模型(Rossi et al., 1992), 且决定系数(R2)值最高者为最优模型。Kriging空间插值是一种估计观测样点间内插值的地统计学方法, 它基于区域化变量理论; 当获得了变量的半变异函数的模拟模型后, 就可以利用样点观测值对研究区域上未取样点的区域化变量值进行估测(Bivand et al., 2013), 进而可得到LAI的空间格局。

在60 m尺度内, LAI在4个方向上的空间异质性均呈增大趋势(图3); 随着尺度增大不同方向上的空间异质性均呈波浪形变化, 除南-北方向(90°)外, LAI在其他3个方向上的空间异质性均整体呈减小趋势; 当尺度大于140 m时, LAI在东-西方向(0°)上的空间异质性随尺度的增大持续降低(图3); 表明LAI的空间异质性不仅与尺度相关, 而且受方向的影响, 各向异性比的结果再次验证了该结论(图4)。在100 m尺度内, 东北-西南和东南-西北(45° vs 135°) 2个方向上的各向异性比接近于1.0, 即各向同性, 但其他尺度内均是各向异性。在60 m尺度内, 东西-南北方向上的各向异性比大于东北-西南和东南-西北, 而随尺度增大, 呈相反结果(图4)。

LAI能够很好地反映植物的生长状况及分布格局(Liu et al., 2015; Zhu et al., 2016), 因此, 揭示谷地云冷杉林LAI的空间异质性及其影响因素对于探究目前谷地云冷杉林处于衰退状态的原因具有重要指示作用, 还可为确定最优的调查样方面积提供科学依据。调查样方面积大, 更具代表性, 且调查结果更可靠, 但需耗费更多的人力、物力和财力; 样方面积小, 虽然降低了工作量, 但缺乏代表性, 调查结果的可靠性较差, 因此, 确定最优的调查样方面积, 对于群落生态学研究至关重要(Bequet et al., 2012b)。本研究发现谷地云冷杉林的LAI在37 m (变程)范围内具有强烈的空间自相关性, 因此, 为避免空间自相关对调查结果的影响, 测定谷地云冷杉林LAI的最优样方面积为37 m × 37 m, 这对日后研究具有重要参考价值。其他学者得到类似结论, 如Zhu等(2016)通过分析7月份亚热带常绿阔叶林和落叶阔叶林内LAI的空间异质性, 得到变程(A)分别为18 m和23 m。然而也有其他学者得到的变程远大于本文结果, 如姚丹丹等(2015)利用半球摄影法分析了吉林省谷地云冷杉林LAI的空间异质性, 得到平均变程为66 m; Zhu等(2016)得到7月份亚热带针阔混交林内的变程为93 m; Liu等(2018)报道小兴安岭阔叶红松林中的变程为92 m。这些不同结论可能主要源于森林类型、调查时期以及LAI测定方法的差异。此外, LAI的空间异质性不仅与尺度相关, 还受方向的影响, 方向不同, LAI呈现不同的空间分布格局(图3, 图4), 其他学者也得到相同的结论(王政权等, 2000; Liu et al., 2018), 表明在日后研究LAI或其他变量的空间异质性时应给予方向性更多关注。