云冷杉针阔混交林叶面积指数的空间异质性

叶面积指数(Leaf area index,LAI)是森林生态过程的关键参数和描述森林冠层结构的重要指标。用半球摄影技术对吉林省汪清林业局金沟岭林场的10块1 hm2云冷杉针阔混交林的LAI进行测定,采用地统计学的半变异函数和普通克里格法对研究区的LAI的空间异质性进行了分析。结果表明:10块样地的10 m×10 m小样方内以及样地间的LAI离散程度较小,但分布有一定的规律。LAI的空间相关性存在不确定性,可由线性、孔状、高斯3种模型有效的描述,空间相关性存在的尺度范围变异大,中等和强空间相关性的影响范围在15—155m之间,10块样地LAI的空间相关性的影响范围平均为65.637m。普通克里格插值结果显示,LAI的空间分布呈明显的条带状和斑块状的梯度变化。LAI与林分平均胸径、林分平均高和林分平均冠长呈显著负相关,与林分平均冠幅、林分密度以及树种个数呈显著正相关。研究结果可为不同尺度云冷杉针阔混交林LAI的估计提供依据。

Spatial heterogeneity of leaf area index of mixed spruce-fir-deciduous stands in northeast China

Leaf area index (LAI) is defined as half the total leaf area per unit horizontal ground surface area. LAI is a significant attribute of forest ecosystem that controls, in part, physiological processes such as photosynthesis, transpiration, and leaf maintenance respiration, as well as physical processes including canopy water interception, evaporation, and light attenuation. There are several procedures to estimate LAI, both direct and indirect. However, the direct methods are slow and labor intensive. Although remote sensing potentially offers the opportunity to access features of ecological phenomena on a large scale, the scaling effect in retrieving LAI is a significant problem because of the heterogeneity in pixels, which can be well described using a semivariogram. In this study, we monitored LAI of ten mixed spruce-fir-deciduous stands in Wangqing Forestry Bureau, Jilin Province, by using a 10 m × 10 m grid in 1 hm² plots and digital hemispherical photography (DPH), the most commonly implemented method in forest research. In total, we obtained 1000 LAI values in these plots. The spatial heterogeneity of the LAI was explored using geostatistical analysis. We found that the average LAI in mixed spruce-fir-deciduous stands was 2.410, ranging from 1.265 to 4.029. LAI had low variability within and among plots, with the coefficient of variation ranging from 9.981% to 19.440%. Linear, hole, and Gaussian models were found to best describe the spatial autocorrelation of LAI. The coefficients of determination of these models were between 0.643 and 0.880. The LAIs had strong, moderate and weak spatial dependence according to their structure ratios. The ranges of these semivariogram models varied greatly, from 15 to 155 m, with an average of 65.637 m. One reason for the large range was the large homogeneous areas at the scale that were longer than the lags for which the semivariograms were calculated. Ordinary kriging was used to analyze the spatial distribution of LAI and showed apparent belt- and spot-shaped massive gradient changes. The relationship between LAI and stand variables was examined. LAI exhibited a significantly negative relationship with average diameter at breast height, average height, and average crown length, but showed a significantly positive relationship with average crown width, stand density, and tree species richness. LAI was closely related to tree species, and plots with few coniferous species had low values. Some unknown factors also contributed to the pattern of LAI in these plots. The results of this study provide a reference for estimating LAI at stand and large scales.