图1 陕西佛坪巴山木竹永久样地位置示意图

Fig. 1 Location of Bashania fargesii permanent plot at Foping, Shaanxi

图2 巴山木竹笋的空间点格局(2002-2008年, 2007年除外)。实线代表样方内每根巴山木竹笋邻近个体数目的观测值减去(如果符合)随机分布时的期望邻近个体数目。在t尺度上, L*(t) = 0, 则符合随机分布, L*(t) ≠ 0, 则偏离随机分布; L*(t)在95%置信区间(Monte-Carlo模拟200次, 用虚线表示)以上, 显著聚集分布(P<0.05); L*(t)在95%置信区间以下, 显著均匀分布(P <0.05)。第一个峰值的t值是斑块长度, 描述聚集的空间尺度; 第一个峰值的高度用来测度拥挤度。

Fig. 2 Spatial point pattern of new shoots of Bashania fargesii from 2002 to 2008, except 2007. The solid line represents the observed number of neighbors per Bashania fargesii new shoot in the plot minus the expected number of neighbors if the new shoots were distributed randomly. At scale t, if L*(t) = 0, the new shoots are distributed randomly; if L*(t) ≠ 0, the distribution of new shoots deviate from random. If L*(t) is above the 95% confidence level of L*(t) from 200 Monte Carlo simulations of the data, shown by the dotted line, the new shoots are significantly clustered (P<0.05); if L*(t) is below the 95% confidence level, the new shoots are significantly regular (P<0.05). The value of t at the first peak in the L*(t) plot is patch length, describing the characteristic spatial scale of clustering. The height of the peak is used as a measure of crowding.

图3 被大熊猫取食的巴山木竹笋空间点格局(2002-2008年, 2007年除外)。实线代表样方内每个被大熊猫取食笋邻近个体数目的观测值减去(如果符合)随机分布时的期望邻近个体数目。在t尺度上, L*(t) = 0, 则符合随机分布, L*(t)≠0, 则偏离随机分布; L*(t)在95%置信区间(Monte-Carlo模拟200次, 用虚线表示)以上, 显著聚集分布(P <0.05); L*(t)在95%置信区间以下, 显著均匀分布(P <0.05)。第一个峰值的t值是斑块长度, 描述聚集的空间尺度; 第一个峰值的高度用来测度拥挤度

Fig. 3 Spatial point pattern of new shoots of Bashania fargesii eaten by giant panda from 2002 to 2008, except 2007. The solid line represents the observed number of neighbors per Bashania fargesii new shoot eaten by giant panda in the plot minus the expected number of neighbors if the new shoots were distributed randomly. At scale t, if L*(t) = 0, the new shoots are distributed randomly; if L*(t) ≠ 0, the distribution of new shoots deviate from random. If L*(t) is above the 95% confidence level of L*(t) from 200 Monte Carlo simulations of the data, shown by the dotted line, the new shoots are significantly clustered (P<0.05); if L*(t) is below the 95% confidence level, the new shoots are significantly regular (P<0.05). The value of t at the first peak in the L*(t) plot is patch length, describing the characteristic spatial scale of clustering. The height of the peak is used as a measure of crowding.

图4 巴山木竹发笋与大熊猫取食之间的空间关系(2002-2008年, 2007年除外)。实线代表样方内每个巴山木竹笋邻近被大熊猫取食笋数目的观测值减去被大熊猫取食笋(如果符合)随机分布时的期望数目。在t尺度上, L12*(t) = 0, 则两类点相互独立, L12*(t) ≠ 0, 则两类点不相互独立; L12*(t)在95%置信区间(Monte Carlo模拟200次, 用虚线表示)以上, 两类点显著正相关(P<0.05); L*(t)在95%置信区间以下, 两类点显著负相关(P<0.05)。第一个峰值的t值描述相关的尺度。

Fig. 4 Spatial correlation between new shoots of Bashania fargesii and giant panda herbivory from 2002 to 2008, except 2007. The solid line represents the observed number of Bashania fargesii new shoots eaten by giant panda per new shoot in the plot minus the expected number of neighbors if the new shoots eaten by giant panda were distributed randomly. At scale t, if L12*(t)=0, the new shoots and new shoots eaten by giant panda are independent; if L12*(t)≠0, the distribution of the two kinds of points are not independent. If L12*(t) is above the 95% confidence level of L12*(t) from 200 Monte Carlo simulations of the data, shown by the dotted line, the two kinds of points are significantly postively associated (P<0.05); If L12*(t) is below the 95% confidence level, the two kinds of points are significantly negatively associated (P<0.05). The value of t at the first peak in the L12*(t) plot describes the scale of association.

图5 巴山木竹笋产量与大熊猫取食的年际动态

Fig. 5 Dynamics of Bashania fargesii new shoots and giant panda herbivory