Variation in trophic cascade strength is triggered by top-down process in an ant-wasp-fig systemVariation in trophic cascade strength is triggered by top-down process in an ant-wasp-fig system

R code and Original data for:Figure 2: Number of wasps from branches with weaver ants or without ants in the cool-dry (A), hot-dry (B), and rainy (C) seasons. Species composition is shown on the top of bar plot for each season, left pie (with ants), right pie (ants excluded). Syconia in the cool-dry season in ant exclusion treatments were all aborted. Linear mixed-effects models were fitted, with the number of fig wasps as the response variable, with treatment (trees with or without Oecophylla smaragdina ) as the fixed factor, tree as a random factor. * (P < 0.05), ** (P < 0.01), *** (P < 0.001), NS = not significant. CF = Ceratosolen fusciceps; SM = Sycophaga mayri; ST = S. testacea; SA = S. agraensis; AS = Apocrypta sp.; AW = A. westwoodi. Columns and bars show mean ± SD. # represent pollinating fig wasp. Numbers under the pie represent replications of syconium sampled.Figure 3: Number of seed from branches with weaver ants or ants excluded in cool-dry, hot-dry, and rainy seasons. * (P < 0.05), *** (P < 0.001). Numbers in the box represent replications of syconium sampled.Figure 4: Mean (±SD) interaction strength (A) of weaver ant, Oecophyla smaragdina, on the gall making fig wasp, Sycophagus mayri, and the strength of the trophic cascade (B) in different seasons. * (P < 0.05), *** (P < 0.001). Numbers in the brackets represent replications of with ants/ant exclusion branchlets pairs.Figure 5: Box plot of the number of weaver ant nests (A), number of weaver ants foraging in fig-bearing branchlets (B), foundress number (C), and oviposition scars of non-pollinating fig wasps (D) in the fruits of Ficus racemosa after receptive stage of Ficus racemosa in different seasons. Different letters represent significant differences between each season, pairwise comparisons conducted using Tukey and Kramer test. Each circle represents a data point. Figure 6: Number (mean ± SD) of weaver ants, Oecophyla smaragdina (A), pollinating fig wasp activity, Ceratosolen fusciceps (B), and gall making fig wasp, Sycophagus mayri activity (C) from 6:00 to 20:00 in different seasons. Symbols represent means and drop lines show the standard deviation. Different letters represent significant differences between each season, significance derived using nonparametric rank-based analysis of longitudinal data with ANOVA-type statistics (ATS).

本数据集配套的R代码与原始数据对应如下：
图2：冷旱季（A）、热旱季（B）与雨季（C）中，有织叶蚁栖息与无蚁栖息的枝条上的榕小蜂数量。各季节柱状图顶部标注了物种组成：左侧饼图为有蚁组，右侧饼图为无蚁组。冷旱季无蚁处理组中的榕果（syconia）全部败育。本研究采用线性混合效应模型，以榕小蜂数量为响应变量，处理方式（有或无黄猄蚁（Oecophylla smaragdina）的植株）为固定因子，植株为随机因子。显著性标记：*（P < 0.05）、**（P < 0.01）、***（P < 0.001），NS表示无显著性差异。物种缩写对应：CF为角胸榕小蜂（Ceratosolen fusciceps）；SM为马来榕瘿蜂（Sycophaga mayri）；ST为S. testacea；SA为S. agraensis；AS为Apocrypta sp.；AW为A. westwoodi。柱状图展示均值±标准差（SD）。# 代表传粉榕小蜂。饼图下方数字为采样榕果的重复数。
图3：冷旱季、热旱季与雨季中，有织叶蚁栖息与无蚁栖息的枝条上的种子数量。显著性标记：*（P < 0.05）、***（P < 0.001）。箱体内数字为采样榕果的重复数。
图4：不同季节下，织叶蚁（Oecophyla smaragdina）对造瘿榕小蜂马来榕瘿蜂（Sycophagus mayri）的交互强度（A），以及营养级联效应强度（B）的均值（±SD）。显著性标记：*（P < 0.05）、***（P < 0.001）。括号内数字为有蚁组/无蚁组枝条对的重复数。
图5：不同季节下，垂叶榕（Ficus racemosa）雌花期（receptive stage）结束后，其榕果中织叶蚁蚁巢数量（A）、栖息于挂果枝条的织叶蚁觅食个体数（B）、寻巢雌蜂（foundress）数量（C）以及非传粉榕小蜂产卵痕数量（D）的箱线图。不同字母代表各季节间存在显著性差异，两两比较采用Tukey-Kramer检验。每个圆点代表一个数据点。
图6：不同季节下，6:00至20:00时段内，织叶蚁（Oecophyla smaragdina）数量（均值±SD，A）、传粉榕小蜂角胸榕小蜂（Ceratosolen fusciceps）活动强度（B）以及造瘿榕小蜂马来榕瘿蜂（Sycophagus mayri）活动强度（C）。符号代表均值，误差垂直线代表标准差（SD）。不同字母代表各季节间存在显著性差异，显著性通过基于秩的非参数纵向数据分析结合ANOVA型统计量（ATS）计算得出。