Data from: Frugivore species richness influences dietary specialisation and network properties in Asian wet tropical forests

We examined how species richness in local plant-frugivore assemblages influences frugivore fruit choice behaviour, their dietary niche and the organisation of the plant-frugivore interaction network. Description of the data and file structure Taxa Coverage: Fleshy-fruited woody plants and avian frugivores. Geographic Coverage: Namdapha Tiger Reserve (27°23′–27°39′N and 96°15′–96°58′E), Pakke Tiger Reserve (26°54–27°16′N and 92°36′–93°09′E), Anamalai Tiger Reserve (10°12′–10°35′N and 76°49′–77°24′E), South Andaman (11°07′–12°15′N and 92°30′–92°50′E), Narcondam (13°30′N and 94°38′E) among the Indian sites, and Hala-Bala Wildlife Sanctuary (5°44′–5°57′N and 101°46′–101°51′E) in Thailand. Temporal Coverage: 2016–2024. Field data was collected for different sites by:1. Namdapha Tiger Reserve: Rintu Mandal, Himanshu Lad2. Pakke Tiger Reserve: Saniya Chaplod, Rohit Naniwadekar3. Anamalai Tiger Reserve: Rintu Mandal, Natasha Desai, Rasika Kadam, Vignesh Chandran4. South Andaman: Arpitha Jayanth, Aditya Gadkari, Vignesh Chandran5. Narcondam: Rohit Naniwadekar, Abhishek Gopal, Sartaj Ghuman, Navendu Page6. Hala-Bala Wildlife Sanctuary: Vatcharavee Sriprasertsil More details about the data can be obtained from R. Naniwadekar from the Nature Conservation Foundation and from (Naniwadekar et al., 2019, 2021, 2025). File Descriptions: Data filesThere are 18 data files in this dataset:1. dat_sxy_anamalai.csv2. dat_tp_anamalai.csv3. dat_p_anamalai.nex4. dat_sxy_andaman.csv5. dat_tp_andaman.csv6. dat_p_andaman.nex7. dat_sxy_namdapha.csv8. dat_tp_namdapha.csv9. dat_p_namdapha.nex10. dat_sxy_narcondam.csv11. dat_tp_narcondam.csv12. dat_p_narcondam.nex13. dat_sxy_pakke.csv14. dat_tp_pakke.csv15. dat_p_pakke.nex16. dat_HMSC_variance_partitioning.csv17. dat_dPrime_NormalisedDegree_combined.csv18. dat_NetworkMetrics_original_and_rarefied.csv The first 15 data files are data files required for running the Hierarchical Modelling of Species Communities (HMSC) for five of our study sites: Anamalai, Andaman, Namdapha, Narcondam, and Pakke. The SXY files contain species and environmental data, TP files contain species traits, and P files correspond to time-calibrated phylogenies of the species of interest. The details of the columns corresponding to these files are as follows: SXY files (“dat_sxy_SiteName.csv”)1. tree_family - refers to the plant family the focal plants belonged to in our study 2. tree_code - unique code given to focal plants in each study site3. tree_sp - refers to the plant species the focal plants belonged to in our study4. seed_size - seed size of the focal plant (small, medium, or large)5. fruit_width - mean width of ripe fruits in mm of the focal plant6. pulp_lipid - lipid content in the pulp of ripe fruits of the focal plant (low, medium, or high)7. log_fcs - ripe fruit crop size of the focal plant (converted to natural log scale)The remaining columns in each SXY file correspond to the presence (1) or absence (0) of the avian frugivore species included in the analysis for a particular site. TP files (“dat_tp_SiteName.csv”)1. bird_sp - avian frugivore species included in the analysis for a particular site2. degree_frugivory - degree of frugivory (percentage of fruits in diet)3. beak_width - beak width of avian frugivores in mm4. hwi - hand-wing index of each frugivore species P files (“dat_p_SiteName.nex”)A time-calibrated phylogenetic tree of the avian frugivore species included in the analysis for a particular site. The data set “dat_HMSC_variance_partitioning.csv” contains the outputs of the HMSC models, specifically, the percentage of variation explained by each fixed effect included in the model for five of our study sites. The details of the columns are as follows:1. Predictors - fixed effects included in the model2. Percentage_variation_explained - percentage of variation explained by each fixed effect3. Site - name of the study sites (five)4. Sp_richness - the total number of avian frugivore species recorded in each study site during the entire study period (and not the number of frugivore species included in the model) The data set “dat_dPrime_NormalisedDegree_combined.csv” contains the specialisation measures (Normalised Degree and Blüthgen’s d′) calculated for each avian frugivore species in our study using network analysis. The details of the columns are as follows:1. Frugivore_code - unique code given to each avian frugivore species2. Scientific_old - binomial names of avian frugivores as per https://www.birdtree.org/3. Scientific_updated - binomial names of avian frugivores as per Clement’s Checklist of Birds4. Normalised_Degree - Normalised Degree of each avian frugivore calculated using network analysis for each site5. D_prime - Blüthgen’s d′ of each avian frugivore calculated using network analysis for each site6. Site - name of the study sites (six)7. Sp_richness - the total number of avian frugivore species recorded in each study site during the entire study period The data set “dat_NetworkMetrics_original_and_rarefied.csv” contains the observed and rarefied values of four network-level metrics, namely, Connectance, Modularity, Specialisation, and weighted NODF for six plant-frugivore assemblages in our study. The details of the columns are as follows:1. Site - name of the study sites (six)2. Sp_Richness - the total number of avian frugivore species recorded in each study site during the entire study period3. Metric - name of the network metric4. Original_Metric_Value - the observed network metric values calculated using network analysis5. Rarefied_Mean - We randomly pulled 696 (the lowest number of interactions among our study sites) interactions from the original plant-frugivore interaction network for each study site to create the rarefied plant-frugivore interaction network, and calculated the aforementioned network metrics. We repeated the process 100 times for each site to calculate the mean network metric values.6. Rarefied_SD - Standard Deviation of the rarefied network metrics were calculated using the protocol mentioned above. CodeThere are three R scripts (.R files) in this dataset:1. code_hmsc.R2. code_correlation_glmm_dPrime_NormalisedDegree_SpRichness.R3. code_correlation_network_metrics_SpRichness.R The R script “code_hmsc.R” contains the R code to fit the model and run the HMSC analysis for each site (L1–L60), evaluate model convergence, compute variance partitioning, extract beta and gamma plots, and check for phylogenetic signal in the data (L65–L126), and finally plot the mean (and 95% CI) percentage variation explained by each predictor (L132–L162). The R script “code_correlation_glmm_dPrime_NormalisedDegree_SpRichness.R” contains R code to check correlations between the species-level metrics Normalised Degree and Blüthgen’s d′ and the site-level frugivore species richness, GLMM codes to model those species-level metrics as a function of site-level frugivore species richness, and plot the outputs. The R script “code_correlation_network_metrics_SpRichness.R” contains R code to check correlations between rarefied network metrics and site-level frugivore species richness, and plot the relationships.本研究探讨了局域植物-食果动物群落中的物种丰富度如何影响食果动物的果实选择行为、其饮食生态位，以及植物-食果动物互作网络的组织模式。 数据与文件结构说明 类群覆盖范围：肉质果木本植物与鸟类食果动物。 地理覆盖范围：印度境内的南达帕老虎保护区（Namdapha Tiger Reserve，27°23′–27°39′N，96°15′–96°58′E）、帕克老虎保护区（Pakke Tiger Reserve，26°54′–27°16′N，92°36′–93°09′E）、阿纳马莱老虎保护区（Anamalai Tiger Reserve，10°12′–10°35′N，76°49′–77°24′E）、南安达曼岛（South Andaman，11°07′–12°15′N，92°30′–92°50′E）、纳康达姆岛（Narcondam，13°30′N，94°38′E），以及泰国的哈拉-巴拉野生动物保护区（Hala-Bala Wildlife Sanctuary，5°44′–5°57′N，101°46′–101°51′E）。 时间覆盖范围：2016年–2024年。 各研究样地的野外数据由以下人员采集： 1. 南达帕老虎保护区：Rintu Mandal、Himanshu Lad 2. 帕克老虎保护区：Saniya Chaplod、Rohit Naniwadekar 3. 阿纳马莱老虎保护区：Rintu Mandal、Natasha Desai、Rasika Kadam、Vignesh Chandran 4. 南安达曼岛：Arpitha Jayanth、Aditya Gadkari、Vignesh Chandran 5. 纳康达姆岛：Rohit Naniwadekar、Abhishek Gopal、Sartaj Ghuman、Navendu Page 6. 哈拉-巴拉野生动物保护区：Vatcharavee Sriprasertsil 如需获取该数据集的更多详细信息，可联系自然保护基金会（Nature Conservation Foundation）的R. Naniwadekar，相关内容亦可参阅（Naniwadekar等，2019、2021、2025）。 文件说明 一、数据文件 本数据集共包含18个数据文件： 1. dat_sxy_anamalai.csv 2. dat_tp_anamalai.csv 3. dat_p_anamalai.nex 4. dat_sxy_andaman.csv 5. dat_tp_andaman.csv 6. dat_p_andaman.nex 7. dat_sxy_namdapha.csv 8. dat_tp_namdapha.csv 9. dat_p_namdapha.nex 10. dat_sxy_narcondam.csv 11. dat_tp_narcondam.csv 12. dat_p_narcondam.nex 13. dat_sxy_pakke.csv 14. dat_tp_pakke.csv 15. dat_p_pakke.nex 16. dat_HMSC_variance_partitioning.csv 17. dat_dPrime_NormalisedDegree_combined.csv 18. dat_NetworkMetrics_original_and_rarefied.csv 前15个数据文件为针对5个研究样地（阿纳马莱、南安达曼、南达帕、纳康达姆、帕克）运行群落物种分层建模（Hierarchical Modelling of Species Communities, HMSC）所需的数据文件。其中，SXY格式文件包含物种与环境数据，TP格式文件包含物种功能性状数据，P格式文件对应目标类群的时间校准系统发育树。各文件的列字段详情如下： SXY格式文件（命名规则为`dat_sxy_SiteName.csv`）： 1. tree_family：本研究中目标植物所属的植物科 2. tree_code：各研究样地中为目标植物分配的唯一编码 3. tree_sp：本研究中目标植物所属的植物物种 4. seed_size：目标植物的种子大小（分为小、中、大三个等级） 5. fruit_width：目标植物成熟果实的平均宽度（单位：mm） 6. pulp_lipid：目标植物成熟果实果肉的脂质含量（分为低、中、高三个等级） 7. log_fcs：目标植物成熟果序的结实量（已转换为自然对数尺度） 其余列字段对应分析中纳入的各鸟类食果动物物种在对应样地的出现情况（1表示存在，0表示不存在）。 TP格式文件（命名规则为`dat_tp_SiteName.csv`）： 1. bird_sp：本研究分析中纳入的鸟类食果动物物种 2. degree_frugivory：食果性程度（即饮食中果实所占的百分比） 3. beak_width：鸟类食果动物的喙宽（单位：mm） 4. hwi：各食果动物物种的手翼指数（Hand-wing Index, HWI） P格式文件（命名规则为`dat_p_SiteName.nex`）：对应研究样地中纳入分析的鸟类食果动物物种的时间校准系统发育树。 数据集`dat_HMSC_variance_partitioning.csv`包含群落物种分层建模（HMSC）的输出结果，具体为模型中各固定效应对变异的解释百分比，覆盖5个研究样地。其列字段详情如下： 1. Predictors：模型中纳入的固定效应变量 2. Percentage_variation_explained：各固定效应的变异解释百分比 3. Site：研究样地名称（共5个） 4. Sp_richness：整个研究周期内各研究样地记录到的鸟类食果动物物种总数（而非模型中纳入的食果动物物种数） 数据集`dat_dPrime_NormalisedDegree_combined.csv`包含通过网络分析计算得到的各鸟类食果动物物种的特化指标（归一化度（Normalised Degree）与布卢特根d'值（Blüthgen’s d′））。其列字段详情如下： 1. Frugivore_code：为各鸟类食果动物物种分配的唯一编码 2. Scientific_old：基于https://www.birdtree.org/ 命名的鸟类食果动物二项式学名 3. Scientific_updated：基于《克莱门茨鸟类名录》（Clement’s Checklist of Birds）更新的鸟类食果动物二项式学名 4. Normalised_Degree：通过各研究样地的网络分析计算得到的各鸟类食果动物的归一化度 5. D_prime：通过各研究样地的网络分析计算得到的各鸟类食果动物的布卢特根d'值 6. Site：研究样地名称（共6个） 7. Sp_richness：整个研究周期内各研究样地记录到的鸟类食果动物物种总数 数据集`dat_NetworkMetrics_original_and_rarefied.csv`包含本研究中6个植物-食果动物群落的4个网络水平指标的观测值与稀化后值，具体为连接度（Connectance）、模块化（Modularity）、特化度（Specialisation）以及加权NODF。其列字段详情如下： 1. Site：研究样地名称（共6个） 2. Sp_Richness：整个研究周期内各研究样地记录到的鸟类食果动物物种总数 3. Metric：网络指标名称 4. Original_Metric_Value：通过网络分析计算得到的观测网络指标值 5. Rarefied_Mean：我们从各研究样地的原始植物-食果动物互作网络中随机抽取696条交互（即所有研究样地中最少的交互数）以构建稀化后的植物-食果动物互作网络，并计算上述网络指标；每个样地重复该过程100次以计算网络指标的均值 6. Rarefied_SD：按照上述流程计算得到的稀化网络指标的标准差 二、代码文件 本数据集共包含3个R脚本文件（.R格式）： 1. code_hmsc.R 2. code_correlation_glmm_dPrime_NormalisedDegree_SpRichness.R 3. code_correlation_network_metrics_SpRichness.R 脚本`code_hmsc.R`包含用于为各研究样地拟合模型并运行群落物种分层建模（HMSC）分析的R代码（第1–60行）、模型收敛性评估代码、变异分区计算代码、提取β与γ绘图数据的代码，以及数据中系统发育信号的检测代码（第65–126行），最终用于绘制各预测变量解释的变异百分比的均值（及95%置信区间）（第132–162行）。 脚本`code_correlation_glmm_dPrime_NormalisedDegree_SpRichness.R`包含用于检验物种水平指标（归一化度与布卢特根d'值）与样地水平食果动物物种丰富度之间相关性的代码，以及用于将这些物种水平指标作为样地水平食果动物物种丰富度的函数进行建模的广义线性混合模型（Generalized Linear Mixed Model, GLMM）代码，同时包含结果可视化代码。 脚本`code_correlation_network_metrics_SpRichness.R`包含用于检验稀化后的网络指标与样地水平食果动物物种丰富度之间相关性的代码，以及关系可视化代码。