Biogeographic affiliation and centers of richness as predictors of elevational range-size patterns for Malesian flora

Our goal was to interrogate the idea that “mountain passes are higher in the tropics” by investigating ecological and biogeographic drivers of elevational range-sizes patterns among equatorial flora. We used herbarium records for sixty species-rich plant families, representing 18535 species total, to estimate distributions over a 4500 m elevational gradient. For each family, we estimated the change in average range-sizes with increasing elevation (i.e. Rapoport’s rule, abbreviated as ERR) and quantified 15 metrics of familial richness distribution, evolutionary age, and biogeographic affiliation. We visualized covariation across families using phylogenetic principal components analysis (pPCA). We then evaluated how family-level ERR slopes correlated with each metric individually, as well as when using multivariate techniques to reduce dimensionality. We hypothesized that if long term climate stability over millions of years promotes habitat specialization, then among taxa with longer-term tropical affiliations, we would expect smaller range-sizes within lowland forests, with greater range-size expansion towards higher elevations, expressed as a positive ERR slope. Conversely, variation in growing conditions should promote larger, relatively consistent, range-sizes at all sections of an elevational gradient, expressed as a neutral ERR slope. Our results support this corollary because of the dichotomy of ERR slopes observed in relation to the elevational distribution of richness and historical biogeographic positioning. We found that families with greater Sundaland endemism, or richness that was restricted to tropical lowland forests, had positive ERR slopes. Families with stronger Sahul affiliation, or montane centered richness, had shallower, neutral, or negative ERR slopes, as did clades with temperate origins. Families with Wallacea affiliation, broader latitudinal or elevational distributions, cosmopolitanism, greater richness, or older evolutionary age had mixed results. We conclude that the relative steepness of an ERR slope is an indicator of a taxonomic group’s tolerance of habitat variation and vulnerability to contemporary climate change.

Methods
This dataset includes species-level elevational ranges and occurrence distribution summaries for >31k plant species from Malesia, which includes the Malay Peninsula and islands of Sumatra, Borneo, Philippines, Java, Lesser Sunda Isles, Sulawesi, Maluki, and New Guinea (excluding the Bismarck archipelago). 

For details on methods, data-sources, data standardization practices, or nomencalture used for “Biogeographic history and centers of richness define elevational range-size patterns for Malesian flora” refer to main document and Supporting information. The data used in the main manuscript is a subset of this Dryad file, selecting only species with two or more observations and from a species-rich family.

Note, the original sources of data included more than a half-million entries from multiple data-sources, most origininating from GBIF.org. The Dryad data file does not include herbarium entries that were excluded due to being flagged as erroneous, non-relevant, or with information that we found to be lacking or conflicting. Elevation data is shown as meters above sea-level (m a.s.l); manually interpreted and converted from feet if necessary. In some instances, nomenclature used for family names has been adjusted to match the phylogenetic mega tree.

Extensive quality control efforts were made to prevent errors with elevation, location, or nomenclature entries, but there is always potential for improvement. This file is provided as-is. Refer to the corresponding author for additional information if needed.

我们的研究旨在通过探究赤道植物区系海拔分布范围格局的生态与生物地理驱动因子，验证“热带山地隘口更高”这一假说。我们共选取60个物种丰富的植物科的标本馆记录，涵盖总计18535个物种，以估算4500米海拔梯度上的物种分布。针对每个科，我们估算了平均分布范围随海拔升高的变化规律（即拉波波特法则（Rapoport’s rule），简称ERR），并量化了科级丰富度分布、演化年龄以及生物地理归属的15项指标。我们通过系统发育主成分分析（phylogenetic principal components analysis，pPCA）可视化了各科之间的协变关系。随后，我们分别评估了科级ERR斜率与各项指标的相关性，同时也采用降维的多元分析方法进行关联分析。

我们提出假设：若数百万年尺度的长期气候稳定性促进了生境特化，那么长期具有热带生物地理归属的类群，其低地森林内的分布范围应更小，而向高海拔区域的分布范围扩张程度更高，表现为正的ERR斜率。反之，生长条件的异质性应促使海拔梯度全段内的分布范围更大且相对稳定，表现为中性的ERR斜率。

我们的研究结果支持这一推论，因为观察到的ERR斜率二分性与类群的海拔丰富度分布和历史生物地理定位相关。我们发现，巽他陆架（Sundaland）特有性更高，或丰富度局限于热带低地森林的科，其ERR斜率为正。与萨胡尔板块（Sahul）关联更强、或丰富度以山地为中心的科，其ERR斜率更平缓、呈中性或为负值，具有温带起源的演化支同样如此。具有华莱士区（Wallacea）归属、更广纬度或海拔分布、广布性、更高丰富度或更古老演化年龄的科，其结果则存在差异。我们最终得出结论：ERR斜率的相对陡度可作为分类群生境变异耐受性以及当代气候变化脆弱性的指示指标。

方法
本数据集包含马来群岛地区（涵盖马来半岛以及苏门答腊、婆罗洲、菲律宾、爪哇、小巽他群岛、苏拉威西、马鲁古和新几内亚（不含俾斯麦群岛））超过31000个植物物种的物种级海拔分布范围及出现分布汇总数据。

关于研究方法、数据来源、数据标准化流程，以及“生物地理历史与丰富度中心定义了马来植物区系的海拔分布范围格局”一文所采用的命名规范，请参阅主文档及补充材料。本主手稿中使用的数据为该Dryad数字知识库（Dryad）数据集文件的子集，仅选取了具有2条及以上观测记录的物种，且来自物种丰富的植物科。

请注意，原始数据来源包含来自全球生物多样性信息设施（GBIF.org）等多个数据源的超过50万条记录。本Dryad数据文件未包含因被标记为错误、无关，或存在信息缺失/冲突而被排除的标本馆记录。海拔数据以米（m a.s.l.，即海拔米数）为单位，必要时已通过人工解读并转换自英尺单位。部分科的分类学名称已进行调整，以匹配系统发育大树形结构。

研究团队已开展严格的质量控制工作，以避免海拔、地理位置或分类学命名方面的错误，但仍存在进一步优化的空间。本文件按“现状”提供。如需获取更多信息，请联系通讯作者。