Scholarlink

<p>We investigated the influence of the data source on taxonomic composition analyses and species distribution model outputs for ants in the Espinhaço Range. Our objectives were to evaluate the taxonomic composition reported on a global repository (GBIF plus GABI) and regional compilations (PELD-CRSC plus Atlantic Ants), to compare stacked species distribution models (SSDMs) built from each source, and to examine climate–richness relationships using a combined model. We compiled occurrences, cleaned the data and then fitted SSDMs with five algorithms and using five bioclimatic variables selected after variance-inflation screening. We also examined richness maps and climate–richness relationships using Spearman correlations. Subfamily and species compositions diverged across repositories. The global dataset contained only ~10% of the records of the regional dataset, resulting in much lower richness magnitudes. Despite this, the models from both datasets were highly correlated. All models consistently indicated higher species richness in the southern Espinhaço, which was strongly associated with lower isothermality and higher precipitation. Despite the importance of global repositories, regional ones are essential for accurate estimates of ant composition and richness. Thus, integrating datasets improved coverage and highlighted significant knowledge gaps in central and northern Espinhaço.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Data source effects on the perceived distribution of ant diversity in the Espinhaço Mountain Range, Brazil

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月15日

查看全文
  • Davi Vilaça,
  • Inácio Gomes,
  • Ricardo Meireles,
  • Ricardo Solar,
  • Frederico Neves

摘要

We investigated the influence of the data source on taxonomic composition analyses and species distribution model outputs for ants in the Espinhaço Range. Our objectives were to evaluate the taxonomic composition reported on a global repository (GBIF plus GABI) and regional compilations (PELD-CRSC plus Atlantic Ants), to compare stacked species distribution models (SSDMs) built from each source, and to examine climate–richness relationships using a combined model. We compiled occurrences, cleaned the data and then fitted SSDMs with five algorithms and using five bioclimatic variables selected after variance-inflation screening. We also examined richness maps and climate–richness relationships using Spearman correlations. Subfamily and species compositions diverged across repositories. The global dataset contained only ~10% of the records of the regional dataset, resulting in much lower richness magnitudes. Despite this, the models from both datasets were highly correlated. All models consistently indicated higher species richness in the southern Espinhaço, which was strongly associated with lower isothermality and higher precipitation. Despite the importance of global repositories, regional ones are essential for accurate estimates of ant composition and richness. Thus, integrating datasets improved coverage and highlighted significant knowledge gaps in central and northern Espinhaço.