对于生活在沙漠中的动物来说，没有什么比一场大雨更令它们开心了。植物短短几天就能发芽、开花，吸引着数百公里外的动物前来觅食。蜜蜂便是最早感知到这一变化的动物之一，因此干旱地区的蜜蜂数量比热带地区还要多。

这是中国科学院动物研究所（以下简称中科院动物所）及合作单位最近的一项研究成果，研究人员通过对2万多种蜜蜂的数据分析，绘制出了全球首张蜜蜂类昆虫多样性分布图。11月20日，相关论文在线发表于《当代生物学》。

蜜蜂分布数据缺失

蜜蜂类昆虫在全球有2万多个物种，是给包括农作物在内的有花植物授粉的主力军。近年来，随着气候变化、栖息地被破坏等，蜜蜂物种多样性和部分物种种群数量在不断减少。

论文合作者、新加坡国立大学生物科学教授John Ascher说：“提起蜜蜂类昆虫，人们或许知道蜜蜂、熊蜂等，事实上蜜蜂类昆虫物种比鸟类和哺乳动物的总和还多。”而对于研究者和公众来说，目前并没有足够系统、全面的数据可供参考。

“此前的物种数据较为零散，而且主要集中在少数几个开展了大规模研究的国家。不少发展中国家的记录和数据都很有限。”论文的另一合作者、中国科学院西双版纳热带植物园副研究员Alice C Hughes说。

截至目前，我国统计已知蜜蜂类昆虫有1400余种。“相关数据主要来自标本馆馆藏定名物种，信息分布也不均衡。要研究中国蜜蜂，有必要先了解全球多样性及分布格局。”论文通讯作者之一、中科院动物所研究员朱朝东说。

全球蜜蜂类昆虫物种丰度分布图

建立全球蜜蜂丰度模型

“我们希望尽可能多地收集数据，只有将分布范围搞清楚，才能有针对性地研究保护措施。”论文第一作者、中科院动物所博士后Michael C. Orr告诉《中国科学报》。

在过去20年里，John Ascher专注于建立蜜蜂分布图，每发现一处新的分布地区，他都会将其更新在生物多样性门户网站Discover Life上。目前，他已记录了2万多种蜜蜂的分布信息。

基于这些数据，结合全球生物多样性信息网络等5个数据库信息、iNaturalist等在线数据库资源中的580万条自然观察蜜蜂记录，研究人员进行了核实比较和统计分析，利用MaxEnt等软件最终建立了全球蜜蜂丰度模型。“许多蜜蜂标本的分布、种类等信息都已缺失，需要分类学专家重新鉴定、核实录入，这也是研究过程中最大的困难。”Orr说。

不按常理“安家”的蜜蜂

对于许多动植物来说，多样性随着纬度的降低而增加，热带地区的物种最丰富。这被称为“纬度梯度”。研究人员发现，蜜蜂类昆虫是一个例外，其远离极点的物种更多，而靠近赤道的物种更少，遵循着“双峰纬度梯度”模式。

“树木给蜜蜂提供的食物有限。当沙漠下雨时，会出现不可预知的大规模开花，可以覆盖整个区域。所以森林和丛林中的物种比干旱的沙漠环境要少得多。”Orr说，由于沙漠中气候、环境变化频繁，每年的资源分布都不相同，可能会有更多的新物种。

据了解，此前已有研究提出过蜜蜂“双峰纬度梯度”假设，但由于相关数据缺失，很难被验证。而本项研究利用蜜蜂类昆虫的全球本底数据，证实了这一假设。

Orr说：“未来希望建立更高分辨率的蜜蜂物种丰度模型。这有利于研究蜜蜂类传粉昆虫保护问题。”

相关论文信息：

（文后附英文）

2020年

渠道越来越多，蜂蜜依然销售不理想？

“绿色通道”时有时无，何时能完全“绿色”？

蜂友群交流越来越多，有效信息却不多？

蜜蜂非常规大量死亡，维权之路如何走？

假蜜市场“依旧繁荣”，好蜂蜜路在何方？

转地养蜂是“路途”辛苦，还是“进场”辛苦？

养蜂有困难，难寻专业系统的学习平台？

看清中国蜂业

看《中国蜂业》

好消息 好消息 好消息

《中国蜂业》杂志社与博看网联合开发

《中国蜂业》电子版

电子版每期

仅需3元

博看网是中文畅销报刊在线阅读平台。其商业模式为:从报刊社购买传统期刊网络版权，对内容进行数字化加工，根据传统期刊出版日期定期更新至统一的网络平台，进行集中或分散销售，供读者免费或付费阅读。

目前博看网已经与4000多家期刊签约，每天有50-100本期刊上线，其信息更新量超过新浪、腾讯等门户网站。

博看网侧重收录发行量大的品牌畅销期刊，如《中国企业家》《中国新闻周刊》等。目前，博看网签约杂志的数量之多、品质之优已经使其内容质量和信息总量远超其他网站，形成了独一无二、不可复制的行业优势，目前已成为全球第一中文报刊网站。

2020年，

《中国蜂业》与博看网签约，

实现了期刊电子化。

再看一篇：

《中国蜂业》杂志是由中华人民共和国农业部主管，中国农业科学院蜜蜂研究所、中国养蜂学会主办的养蜂科技刊物，1934年1月创刊。是中国期刊方阵双效期刊、中国农业核心期刊、中国学术期刊综合评价数据库统计源期刊、全国优秀科技期刊。

点击左下角原文链接可购买《中国蜂业》杂志

Summary

Insects are the focus of many recent studies suggesting population declines, buteven invaluable pollination service providers such as bees lack a modern distributional synthesis. Here, we combine a uniquely comprehensive checklist of bee species distributions and >5,800,000 public bee occurrence records to describe global patterns of bee biodiversity. Publicly accessible records are sparse, especially from developing countries, and are frequently inaccurate throughout much of the world, consequently suggesting different biodiversity patterns from checklist data. Global analyses reveal hotspots of species richness, together generating a rare bimodal latitudinal richness gradient, and further analyses suggest that xeric areas, solar radiation, and non-forest plant productivity are among the most important global drivers of bee biodiversity. Together, our results provide a new baseline and best practices for studies on bees and other understudied invertebrates.

Keywords

insects

Hymenoptera

Apoidea

Anthophila

biodiversity

latitudinal gradient

species richness

bimodal gradient

drivers of diversity patterns

Introduction

Insects are reportedly declining at alarming rates worldwide, yet we do not understand even the most basic elements of their distributional dynamics.1Despite their importance, knowledge of insect biodiversity remains remarkably poor; the sheer number of species and the difficulty of identifying them preclude typical monitoring approaches, and the requisite funding is lacking.2,3Consequently, millions of museum specimens await identification or even formal description, remaining inaccessible to researchers.

Understanding insect distribution is key to evolutionary studies of origin and diversification, as well as ecological or conservation-oriented studies of how specific groups will respond to threats such as climate change or other human-induced phenomena.4,5In light of this, building and sharing our knowledge of insect distribution is one of the greatest, most important challenges that biologists and conservationists face, but the challenges of studying insects mandate the study of representative areas or specific groups.

As ecologically and economically invaluable pollinators, bees represent an ideal case study.4,6,7,8However, comprehensive analyses of bee distribution are nearly non-existent, with most focusing on limited regions9,10,11or site-based studies.12,13Well-known, eusocial bumblebees (Bombus) and the less-studied, solitary polyester bees (Colletinae) are exceptions.14,15However, these groups comprise 16). Furthermore,Bombusdominate at higher latitudes and elevations, whereas Colletinae are more species rich in xeric areas, suggesting that neither alone can represent overall bee biodiversity.

Those few efforts that explore worldwide bee distribution are descriptive, reliant on comparisons between small, well-sampled areas such as Palm Springs and Riverside in California.17Nonetheless, some general patterns have been hypothesized: bee species richness is highest in relatively xeric areas while tropical environments, famed for extraordinary insect species richness, have few.17This leads to a bimodal latitudinal gradient in some bee groups.15,18,19However, data remain limited for testing reported global trends, and supposed bimodal latitudinal gradients of other Hymenoptera are uncertain due to sampling biases and taxonomic under-description,20leaving few documented examples.21To date, these hypotheses remain untested for bees globally. The primary cause of this bee distribution knowledge gap is insufficient reliable occurrence data,22although the analytical and taxonomic expertise required have also precluded exhaustive analysis of bee distribution.

Here, we map and model the known distribution of bees based on a uniquely comprehensive checklist collated from specimens, verified observations, and published records, and quantitatively compare this to occurrence data from five public databases.16In doing so, we reveal the biases of public bee occurrence data and provide best practices for future analyses. By combining multiple, mutually informative data sources, we generate the most comprehensive assessment of global bee distribution, delimiting world hotspots of bee species richness. We then assess the drivers of these patterns and, in turn, use these predictions to model bee richness worldwide.

Results

Public Database Cleaning and Comparison to Checklist

Of the 5,857,811 occurrence records compiled, under 16% (907,001) passed all filters (Table 1). When excluding duplicate removal steps, which are not indicative of error and constituted 75%–90% of records (and also included duplicates between the databases;Table S1), there is an overall error rate span of 1%–8% for the datasets. 43,857 synonyms were compiled for the world total of 20,555 bee species, and 10,724 records were corrected across 6,340 species (Table S2). Although the East-West hemispheric check detected negligible error rates(