By Paul H. Williams, Jiaxing Huang, and Jiandong An
Bumble bees are well known for being among the most important pollinators in the world’s north-temperate regions. Perhaps more surprisingly, half of the world’s bumble bee species are concentrated in just one country, China. With an area only slightly smaller than the U.S., China has almost three times as many species.
Understanding the bumble bees of China is more difficult because the mountains are high and the type specimens far away. So despite much invaluable work over many years, China also has the least well known bumble bee fauna worldwide. For the last 10 years, we have been trying to address this — to discover the full richness of China’s bumble bee diversity, to map the species’ distributions, to assess their conservation status, and to assess their potential for increasing crop yields.
Sorting out China’s bumble bees is not simply a question of matching a bee’s appearance with published descriptions or with museum specimens. Bumble bees have been described as morphologically homogeneous compared with other bees. However, they do have strikingly varied color patterns of the “hair,” although many analyses show that even single bumble bee species can be impressively variable in color pattern, not just locally, but more especially among different geographical regions. In addition, many of the different species within the same region can look identical. Despite this, people still try to use color patterns to identify them, even when describing new species. This is like trying to study plant taxonomy by looking just at leaves, without flowers, let alone DNA. It can lead to problems.
In 1998, an analysis of the rates of discovery of all the bumble bee species accepted at the time showed that rates varied widely around the world, with recent rates of discovery being highest in the Oriental Region, including China. This pattern is complicated by intraspecific variation. When people were focusing on the variation in color patterns, they published far more names for supposedly different species than were accepted as “good” species when those groups were subsequently revised in more detailed morphological studies. The result in 1998 was an average of more than 10 names per species. This illustrates the pitfalls of simply going out into the field and naming something that appears to be different. So we are still in need of reliable identifications and names to allow the information on species to be linked together and searched easily.
Since then, work on Chinese bumble bees has been gathering pace. Our team has made the project a priority for the last decade as a joint endeavor among institutes across China. This has been coordinated by the CAAS Institute of Apicultural Research in Beijing, with taxonomic research by the Natural History Museum in London, funded largely through grants from China. With the help of enthusiastic students, we have been surveying bumble bees from new sites across China every summer. We seek to follow best practice in making every step in the work transparent and accountable — by collecting reference voucher specimens, labeling each with a unique specimen-identifier number together with precise GPS location and elevation data, identifying species as far as possible using consistent and repeatable standards, and data-basing all of this information so that it can be easily accessed, queried, and mapped. We expect to have close to 50,000 fully data-based specimens in the project’s reference collection by the end of 2016. For identification, morphological comparisons have been an important start, although now we can use DNA data to discover more about the evolutionary relationships among the bumble bees we see. We are making progress with DNA sequencing for at least the most problematic groups, which should help to support better-informed interpretations. This genetic evidence is leading, on the one hand, to synonymizing some taxa with very different color patterns as parts of some highly variable species, and, on the other hand, to discovering new cryptic species that scarcely differ in color and morphology.
In 2007, soon after the beginning of this project, we had a list of 108 bumble bee species for China. Now in 2016, we have a provisional list of 130 species. This is exactly half of the currently estimated total number of bumble bee species world-wide. A few of the new species have never been recorded before. But most have been added because, although they had been considered previously as parts of known species, recent DNA studies now support them as separate species in the currently accepted sense of evolutionarily-independent lineages. This kind of work is never finished, as information changes. Frustrating in some ways, such shifting ground is a necessary consequence of progress in ideas on the nature of species, in analytical techniques, and in data compilation.
As a result of both the recent increase in survey efforts across China and of the changing concepts of species, we have added to the faunal lists for nearly all of the provinces. This shows that information has improved across the entire breadth of the country. There are most new records for some of the more mountainous provinces in northern China (especially Ningxia and Gansu).
The new records provide an improved quantitative estimate of the pattern of bumble bee species’ diversity across China. It had long been known in broad terms that the mountains around the edge of the Qinghai-Tibetan plateau (particularly in Sichuan and Gansu) are the principal hotspot for bumble bee species’ diversity worldwide. In contrast, the lower mountains and lowlands of southern and eastern China are warmer and have fewer of these cool-climate-loving bees.
However, the pattern in the species’ richness map (above) is complicated by variation in the sizes of the provinces, especially because the largest provinces are all concentrated together in the north and west of China. To reduce the effect of area, we can map any excess (or deficit) in species’ richness relative to the China-wide relationship between species’ richness and province size. This map shows that the provinces with the largest excess species’ richness for their size are those along the eastern edge of the Qinghai-Tibetan plateau, and to a lesser extent those provinces to the east along the mountains of North China. Mountain regions may have richer bumble bee faunas for several reasons: mountains have a greater variety of habitats in the larger range of elevational zones, each with different faunas; the high subalpine meadows tend to be especially rich in the cool-adapted bumble bee species; and areas with high relief have been suggested to facilitate the speciation process. One of the largest provinces (Niemenggu in the north) has slightly fewer species than expected, probably because it includes large areas of desert that are inhospitable to bumble bees. Perhaps surprisingly, the same is not true of Xinjiang in the northwest, also with extensive deserts, but only because the fauna is augmented with the rich and distinctive Altai and Tian Shan mountain faunas shared with Russia, Mongolia, and Central Asia.
Ultimately the Bumblebees of China initiative is aiming for practical applications, including selecting the best indigenous bumble bee species for rearing to improve crop pollination. Currently, four species have been selected by this project and are being reared through their complete life cycles at the institute in Beijing. They are now being used in trials with fruit crops on local farms. But one of our first aims is to provide identification tools. We are in the process of writing keys to help with the identification of all of the bumble bee species in China, and we are working on mapping their distributions. We hope to complete a preliminary atlas of the bumble bees of China in 2017. We will then be in a position to assess all of the species for their IUCN Red List threat status.
Paul Williams is Visiting Professor at the CAAS Institute of Apicultural Research (IAR), Beijing, China, and a Research Entomologist at the Natural History Museum, London, UK, working in collaboration with Dr. Jiaxing Huang and Professor Jiandong An at the IAR in Beijing.