PLANT-POLLINATOR INTERACTIONS IN EAST ASIA: A REVIEW

Pollination studies in East Asia have been developing rapidly in recent decades. East Asia may provide important information on many aspects of plant-pollinator interactions because of the rich fauna and flora and highly heterogeneous environments that occur there. In this review, plant-pollinator interactions in East Asia were summarized. Bumblebees are important pollinators of many plant species in East Asia, as well as in Europe and North America. Native honeybees may also have important roles in pollination in East Asia. Bird pollination and hawkmoth pollination may be less common in East Asia than in North America. Geographic variation in pollination interactions is expected because several types of pollinators are rare or absent in some habitats or geographic regions. For example, specialized nectar-feeding vertebrates like sunbirds and pteropodid bats are absent from most of East Asia except for some areas in its southern part. Opportunistic nectar-feeding vertebrates may have important roles in pollination where specialized nectar-feeding vertebrates are absent. Human impacts on plant pollinator interactions are understudied in this region. However, climate change, habitat degradation, and invasive species may have negative impacts on plantpollinator interactions and thus plant reproductive success there. The information available on the plant-pollinator interactions in East Asia is still limited because many plant and pollinator taxa and many types of habitats are understudied.


INTRODUCTION
Animal pollinators play significant roles in the sexual reproduction of many flowering plants (Ollerton et al. 2011;Ratto et al. 2018). The ecological and evolutionary consequences of plant-pollinator interactions are geographically variable (Hiraiwa & Ushimaru 2017;Johnson et al. 2017;Ollerton 2017;Zanata et al. 2017). However, the worldwide variation in plant-pollinator interactions is still poorly explored (Ollerton 2012). For example, some globallevel studies of plant-pollinator interactions and plant reproductive ecology included no or very few East Asian examples due to the lack of available data (Ollerton et al. 2009;Ollerton 2012;Schleuning et al. 2012;Rech et al. 2016;Johnson et al. 2017;Zanata et al. 2017). Additionally, pollinators like hawkmoths and nectar-feeding birds have been poorly studied in East Asia (Johnson et al. 2017;Ren et al. 2018;Liu & Sun 2019).
Geographically, East Asia comprises China, the Korean peninsula, the Japanese archipelago, and Taiwan (Prescott 2015). The northern part of the region (central and northern China, the Korean Peninsula, and the main islands of Japan) is Palaearctic, the southern part of the region (southern China, Taiwan, and the Ryukyu Islands) is Oriental, and the Bonin Islands are part of the northern limits of Oceania (Udvardy 1975;Corlett 2004). For several reasons, East Asia could be an important temperate region of the Northern Hemisphere for use in comparative studies with North America and Europe, wherein the field of pollination biology is betterdeveloped. First, East Asia has remarkably rich and highly endemic flora and fauna (Zhu 2016;Tojo et al. 2017;Lu et al. 2018;Tang et al. 2018). For example, more than 30000 species of vascular plants have been recorded in China, and more than half of these species are endemic (Blackmore et al. 2015). East Asia acted as a refugium for many plant taxa because it has historically been covered by glaciers less than North America and Europe . Second, the environment of East Asia is highly heterogeneous. For example, the Qinghai-Tibet Plateau is the largest and highest plateau in the world, and the species richness of both plants and pollinators on the plateau and in the surrounding region is very high Sun et al. 2014). East Asia is physiographically more heterogeneous than eastern North America (Qian & Ricklefs 2000). The complex combinations of different geographic, climatic, and ecological factors that occur in East Asia may be associated with the high diversification of plants and animals there (Qian & Ricklefs 2000;Tojo et al. 2017). The heterogeneous environment in East Asia provide opportunities to study how plant-pollinator interactions differ across environmental gradients and among habitats. For example, islands in East Asia like Hainan Island, the Japanese archipelago, and Taiwan are good models to study how plant-pollinator interactions and plant reproductive characteristics can differ between islands and the mainland (Inoue 1993; Ling et al. 2017). Third, some plant genera have disjunctive distributions in East Asia and North America. These disjunctively distributed genera have more species in East Asia than in North America, and this component of the East Asian temperate flora is thus a good model system to understand the diversification of plants in temperate regions (Qian & Ricklefs 2000). Therefore, such genera provide opportunities to study how plant-pollinator interactions and the evolution of pollination systems differ among continents ).
The study of plant-pollinator interactions and plant reproductive ecology in East Asia has been developing rapidly in recent decades (Barrett 2015;Ren et al. 2018). Ren et al. (2018) summarized the pollination research done so far in China, and suggested future directions it could take. However, a comprehensive review focused on the characteristics of plant-pollinator interactions across the entire region of East Asia has not yet been published. Some reviews have been published that were concerned with specific areas of East Asia, including the Oriental Region in East Asia (Corlett 2004) and the Qinghai-Tibet Plateau . Some reviews of the pollination systems of East Asian Orchidaceae and Chinese Gesneriaceae were also previously published (Van Der Cingel 2001;Tang et al. 2014;Ling et al. 2017).
The present review aims to provide a general overview of the plant-pollinator interactions involving wild plants in East Asia. This review specifically highlights the diversity of pollinators and the plants with which they interact in this region. Although this review focuses on biotic pollination, the occurrence of mixed pollination systems involving pollination by both animals and the wind, called ambophily, is also noted. Additionally, the potential impacts of human activities on the plant-pollinator interactions in East Asia are discussed. Finally, I discuss future prospects for research on plantpollinator interactions in East Asia.
This review was not a formal systematic review or metaanalysis. Thus, the literature survey conducted herein was not very systematic. I selected representative papers published on East Asian pollination from 1956 to 2019. This time frame largely covered the development of research on plantpollinator interactions in East Asia (Tanaka 1996;Ren et al. 2018). Published studies were located using Google Scholar searches for keywords representing the regions (e.g., China), plant taxa (e.g., Rhododendron), and pollinator groups (e.g., hawkmoth) in East Asia. I also located additional studies by checking the references lists of the published East Asian pollination studies obtained. The studies citing the East Asian pollination studies obtained were also examined. Additionally, I located local publications that are not easily accessible. Studies focused on individual plant species were mostly referred to so that the pollination systems of particular plant taxa could be described. Some community-level studies of plant-pollinator interactions and books on floral ecology suggested the existence of many specialized pollination systems (e.g., Kato 2000;Corlett 2001;Tanaka 2009). However, these studies were generally not cited herein to describe the pollination systems of particular plant taxa because detailed studies are needed to predict the most effective pollinators of particular flowers (King et al. 2013;Funamoto & Ohashi 2017). The community-level pollination studies examined were thus mostly cited to describe trends in the flower-visiting habits of particular animal taxa and plantpollinator interactions in specific habitats.
I searched within various community-level studies to evaluate the relative abundances of different types of pollinators in various habitat types (Fig. 2). Studies providing the number of visits by different types of pollinators were selected. The number of visits by the different types of pollinators were then obtained from the published literature. Data were extracted from the tables, figures, and supporting information in published studies. When a study encompassed multiple years and sites, its data were pooled across these to find a value for each study. I categorized pollinators into 9 functional groups: Bee, Wasp/Sawfly, Diptera, Syrphidae, Non-syrphid Diptera, Coleoptera, Lepidoptera, Hemiptera, and Others. The Others group included various groups of pollinators, such as Orthoptera, Blattodea, spiders, birds, and ants. These categories generally followed those used in other pollination studies (Fenster et al. 2004;Nikkeshi et al. 2015;Zhang & He 2017). However, some functional groups defined in these studies were combined or divided herein because all functional groups were not always distinguished in the selected community-level studies. For example, Diptera was used in a study by Inoue & Endo (2006b), wherein Syrphidae and Non-syrphid Diptera were not distinguished. Wasp and sawflies were lumped together herein because nonbee hymenopterans were combined in a study by Zhao et al. (2016). Although species of Hemiptera are not usually considered pollinators (Wardhaugh 2015), Hemiptera was treated as an independent category herein because hemipterans are frequently observed on flowers in many communities.

Eusocial bees
The eusocial bees recorded from East Asia include bumblebees (Bombus), honeybees (Apis), and stingless bees (Trigona) (Michener 2007;Tadauchi et al. 2014). Bumblebees are distributed over most of East Asia (Michener 2007;Williams et al. , 2010Tadauchi et al. 2014), but they are absent on islands like the Izu Islands (except for Oshima Island), Bonin Islands, Ryukyu Islands (except for Yakushima Island), and the adjacent islands of Taiwan (Starr 1992;Inoue 1993;Tadauchi et al. 2014). Almost half of the world's bumblebee species are found in China (Williams et al. , 2017. The edge of the eastern Tibetan Plateau is one of the greatest hotspots of bumblebee diversity in the world, where 56 species of bumblebees have been recorded ). Native honeybees, such as Apis cerana and Apis florea, and managed honeybees (Apis mellifera) are distributed in the temperate and subtropical regions (Hepburn et al. 2005;Michener 2007;Tadauchi et al. 2014). Several species of native honeybees co-occur in southern China (Cui & Corlett 2016). Native honeybees are absent from the Bonin Islands and most of the Ryukyu Islands (Kato et al. 1999;Kato 2000;Tadauchi et al. 2014). Stingless bees have been recorded in southern China and Taiwan (Sakagami & Yamane 1984;Michener 2007).  (Yumoto 1986;Mizunaga & Kudo 2017). Habitat and flower preferences differ among bumblebee species. For example, the shorttongued species Bombus ardens prefers shallow tree flowers and appears in forests, whereas the long-tongued species Bombus diversus prefers deep-tubed flowers and appears on both open-land habitats and in forests (Ushimaru et al. 2008).
Stingless bees act as important pollinators of many plant species in Oriental Region (Corlett 2004). However, in East Asia, stingless bees are less frequently reported as flower visitors compared to the honeybees and bumblebees. They are reported as nectar and pollen robbers of several flowers (Zhang et al. 2005;Luo et al. 2008;Deng et al. 2015).
Pollinating gall midges of the genus Resseliella oviposit their eggs in flowers, and their larvae feed on the resin secreted from wounded floral tissues (Luo et al. 2017. One plant species' flowers are usually pollinated by one species of gall midge ).

Coleoptera
Species of beetles belonging to the Cerambycidae, Elateridae, Oedemeridae, and Scarabaeidae visit many flowers with generalized morphologies (Fig. 1A;Kuboki 1980;Kato et al. 1993;Kato 2000;Osada et al. 2003;Yang et al. 2017). Beetles can be effective pollinators of flowers. For example, flower chafers act as important pollinators of Magnolia obovata because they carry a higher proportion of outcrossed pollen than bumblebees do (Matsuki et al. 2008). Beetles may also act as pollinators when the activity of other types of pollinators is low. For example, Pidonia (Cerambycidae), a genus of diverse and abundant coleopteran flower visitors in Japanese temperate forests, can visit flowers in shady conditions (Kuboki & Shimamoto 1979;Kuboki 1980;Sakakibara et al. 1996). Staphylinid beetles were observed among the flower visitors of Symplocarpus renifolius, the flowers of which apparently mimic dung or carrion (Uemura et al. 1993).     (Okochi et al. 2006).

Plant-pollinator interactions in East Asia
The available information on plant-pollinator interactions in East Asia is very scattered. Thus, any generalizations drawn at this point must be highly tentative. Further studies are needed to reveal the general trends in plant-pollinator interactions in East Asia. However, some generalizations can be made to allow East Asian plant-pollinator interactions to be compared with those in other temperate regions in the Northern Hemisphere, like those in Europe and North America.
Bumblebees are important pollinators of many plant species in East Asia, as in Europe and North America. Many plant species are visited by honeybees, which are sometimes considered one of the most important pollinators in East Asia . However, the roles of Apis species as pollinators in natural ecosystems are still poorly known. For example, only a few species with specialized honeybeepollinated flowers have been reported, and only among the Orchidaceae and Apocynaceae, which have highly specialized floral traits like pollinaria (e.g., D-K. Tsuji & Kato 2010). Few studies have documented the importance of Trigona species as pollinators in East Asia. Therefore, future studies are needed to elucidate the importance of Apis and Trigona as pollinators in East Asia. At least two approaches are required to determine the importance of eusocial bees as pollinators in natural ecosystems. First, the proportion of plant species in natural communities that depend on pollination by eusocial bees should be evaluated. Second, the effectiveness of pollination by eusocial bees and other pollinators should be compared among many plant species.
Some pollination systems may be less common in East Asia than in other regions. Vertebrate pollination is less common in East Asia than in North America. For example, numerous hummingbird-pollinated plants have been reported from North America (Grant 1994;Fleming & Muchhala 2008). Although sunbirds and pteropodid bats are present in the southern part of East Asia, specialized nectar-feeding birds and nectar-feeding bats are absent in Europe and the northern part of East Asia (J-S. Vincenot et al. 2015;Krauss et al. 2017). In contrast, hummingbirds and nectarfeeding bats are present in North America (Fleming & Muchhala 2008). The lack of specialized vertebrate nectarivores in the northern part of East Asia may explain the limited diversity of the vertebrate-pollinated plants found there, as well as in Europe. However, East Asia has more vertebrate-pollinated plants than Europe. Vertebratepollinated flowers are extremely rare in continental Europe (Ortega-Olivencia et al. 2005). Unlike Europe, the northern part of East Asia has several relatively specialized nectarfeeding vertebrates, including white-eyes, callosciurine squirrels, and civets (Corlett 2004;Kobayashi et al. 2015Kobayashi et al. , 2017Kobayashi et al. , 2019Funamoto & Sugiura 2017). Non-flying mammals and opportunistic nectar-feeding birds may partly fill the pollination niche of specialized nectarivores in the northern part of East Asia.
Nocturnal hawkmoth pollination is observed less in Asia than in Africa and the New World (Johnson et al. 2017), but the nocturnal activity of hawkmoths on flowers can be underestimated (Liu & Sun 2019). Europe also has only a few examples of nocturnal hawkmoth-pollinated plants, such as Platanthera, Silene, and Lonicera (Pettersson 1991;Proctor et al. 1996). However, fewer hawkmoth-pollinated flowers are observed in Europe than in East Asia, likely because Europe has fewer hawkmoth species than East Asia (Ballesteros-Mejia et al. 2017).
There is substantial geographic variation in the pollinator fauna across East Asia. Some regions lack particular groups of pollinators. Oceanic islands, such as the Izu Islands and Bonin Islands, typically lack long-tongued pollinators like bumblebees and butterflies (Inoue 1993;Mizusawa et al. 2014;Hiraiwa & Ushimaru 2017). Although the Ryukyu Islands are continental islands, bumblebees are also absent from most of these islands (Kato 2000;Tadauchi et al. 2014). Hawkmoths and solitary bees may thus act as important pollinators on these islands (Inoue 1993;Kato 2000;Yamada et al. 2010;Mizusawa et al. 2014;Yamada & Maki 2014;Hiraiwa & Ushimaru 2017;Nakajima et al. 2018).

Human impacts on plant-pollinator interactions
The distribution and phenology of organisms is expected to be affected by climate change, which may affect plantpollinator interactions ( Habitat degradation, for example through urbanization, has been proposed to have negative impacts on biodiversity and plant-pollinator interactions (Harrison & Winfree 2015). However, few studies have explicitly examined the effects of habitat degradation on plant-pollinator interactions, and the studies that have been done found mixed results. For example, pollinator visitation and pollinator-related reproductive success were low in urbanized populations of Commelina communis (Ushimaru et al. 2014). In contrast, the visitation frequency of effective pollinators was similar or even higher in an urban population than that in the forest habitats of Island ecosystems are especially vulnerable to both habitat destruction and invasive species (Sugiura 2016). For example, plant-pollinator interactions are threatened on the Bonin Islands, a Japanese oceanic archipelago (Kato et al. 1999;Abe 2006). The pollinator fauna of the Bonin Islands is now dominated by invasive species like Apis mellifera that compete with native pollinators (Kato et al. 1999;Sugiura 2016). A. mellifera may facilitate the reproduction of alien flowers because these bees prefer to visit the flowers of alien species more than those of the islands' endemic species (Abe et al. set of a species endemic to the Bonin Islands, Leptopetalum grayi, was found to be lower on a main island where invasive species like honeybees and anoles were abundant than that on a satellite island where these invasive species were absent (Tsuda et al. 2018). The flower visitors of L. grayi on the main island were dominated by A. mellifera, whereas diverse native visitors were observed on this species' flowers on the satellite islands (Tsuda et al. 2018). On the Bonin Islands, many plants present pollination syndromes toward particular types of pollinators, but virtually no candidate pollinators visit these flowers (Abe 2006;Watanabe et al. 2018). For example, some flowers present a hawkmoth pollination syndrome, but these flowers were observed to receive no or only very rare visits by moths (Abe 2006;Watanabe et al. 2018).
Extensive browsing and grazing by vertebrate herbivores like Sika deer and yak significantly reduce the abundance and diversity of insectary plants (Kato & Okuyama 2004;Xie et al. 2008;Sakata & Yamasaki 2015). In contrast, traditional light and moderate grazing increase the species richness of plants and the nectar production in the community . The decline in flowering plants caused by extensive browsing and grazing can reduce the abundance of bumblebees in the habitat (Kato & Okuyama 2004;Xie et al. 2008;Sakata & Yamasaki 2015), and thus reduce plant reproductive success (Sakata & Yamasaki 2015).

Future work
Research on plant-pollinator interactions in East Asia has been developing rapidly during recent decades. However, the available information on plant-pollinator interactions in East Asia is still limited. Both community-and species-level studies are required to understand the general trends in plantpollinator interactions in East Asia. The accumulation of research carried out in these two directions on plant-pollinator interactions in East Asia would contribute to our understanding of global-level patterns in plant-pollinator interactions.
The available information on plant-pollinator interactions in East Asia is biased towards particular taxa of plants and pollinators. Although pollination studies have been conducted on 105 families of Chinese flora, in most of these families the pollination biology of less than ten species has been examined . Bumblebees and flowers specialized for pollination by bumblebees are frequently studied. However, few studies have examined the role of bumblebees as pollinators of flowers with more generalized pollination systems. Non-bee pollinators are clearly understudied. However, pollinators like butterflies, moths, wasps, flies, birds, and mammals can be important pollinators of many plants. Few studies have documented hawkmoth and bird pollination in East Asia (Johnson et al. 2017;Zanata et al. 2017;Ren et al. 2018). Nocturnal pollinators like bats, moths, and non-flying mammals are also understudied compared with diurnal pollinators. The importance of small pollinators, such as microdipterans, small beetles, and thrips, is also likely underestimated. More studies are required to reveal the importance of these understudied pollinators in East Asia. Experimental approaches could contribute to evaluations of the importance of these understudied pollinators. For example, the relative contribution of different types of pollinators to pollination success in particular plant species could be evaluated by selective exclusion experiments directed at particular pollinators, and also by comparing the single-visit pollination effectiveness among different types of pollinators (e.g., King et al. 2013;Kiepiel & Johnson 2014). Such studies have recently been conducted on many plant species in East J Poll Ecol 25 (6) Asia (e.g., Funamoto & Ohashi 2017;Nakajima et al. 2018;Liu & Sun 2019;Tian & Ren 2019).
Studies carried out in a wide array of ecosystems and regions in East Asia are required because the research on plantpollinator interactions done so far has not been evenly distributed geographically. The majority of pollination studies in China have conducted in mountainous regions . In East Asia, relatively few studies have been conducted in the subtropics, coastal ecosystems, and wetland ecosystems. Although the Korean peninsula and Taiwan are biologically diverse and biogeographically important components of East Asia (Kong & Watts 1999;Choe et al. 2016;Zhu 2016;Tojo et al. 2017), very few studies on plantpollinator interactions have been conducted in these regions.
The relationship between habitat type and the composition of pollinator fauna in East Asia is still unclear because few community-level studies that were comparable among habitat types have been conducted. However, some trends can be observed (Fig. 2). Although very few studies have been conducted in wetland habitats, dipteran pollinators are known to be abundant in wetland habitats (Kato & Miura 1996). Bees, especially bumblebees, and Diptera are dominant flower visitors in alpine regions (Yumoto 1986;Fang & Huang 2012;Mizunaga & Kudo 2017;Ishii et al. 2019). Wasps tend to be more abundant in coastal sand dunes than in other habitats (Inoue & Endo 2006b;Hiraiwa & Ushimaru 2017). More studies are needed to definitively elucidate the relationships between different habitat types and their pollinator fauna.
Comparing East Asian plant-pollinator interactions with plant-pollinator interactions in other regions, especially Europe and North America, can contribute to the improved understanding of the large-scale ecological and evolutionary consequences of plant-pollinator interactions . For example, taxa with disjunctive distributions across North America and East Asia provide opportunities to investigate how plant-pollinator interactions differ across continents (e.g., Fan et al. 2011;Ren et al. 2019).
Rapid economic growth in East Asia is considered to have had negative impacts on the East Asian ecosystem (Fang et al. 2018). The effects of human impacts on plant-pollinator interactions are still poorly investigated in East Asia. Observational and experimental studies are needed to understand these human impacts on plant-pollinator interactions because the effects of humans on plant-pollinator interactions can be case-specific (Ollerton 2017;Kobayashi et al. 2018a). For example, pollinator monitoring is required in East Asia. Although declines in the diversity and abundance of pollinators have been reported in North America and Europe (Ollerton 2017) and declines in Chinese bumblebee populations have been suggested (Williams & Osborne 2009), pollinator monitoring studies have still rarely been conducted in East Asia. In addition, the plant species richness and species richness of flower visitors found on the plants within a community are strongly and positively correlated across communities (Ollerton 2017). Evaluating the relationships among pollinator species/functional diversity, plant species diversity, and plant reproductive success can provide valuable insights for conservation purposes (Hiraiwa & Ushimaru 2017;Ollerton 2017).
In conclusion, this review suggests that there is considerable variation in plant-pollinator interactions across the different regions and habitat types within East Asia. However, the available information on plant-pollinator interactions in East Asia is still limited. Many types of pollinators, habitats, and regions have received little attention. Pollination by nocturnal pollinators is particularly understudied. Community-level studies on many ecosystems across the entirety of East Asia are needed to understand general trends in plant-pollinator interactions in East Asia.

APPENDICES
Additional supporting information may be found in the online version of this article: APPENDIX I. A summary of pollinator groups in East Asia and the plant genera they visit/pollinate.