Optimization of pollination of greenhouse-grown tomatoes by bumblebees

蜜蜂作为大多数水果和蔬菜作物的授粉者发挥着关键作用。温室种植的番茄需要主动的花粉转移，无论是通过机械手段还是通过觅食蜜蜂，如驯化的大黄蜂，熊蜂。在温室里，由熊蜂授粉的番茄比机械授粉的番茄果实更大，每公顷产量更高。然而，使用大黄蜂也有一些缺点。种植者只能使用英国本土的浅黄色尾熊蜂亚种B。terrestris audax，以防止疾病传播给野生熊蜂。然而，这些菌落通常在放置后2-4周内下降，并且在软质水果和蔬菜作物（包括有商业价值的番茄品种）上表现出相对较差的授粉率。其业绩不佳的原因尚不清楚。目前，种植者将殖民地放在温室里，没有对殖民地进行任何直接监测。除了计算来访的蜜蜂在花朵上留下的咬痕外，他们也没有办法将蜜蜂的行为与实际的授粉率和番茄产量联系起来。在这里，我们建议与英国最大的温室番茄种植者Thanet Earth（肯特）和基于传感器的蜜蜂技术公司Bee-Ops合作，开发监测和改善英国种植番茄上B.terrestris audax管理的方法。我们的项目有四个主要目标：（1）测定熊蜂群体在温室番茄作物上的授粉效率（2）探讨蜂群发育阶段和营养状态对熊蜂群体寿命和觅食率的影响;（3）利用B提高番茄授粉效果。（4）使用目标1- 3的数据创建一种改进的菌落管理策略，优化授粉率。在目标1中，通过与Bee-Ops合作，我们将采用最新的人工智能和传感器技术来监测Thanet Earth商业温室中B.terrestris audax菌落的活动。传感器技术将使我们能够为种植者提供有关当前授粉状态的重要数据：从首选供应商那里收到的殖民地质量评级，以及有关特定品种的花卉访问和殖民地觅食率及其对特定条件下产量的影响的高分辨率信息（例如一年中的时间，温度）。在目标2中，我们将全面测试引入温室的殖民地的发育阶段如何影响觅食率。使用Wright和史蒂文森实验室开发的熊蜂优化花粉替代品，我们的目标是通过研究在殖民地发育早期营养干预如何影响殖民地的人口规模和觅食率来提高殖民地寿命。在目标3中，我们将测量番茄花香散发的挥发性化合物。利用史蒂文森实验室开发的方法，我们将通过将蜂群中的蜜蜂暴露在与含咖啡因花蜜相关的番茄花香中，来训练蜜蜂学习对番茄花香的定向。这些数据将构成目标4中设计的策略的基础，以显示特定的干预措施如何提高熊蜂对特定番茄品种的授粉效率。与我们的项目合作伙伴一起，我们将制定新的策略，以实现最佳的番茄授粉。我们预测，我们通过这一建议开发的方法组合将推广到大黄蜂授粉的其他水果和蔬菜作物生长在温室或polytunnels。

Bees play a critical role as pollinators of most fruit and vegetable crops. Greenhouse grown tomatoes require active pollen transfer, either through mechanical means or by foraging bees such as the domesticated bumblebee, Bombus terrestris. In greenhouses, tomatoes pollinated by bumblebees have larger fruit and greater yield per hectare than those pollinated mechanically. However, using bumblebees has some drawbacks. Growers are limited to use the UK native subspecies of the buff-tailed bumblebee, B. terrestris audax, to prevent the spread of disease to wild bumblebees. However, these colonies generally decline within 2-4 weeks of placement and exhibit relatively poor pollination rates on soft fruit and vegetable crops including commercially-valuable tomato varieties. The reasons for their poor performance are unknown. At the present, growers place colonies in glasshouses without any direct monitoring of the colonies. With the exception of counting bitemarks on flowers left by visiting bees, they also have no way to relate the behaviour of the bees to actual pollination rate and yield of tomatoes. Here, we propose to develop methods to monitor and improve the management of B.terrestris audax on UK grown tomatoes in collaboration with the UK's biggest glasshouse tomato grower, Thanet Earth (Kent) and sensor-based bee technology company, Bee-Ops.Our project has four main objectives: (1) To measure the pollination efficiency of bumblebee colonies in tomato crops in greenhouses (2) To identify how colony phase and nutritional state influence colony longevity and foraging rate of bumblebee colonies; (3) To improve the pollination of tomatoes by B. terrestris audax through in-colony exposure to floral chemical cues; (4) To create a strategy for improved colony management that optimizes pollination rate using data from Objs 1-3.In Obj 1, through collaboration with Bee-Ops, we will employ the latest AI and sensor technology to monitor the activity of B.terrestris audax colonies in commercial glasshouses at Thanet Earth. Sensor technology will enable us to provide the growers important data about the current state of pollination: a rating of the quality of the colonies received from their preferred suppliers and high resolution information about floral visitation and foraging rate of colonies for specific varieties and its influence on yields under particular conditions (e.g. time of year, temperature). In Obj 2, we will comprehensively test how the developmental phase of the colony introduced to the glasshouse influences foraging rate. Using an optimized pollen substitute for bumblebees developed by Wright and Stevenson's laboratories, we aim to improve colony longevity by studying how nutritional intervention early in colony development affects the colony's population size and foraging rate. In Obj 3, we will measure the volatile compounds emitted by tomato floral odours. Using methods developed in Stevenson's laboratory, we will train bees to learn to orient towards tomato floral scents by exposing bees in colonies to tomato floral scent in association with caffeinated nectar. These data will form the basis of a strategy devised in Obj 4 to show how specific interventions improve the pollination efficiency of bumblebees on specific tomato varietals. Together with our project partners, we will devise new strategies for optimal tomato pollination. We predict that the combination of methods we develop through this proposal will generalize to bumblebee pollination of other fruit and vegetable crops grown in greenhouses or polytunnels.