Standardization of breeding processes of the native bumblebee colonies Bombus atratus for its implementation as a directed pollination strategy for economically important crops in Colombia

Duration: 2017

Funded by: Colciencias y Vicerrectoria de investigaciones

Coordination: Diego Riaño Jimenez

The UMNG Wild Bees Biodiversity and Ecology research group has developed a breeding system for the native species Bombus atratus, which shares the characteristics of this genus that make them efficient pollinators and presents a wide range of altitudinal dispersion and a high adaptability to disturbed areas that allow us to consider it as a promising species to be implemented in pollination programs of crops at national and South American level.

The development of a colony breeding system under captive conditions has allowed to carry out experimental studies on the activity and pollinator effect of Bombus atratus in different crops of economic interest: tomato. These studies have shown that B. atratus mediated pollination increases the fruit quality in terms of seed numbers and weight by 20-35% compared to fruits produced by self-pollination. In addition, a breeding scheme was developed in semicontrolled conditions – Bombinario, which is a mixed covered (greenhouse) and open area, in which different floral sources are offered with the purpose of supplying the requirements of pollen and nectar needs of B. atratus. The colonies maintained in this scheme can reach sexual maturity (production of queens and males), however is not a continuous process, so it is necessary to establish some criteria for the location and maintenance of colonies in this scheme, among which are to determine the pollinic sources used throughout the life cycle of the colonies, the time that the colonies take for the production of sexuates, the number of productive cycles of sexuates and the number of sexuates per cycle. Although the production of sexuates is not continuous, this system allowed the development of a scheme of breading under controlled conditions, however the efficiency of these in terms of produced colonies is still very low. This aspect has been a constant limitation not only in the breeding system of B. atratus but also in the breeding systems of colonies of other species of bumblebees. Some variables have been defined that can influence the efficiency of the initiation of cross queens such as the maturity age of the queen, the photoperiod and the temperature at which the queens are exposed.

For the analysis of the floral resources used by the colonies of B. atratus, a palynological analysis will be carried out. In order to establish the floral sources used by the colonies during their life cycle, including the phase of production of sexuates, will be carried out a survey of flowering plants located inside and outside the greenhouse, the abundance of such plants will be recorded and the flowering season with which will build a floral calendar of the floral resources available for feeding the colonies for a year. In order to know the pollen of these plants, flower buds will be collected and processed by the acetolysis method (Erdtman 1960) under the protocol of Silva et al. (2010) and will be made permanent sheets with this material which will constitute the reference library. Photographs of the pollen grains will be taken to make a reference catalog for subsequent palynological analyzes. The plates will be stored in the palynological collection of the Entomology laboratory of the Faculty of Basic and Applied Sciences in the Campus Nueva Granada of the UMNG.

In addition, weekly loads of the workers’ corbiculas and the pollen and nectar in the nests will be collected from the time of their transfer to field conditions until the end of the production of sexuates. Observations of pollen and nectar accumulation within the nests and entry frequency observations of workers with pollen loads for 5 minutes in each colony will be recorded. The collected pollen loads will be processed under the same methodology of acid acétolisis (Erdtman 1960; Silva et al., 2010) and will be analyzed with the help of the reference catalog using the comparison method (Silva et al., 2010).

A quantitative analysis will be performed by counting 400 grains in each sheet of worker loads to determine the most frequent pollen type. In addition, the pollen types will be classified according to the percentage and classes of occurrence according to the classification proposed by Barth (1970) and Louveaux et al. (PAI between 3 and 14%), occasional isolated pollen (PAO between 1 and 5%), pollen (PA, between 15% and 45%), pollen 3%) and pollen present (PP, 1%).


Melissa Yvette Guerrero Torres

Paola Alarcon

Natali Plazas Usategui

Laura Rojas


Universidad Militar Nueva Granada Colciencias