DISCUSSION
By studying 50 colonies from five species of stingless bees, we showed that the number of brood cells and external activity are the biological parameters that best relate to colony size, regardless of the striking differences in their size and life history traits. To measure bee colony size and relate it to biological parameters closes important knowledge gaps in the life histories of several widely distributed Amazonian stingless bees, but also identifies useful proxy measures of colony size for colony management and multiplication.
The number of brood cells was the best predictor of colony size; however, its measurement involves an intensive handling of colonies, and it is impossible to obtain in natural nests. External activity was also positively correlated with population size, and it is feasible to obtain in managed or natural colonies; we hypothesize this relationship occurs because it is linked with the number of foragers (the last task in the life of workers; Michener 1974; Sakagami 1982; Wille 1983), which are responsible for keeping the food stocks in optimal levels, and thus in capacity to nourish the immatures and adult bees.
External activity is commonly used for measuring the “strength” of the colony; together with the number of adult bees, the number of brood cells and the number of food pots (Hilário et al. 2000; Gostinski et al. 2017). External activity is a metric that can be easily assessed even in natural colonies, where the measurement of brood combs and food stocks is impossible to obtain without destroying the colony (Hilário 2007). The relationship of colony size with the other measured biological parameters varied among the different species. The food stocks only presented a positive relationship with the number of adult bees inM. fasciculata . For the other species, there was no relationship, as shown in the general model, including all species. Regarding the egg-laying rate, in S. aff. postica we observed high levels of variation in the relationship with the adult population, probably due to the high mortality of immature bees observed in this species (Figure S4).
Among the five species studied, S. aff. postica had the largest population, corroborating previous estimates made by Lindauer and Kerr (1960; around 15,000 adult workers), while P. minimapresented the smallest population (mean colony size around 400), much higher than the estimate of 175 bees made by Wille and Michener (1973). For M. fasciculata , our measures presented a smaller population than the estimate of 776 adult bees, which may be due to methodological differences between studies (Kerr et al. 2001). Another possibility is that the since the study of Kerr et al. (2001) was made with other populations of M. fasciculata , in a region approximately 600 km from our study site and with different vegetation physiognomies. Thus, differences in the estimates made for M. fasciculata (and forP. minima ; Wille and Michener 1973) may be due to populational variation and the resources available to colonies in the different areas. For F. longipes and M. flavolineata , there were no previous estimates of their population size.
The size of a colony in social insects is linked with several aspects of their life-history, such as foraging strategies and reproduction rates (Oster and Wilson 1978; Planque et al. 2010). Species with large populations, like we found in S. aff. postica and F. longipes potentially have large numbers of workers involved in defense and resource gathering, and therefore the colony’s consumption of resources may increase proportionally. In contrast, small populations, such as P. minima , although potentially not able to collect large amounts of resources, may have smaller resource demands, thus compensating for the small number of available foragers. However, relationships between consumption/gathering of resources are yet to be studied in stingless bees, with one of the reasons being the perceived difficulty in estimating colony size. As such, our results will help future studies investigate this relationship.
From a practical perspective, knowledge on stingless bee colony sizes is highly relevant for their use as crop pollinators (Giannini et al., 2020). For example, in the Amazon region, Scaptotrigona aff.postica visit crops of economic importance, such as Rambutan (Nephelium lappaceum L.) and pollinate others, such as Açaí palm (Euterpe oleracea Mart) (Ricon-Rabanales 2015; Campbell et al. 2018). However, for crop pollination, clear recommendations on colony stocking densities are necessary for the development of effective managed pollinator programs, as already defined for Apis mellifera (Vaudo et al. 2012). This number depends on the foraging range of the colonies, the number of foragers, and the number of flowers that need to be pollinated within croplands (Kuhn-Neto et al. 2009; Rands and Whitney 2011). Thus, our data can be combined with existing information on species’ foraging ranges to develop novel managed pollinator protocols using native bee species for several important regional crops (Campbell et al. 2019; Araújo et al. 2004).
Finding an efficient method for estimating population size is important for commercial rearing of stingless bee colonies, for monitoring colony health, and for future scientific studies. Our study is the first to count the total number of bees (adult and immature) and correlate it with intrinsic factors in meliponine colonies. Previous studies provided formulae to estimate colony size but did not validate these formulae with biological data (Ihering 1930; Aidar 1996), and a recent study showed that colony size of stingless bees has been overestimated (Valadares et al. 2021). Since the direct measure of the number of adult bees involves invasive sampling methods (Valadares et al. 2021), our study provided a reliable approach to estimate the number of adult bees by using the external activity. On the other hand, our study shows that estimating population size will not be possible with a general formula that works for all Meliponini.
For building a reliable and feasible estimation method for population size, future studies should focus on single species or genus. It is also important to collect larger and broader samples to deal with disturbances caused by natural variation and thus understand the error range of the method. External activity is a good parameter to start developing a formula, because it is a variable that is easy to obtain in natural or managed colonies, and it is highly correlated with population size in most species. Future studies also must focus on other bee genera and verify possible variation in interspecific population size related with different life-history traits, as well as with temporal fluctuations.