Impacts of pests and diseases on the decline of managed bees in Brazil: a beekeeper perspective

Abstract Brazilian beekeepers have experienced massive colony losses in recent years. Possible causes include bee pests and diseases, although there have been few relevant studies in the country and no official data are available. The aim of this work was to assess, from a beekeeper's perspective, the incidence of pests and diseases in colonies, and their impact on the decline of bee populations in Brazil. The research used two online questionnaires that included disease symptoms to facilitate participant identification of diseases. A total of 674 valid responses were obtained from beekeepers of Apis mellifera (49% of the total) and stingless bees (51%), distributed among 24 of the 27 Brazilian states. Most beekeepers observed pests (average 78.1%) and disease symptoms (65.5%) in their colonies. Based on beekeepers' responses, pests and diseases affect beehive losses in Brazil. Colony loss percentage for A. mellifera was higher in apiaries with ants (19.8% of colony losses) and wasps (46.7%) than in those without pests (p < 0.001). While for stingless bees, meliponaries with robber bees (15.8% of colony losses), ants (14.7%), and phorid flies (12.5%) had higher colony losses than those without pests (p < 0.001). Apiaries and meliponaries with worker bee deaths had higher losses than those without symptoms. The results also indicate that monitoring can reduce colony losses in apiaries and meliponaries that have some kind of pests and diseases. Using a beekeeper perspective, this study provides information about the impact that pests and pathogens have on bee losses that can be used to guide improved management practices.


Introduction
Bees are essential for maintaining biodiversity in wild and agricultural ecosystems as they are the main pollinators (Ashworth et al., 2009;Giannini et al., 2020;Potts et al., 2016).However, their presence in these ecosystems has decreased, which has aroused great concern worldwide for their conservation (Gregorc, 2020;Potts et al., 2016).A reduction of 3 to 8% in world agricultural productivity would be among the main consequences of the disappearance of bees (Bretagnolle & Gaba, 2015) and countries that depend economically on food exports, such as Brazil, are more vulnerable to pollinator crisis (Novais et al., 2018).It is estimated that the loss of pollinators in Brazil would reduce its agricultural production by up to US$14.56 billion annually (Novais et al., 2018).Between 2013 and2017, 19,296 of 37,453 inventoried bee colonies and nests in Brazil were lost (Castilhos et al., 2019).Mortality of these essential pollinators in agriculture in Brazil has been mainly attributed to pesticides (Castilhos et al., 2019).However, there is consensus that the interaction of multiple factors (such as parasites and pathogens, lack of floral resources, habitat loss and climatic change) has contributed to this bee decline (Maggi et al., 2016;Requier, 2019;vanEngelsdorp et al., 2009).
Besides pesticides, bee losses can also be affected by pathogens and parasites (Pires et al., 2016).Maintenance of bee colony health is extremely important in Brazil because beekeeping has a major impact on the export market of bee products (Oliveira et al., 2021).There is a lack of official records, and knowledge about bee health in Brazil is currently restricted to studies conducted in the South and Southeast regions (Message et al., 2012;Pires et al., 2016).Most studies have focused on A. mellifera (Pires et al., 2016), which has the greatest commercial exploitation (Paula et al., 2016).These studies have assisted A. mellifera beekeepers (Ap-beekeepers) in the care of their colonies.Despite the increasing number of stingless bee beekeepers (Sb-beekeepers) in recent years in Brazil, there is still little epidemiological knowledge concerning these bees (Jaff e et al., 2015).It is possible that many of the parasites and pathogens observed in A. mellifera also affect stingless bees.Apis mellifera interacts with stingless bee species when they share floral resources, thus causing pathogen spillover between species (Ngor et al., 2020;Purkiss & Lach, 2019).There are increasing reports in the literature of the occurrence of parasites and diseases of A. mellifera in stingless bee colonies (e.g., Porrini et al., 2017;Teixeira et al., 2020;Toufailia et al., 2017).Moreover, some management practices, such as nutritional supplementation with honey bee products, can facilitate the dissemination of diseases in both bee groups (Gisder & Genersch, 2017;Mallinger et al., 2017).Disease and pest resistance depend on genetic predisposition and hygienic behavioral characteristics, which differ significantly between these bee groups (Toufailia et al., 2016).Also, stingless bees have a great diversity of genera and species (Michener, 2000).Therefore, there is a need to better understand the composition of natural enemies that occur among managed bees to develop assertive control strategies and prevent the emergence of new diseases.
Despite pests and pathogens being two of the main factors associated with the disappearance of bees worldwide (e.g.Ant unez et al., 2017;Gregorc, 2020;Porrini et al., 2017;Sharif et al., 2020;Teixeira et al., 2020), there are still few studies and there is no active monitoring in Brazil, except for some localities (e.g., Requier & Ant unez, 2019).Such studies are necessary for a better understanding of the distribution of pests and pathogens in the country and how they are shared by different bee species.To obtain information about bee health in an extensive territory such as Brazil, a citizen science approach involving the perspective of beekeepers can be a useful strategy (See Appenfeller et al., 2020).The inclusion of local knowledge of beekeepers is common in scientific investigations of bees (e.g., Ant unez et al., 2017;Brodschneider et al., 2016), and in many cases, participatory research projects have produced legitimate scientific knowledge quickly and cheaply.
Another influential factor for bee health is the surrounding landscape, as it defines the food that bees obtain (Ricigliano et al., 2019;Steinhauer et al., 2018) as well as other important resources for bee health, such as resins (Requier & Leonhardt, 2020).Poor nutrition reduces the physiological defenses of bees against pathogenic infections and pest invasions (Di Pasquale et al., 2016;Dolezal & Toth, 2018).Low quantity and diversity of floral resources for bees have been associated with altered environments, such as areas converted for pasture, silviculture and agriculture (Smart et al., 2016).Colonies near forest remnants can provide better conditions for bees and ensure their survival (McNeil et al., 2020;Ricigliano et al., 2019).In addition, management strategies for beekeeping, such as artificial complementary feeding or even planned gardens to ensure colony health during periods of food shortage are decisive in maintaining bee health (Steinhauer et al., 2021).Knowledge acquired through personal and interpersonal experience, courses and training in bee management can also be valuable for ensuring colony health (Jacques et al., 2017;Jaff e et al., 2015).Knowledge about some beekeeping practices, such as monitoring, queen replacement and renewal of wax and food stocks in colonies, can also help reduce the incidence of parasites and pathogens (Jacques et al., 2017).
Our study assessed the bee pests and pathogens in colonies of A. mellifera and stingless bees and their impacts on bee decline in Brazil, from a beekeeper perspective provided through questionnaires.Our hypotheses proved from the beekeeper's perspective were: (i) bee pests and pathogens are similar between A. mellifera and stingless bees; (ii) bee pests and pathogens increase losses of A. mellifera and stingless bees; (iii) colonies near environments altered by human activities (pasture, silviculture, agriculture and urbanization) have higher incidences of bee pests and pathogens than colonies near natural vegetation; and (iv) beekeeper qualifications and good management practices reduce the incidence of bee pests and pathogens in colonies.The information obtained from these questionnaires allowed us to better understand the status of bee health in the country and the challenges that beekeepers face in relation to bee health based on their knowledge, as well as the role pests and diseases play in the loss of colonies and possible associated causes.

Questionnaire
Two semi-structured questionnaires were prepared in the Portuguese language: one aimed at Ap-beekeepers (Supplementary Questionnaire 1) and one for Sb-beekeepers (Supplementary Questionnaire 2).The questionnaires were developed to assess the occurrence of bee pests and pathogens, their respective impacts on colony losses and the possible causes from the perspective of beekeepers.The questionnaires were based on previous studies, such as the those developed by the COLOSS association (van der Zee et al., 2013), the Solatina Society (Ant unez et al., 2017;Requier & Ant unez, 2019), and two studies developed in Brazil (Castilhos et al., 2019;Jaff e et al., 2015).No questionnaires addressing the specific theme of pests and diseases aimed at beekeepers were found in the literature.
We listed the main insect groups that are related to colony loss in Brazil and classified them as parasites, cleptoparasites, cleptobiotics, predators, and others (Supplementary Table 1).Questions about the main diseases were presented to beekeepers in the form of symptoms to make questions more liable to interpretation by someone who is not knowledgeable about diseases but is able to recognize symptoms (Supplementary Table 2).The questionnaires were reviewed by national and international researchers with extensive knowledge about bees and by leaders of Brazilian beekeeping associations.In addition, pre-tests of the questionnaires were applied to a group of 10 beekeepers in the Brazilian state of Minas Gerais in 2019 to eliminate, add or reformulate the final questions.
Each questionnaire contained 30 questions.Due to the different management characteristics of Ap-beekeepers and Sb-beekeepers, specific questions were developed for each group, with adjustments to alternatives for multiple-choice questions.The questions were divided into four sections: (1) Socioeconomic profile; (2) Evaluation of colony loss; (3) Characterization of the surrounding area; and (4) Management practices.The answers were collected in reference to the period from January 2017 to December 2019 for each apiary/meliponary.
Beekeepers were invited to participate by an email distributed to a list created from contacts from associations, federations, cooperatives, research institutes, NGOs and beekeeping stores.The list comprised 503 email addresses covering all Brazilian states.The email invitation included two access links to the virtual questionnaires hosted on the web page of the Federal University of Minas Gerais, a well-known Brazilian public teaching and research institution.One of the links was addressed to Apbeekeepers, while the other to Sb-beekeepers.The questionnaires were also sent to beekeeping groups via messaging applications.Participants were invited to publicize the project at the end of the questionnaire ("snowball sampling" Goodman, 1961).Questionnaires were answered between December 2019 and December 2020 throughout Brazil.All participants gave their informed consent in writing prior to inclusion in this research.The information, even from those allowing identification, was kept confidential.The research did not involve indigenous populations and did not reveal any personal information of its participants, thus, based on Jaff e et al. (2015) and some waiver criteria for submission to an ethics committee, no application for approval was made.
A total of 677 responses were obtained from Apbeekeepers and 765 from Sb-beekeepers.The answers were submitted to a validation process (details in Supplementary Figure 1), which eliminated 51% of the responses from Ap-beekeepers and 55% from Sb-beekeepers, resulting in 330 and 344 valid questionnaires, respectively.The total of 674 valid questionnaires was greater than the sample size of 246, which was considered the minimum sample size (Castilhos et al., 2019) needed to guarantee the accuracy of analyses with a 95% confidence level and 5% estimated error for an estimated national beekeeper population of 350,000 (Maggi et al., 2016), according to van der Zee et al. (2013).The spatial distribution of responses was obtained from the geographical coordinates of the centroid of the informed region (municipality/neighborhood/district) on a cartographic basis of Brazilian territorial limits at a scale of 1:250,000 (IBGE, 2010) using R software (R Core Team, 2020;Tennekes, 2018).

Statistical analysis
Calculations of losses were made according to the average percentage of losses experienced by each participant so that enterprise size (number of colonies) had less influence on loss rates (Castilhos et al., 2019;van der Zee et al., 2013).Individual questions of the questionnaires were optional, which resulted in different numbers of answers for each question.
The Mann-Whitney U test (Mann & Whitney, 1947) was used to compare parameters that considered differences between two options (e.g., with pests and without pests, with symptoms and without symptoms).The Kruskal-Wallis non-parametric test (Kruskal & Wallis, 1952) was used to compare parameters that considered differences among more than two treatments (e.g., types of pests and symptoms, places near colonies and management techniques), with post-hoc analysis using Dunn's Pairwise Multiple Comparison Test (Dunn, 1961) to identify differences between treatments.Comparisons between two or more categorical variables were performed by the Chi-Square Independence Test.
The effects of interactions between the variable of experience and those of management techniques, as well as the effects of interactions between the variables of colony health conditions (pests presence/ absence, disease symptoms presence/absence) and those of management techniques, on the response variable of colony loss, were evaluated.The effects of pair-wise interactions between beekeeper experience/qualification and variables of management techniques on the response variables of colony health conditions (number of types of pests and symptoms) were also investigated.These interactions were analyzed by two-way interaction analyses using General Linear Models (GLMs) followed by a posthoc analysis by the Mann-Whitney U test when comparing categories of variables that had only two options and the non-parametric Kruskal-Wallis test for comparing the categories of variables with more than two options.All tests considered a 95% confidence level (a ¼ 0.05) and were performed in R (R Core Team, 2020).The results expressed in Figures or Tables are presented as means and standard error (±SE).

General data
A total of 674 valid questionnaires were obtained, including 330 from Ap-beekeepers and 344 from Sbbeekeepers (Supplementary Table 3).Responses were from 24 of the 27 Brazilian states and were concentrated in the Southeast (50.5% of all responses), South (23.1%) and Northeast (18.4%) regions of the country (Figure 1).Most of the respondents were male (87.5% of Ap-beekeepers, 88.2% of Sb-beekeepers).The average age was 47 ± 0.7 years old for Ap-beekeepers (47.3 ± 2.8 years for women, 47.2 ± 0.79 years for men) and 43 ± 0.7 years old for Sb-beekeepers (44.1 ± 2.14 years for women, 43.4 ± 0.78 years for men).The average experience for Ap-beekeepers was 13.3 ± 0.7 years and for Sb-beekeepers 5.9 ± 0.5 years.
The total number of colonies declared by the beekeepers between 2017 and 2019 was 85,064, including 76,886 and 8,178 colonies of Apis mellifera and stingless bees, respectively.Most of the participants (66.1% of Ap-beekeepers, 90.1% of Sb-beekeepers) were small-scale beekeepers (less than 50 colonies) and exercise the activity as a complement to income (81.3% of Ap-beekeepers, 99.7% of Sb-beekeepers).About one quarter (24%) of Sb-beekeepers answered that they perform the activity for recreation or interest in preserving nature without generating any financial income.
Most Sb-beekeepers (87.7% of responses) declared the common name of bees, and only 12.3% declared the scientific name of the bee species in their answers.In all, 93 common names and 36 scientific names were declared in the questionnaires.The most common species was Tetragonisca angustula (popular name jata ı; 20.4% of all Sb-beekeeper responses) (Supplementary Table 4).
The results also show that 90.9% of the Ap-beekeepers (v2 ¼ 65.403; p < 0.001) and 85.8% of the Sb-beekeepers (v2 ¼ 26.354; p < 0.001) who mentioned having observed pests in their colonies also noted disease symptoms.

Association between colony loss and pest and disease symptom presence
During the three years analyzed (2017 to 2019), the average individual colony loss was 15.3 ± 0.99% for Ap-beekeepers and 10.3 ± 0.78% for Sb-beekeepers.
The influence of land use and occupation around apiaries on the occurrence of bee pests and disease symptoms Land use and occupation around apiaries did not influence the number of types of observed pests (v 2 ¼1.389; n ¼ 249; p ¼ 0.751) or the symptoms (v 2 ¼3.368; n ¼ 234; p ¼ 0.498) for A. mellifera (Figure 3).However, land use and occupation did influence the type number of pests (v 2 ¼21,573; n ¼ 269; p < 0.001) and symptoms (v 2 ¼12.44; n ¼ 254; p < 0.001) for stingless bees.Stingless bees had the lowest number of types of pests (1.26 ± 0.09) and symptoms (0.87 ± 0.09) in urban areas.The average number of types of pests and symptoms did not differ among the other environments: remaining areas of natural vegetation (pests ¼ 1.75 ± 0.11; symptoms ¼ 0.90 ± 0.10), silviculture (pests ¼ 2.54 ± 0.31; symptoms ¼ 1.7 ± 0.45) and pasture (pests ¼ 1.9 ± 0.34; symptoms¼ 1.68 ± 0.29).Average number of types of pests or disease symptoms of A. mellifera followed by the same small-case letter do not differ significantly (a ¼ 0.05).Average number of types of pests or disease symptoms of stingless bees followed by the same capital letter do not differ significantly a ¼ 0.05.

The influence of experience, qualification and management techniques on the occurrence of bee pests and disease symptoms
The level of experience of Ap-beekeepers did not influence the number of types of pests (p ¼ 0.123) or diseases (p ¼ 0.451) (Supplementary Table 6).However, experienced Sb-beekeepers (more than three years of experience) observed more pests in their meliponaries (1.77 ± 0.123) than those with less experience (1.29 ± 0.09) (p ¼ 0.002; Supplementary Table 6).The enterprise size did not affect the observations of pests and diseases for both groups of bees (Supplementary Table 6) Ap-Beekeepers who learned their trade formally (e.g.courses, academic qualification and cooperatives) observed fewer disease symptoms (average number of types of symptoms ± standard error ¼ 1.05 ± 0.09, p ¼ 0.02) than those who had an informal education about bees (1.47 ± 0.16) being Self-taught, Friends/Family (Supplementary Table 6).However, the form of professional qualification did not influence the number of types of pests (v 2 ¼0.246; n ¼ 377; p ¼ 0.969) or disease symptoms (v 2 ¼0.365; n ¼ 349; p ¼ 0.365) for stingless bees.
All management techniques (i.e., timing of monitoring colonies, ways to acquire colonies, types of complementary feeding, planting of bee-friendly gardens, movement of colonies to new locations, queen replacement, old wax replacement) that Ap-beekeepers indicated as helping the colony quality and productivity had no effect on reducing the number of types of pests and disease symptoms (p > 0.05; Supplementary Table 6).Sb-beekeepers who did not plant a bee-friendly garden (1.53 ± 0.08) had more pests in their meliponaries than those who did (1.05 ± 0.14; p < 0.001).Meliponaries that were not transported to another location (1.28 ± 0.09) had fewer pests than those that were moved to another location (1.60 ± 0.12; p ¼ 0.022).Meliponaries without any kind of complementary feeding (0.83 ± 0.18) had fewer pests than those with complementary feeding of protein (1.64 ± 0.19), honey (1.75 ± 0.11), sugar syrup (1.64 ± 0.09), and pollen (1.80 ± 0.14; p < 0.001) (Supplementary Figure 3).Meliponaries with complementary feeding based on protein had the highest increased incidence of predators (v2 ¼ 10.754; p ¼ 0.029).Parasites, such as mites, phorids, wax moths and beetles, were frequently observed in meliponaries that received any type of complementary feeding (v 2 ¼29.014; p < 0.001), while other pests (bees and ants) were more frequently observed in meliponaries with complementary feeding of honey and pollen (v 2 ¼1,404; p ¼ 0.02).

The influence of beekeeper knowledge (experience and qualification) and management techniques on the quantities of pests and disease symptoms
Ap-beekeeper experience and complementary feeding type provided had no interaction effect on the number of types of pests or on the number of disease symptoms in apiaries (Supplementary Figure 4A and B; Supplementary Table 7).However, for Sb-beekeepers, inexperienced respondents who provided a mixed diet observed more pests (X 2 ¼10.772; p ¼ 0.013) and more disease symptoms (X 2 ¼10.849; p ¼ 0.013) than did those who did not provide such diets (Supplementary Figure 4C and D; Supplementary Table 8).AP-beekeeper qualification and complementary feeding types provided had no interaction effect on the number of types of pests or on the number of disease symptoms in apiaries (Supplementary Figure 5A and B; Supplementary Table 7).However, Sb-beekeepers with informal training who did not perform complementary feeding observed more pests (X 2 ¼9.005; p ¼ 0.029) and more disease symptoms (X 2 ¼8.424; p ¼ 0.038) than did those who provided protein and mixed diets (Supplementary Figure 5C and D).Sb-beekeepers with formal training who provided mixed diets observed more pests (X 2 ¼11.506; p ¼ 0.009) than did those who provided protein or no complementary feeding.
The interaction between training type and monitoring frequency affected the number of types of pests found in apiaries (X 2 ¼7.022; p ¼ 0.030; Supplementary Figure 6A).Among Ap-beekeepers who monitored at intervals longer than two months, those who had formal training observed fewer pests than those who were informally trained.Among Apbeekeepers who monitored monthly, those with formal training observed fewer disease symptoms than those with informal training (W ¼ 507.5; p ¼ 0.036; Supplementary Figure 6B).For Sb-beekeepers, there was no effect of an interaction between training and monitoring frequency on the number of types of pests or the number of disease symptoms (Supplementary Figure 6C and D; Supplementary Table 8).
The influence of pests and disease symptoms, and management techniques on the colony loss Among Ap-beekeepers who checked apiaries weekly, those observed pests had similar losses to those who did not (W ¼ 277.5; p ¼ 0.161).When monitoring was less frequent, Ap-beekeepers reporting pests had higher losses than did those who did not (Figure 4A).Among Sb-Beekeepers reporting pests, those who checked meliponaries weekly had fewer losses than those who checked less frequently (X 2 ¼7.2; p ¼ 0.027; Figure 4B).
Among Ap-beekeepers who monitored weekly, those who observed the presence of disease symptoms had greater losses than those who did not (W ¼ 333.5; p < 0.001), which was also true for those who monitored monthly (W ¼ 333; p < 0.001; Figure 4C).However, when monitoring was done at intervals longer than two months, losses for Ap-beekeepers reporting symptoms were similar to those who did not (W ¼ 18.5; p ¼ 0.48; Supplementary Table 9).In contrast, among Sb-Beekeepers reporting disease symptoms, those who checked weekly had fewer losses than those who checked less frequently (X 2 ¼7.000, p ¼ 0.030; Figure 4D; Supplementary Table 10).
Among Ap-beekeepers who provided their colonies with an energetic diet (e.g., sugar), those who did not observe pests had fewer losses than those who did (W ¼ 135.5, p ¼ 0.004; Figure 4E).Sb-Beekeepers had greater colony loss when complementary feeding was not provided (W ¼ 50, p ¼ 0.023) or when protein was provided (Figure 4F).
Among Ap-beekeepers who provided an energetic diet, those reporting disease symptoms had greater losses than those who did not (W ¼ 78.5; p < 0.001) which was also true for those who provided a mixed diet (W ¼ 239; p < 0.001; Figure 4G).For Sb-beekeepers, healthy colonies resulted in fewer losses when complementary feeding was not given (W ¼ 24.5; p ¼ 0.032) or when protein was fed (W ¼ 23; p ¼ 0.05; Figure 4H; Supplementary Table 10).

Discussion
This research, guided by a citizen science perspective (participation of the general public in the scientific process), was the first survey of the impacts that bee pests and pathogens have on the decline of managed bees in Brazil based on experiences shared by beekeepers.The contributions of beekeepers came from nearly all Brazilian states and included an unprecedented contribution from the Northeast, as most bee health studies have concentrated on the South and Southeast regions of the country (Pires et al., 2016).This study includes similar contributions by Ap-beekeepers and Sb-beekeepers for the assessment of the health status of Brazilian apiaries and melinoparies.This study indicates that both groups experience significant losses of colonies, with rates similar to those observed in temperate zone countries where colony loss rates of 5.8 to 32% were reported from 2018 to 2019 for A. mellifera (Gray et al., 2020).On the other hand, colony loss rates of 30% and 50% per year were reported for Argentina and Chile, respectively (Maggi et al., 2016), which is higher than observed here.The use of Africanized A. mellifera appears to be a crucial factor that positively affects the survival of the species.Beekeepers and researchers have found these bees to be more resistant to pathogens and parasites than European A. mellifera (De Jong, 1996).For example, A. mellifera in Venezuela, Brazil and some provinces in Uruguay and Argentina are able to coexist with the mite, Varroa destructor.Apparently, the most problematic countries in terms of bee health are Argentina and Chile, due to the fact that most of their bee populations belong to European lineages, which are highly susceptible to attacks by parasites and pathogens (Maggi et al., 2016).Colonies heavily infested with mites are sometimes treated with acaricides, which can further reduce the bee immune system (Boncristiani et al., 2012).In addition, treatment options tend to become ineffective as the mites become more resistant (Roth et al., 2021).To avoid excessive use of acaricides, the selection of miteresistant bee lines may be an alternative for the control of this important parasite (Traynor et al., 2020).
Our study provides evidence, from the perspective of beekeepers, that pests and pathogens contribute to the loss of colonies in Brazil.Moreover, colonies that had pests also had, simultaneously, disease symptoms.In fact, some pests, such as mites, may favor the entry of viral infections, such as deformed wing virus (DWV), acute paralysis virus (ABPV), and weakened colonies (Levin et al., 2016).Phorids are the main parasites of stingless bees and are known to transmit viruses and other diseases (Core et al., 2012;Ueira-Vieira et al., 2015).Ants were the insects most associated with higher colony loss in apiaries.Some of these diseases, depending on the extent of damage, can weaken colonies and, consequently, make them more susceptible to parasite attack (Barroso-Ar evalo et al., 2019).
Abrupt reduction in the number of workers, the main symptom of colony collapse disorder (vanEngelsdorp et al., 2009), was identified in both groups of bees, but this symptom only resulted in significant colony loss for A. mellifera.Some studies have indicated that a combination of pesticides and pathogen infection is associated with large losses of colonies, referred to as colony collapse disorder (Doublet et al., 2015;S anchez-Bayo et al., 2016).Daily sublethal doses of pesticides cause oxidative stress and reduce the ability of insects to produce protective peptides, which prevent the proliferation of microorganisms (James & Xu, 2012), thus reducing bee longevity and resistance of bees to diseases (S anchez-Bayo et al., 2016).Pesticides can also reduce cognitive capacity, flight capacity, foraging activity, and can be lethal (S anchez-Bayo et al., 2016).Therefore, and in the context of new public policies currently adopted by the Brazilian Agriculture Ministry that can lead to an increase in the use of pesticides (Barbosa et al., 2021;Braga et al., 2020), it is imperative to monitor this problem to ensure the health of colonies.
Nevertheless, a reduced number of workers may also be related to worker death, as observed in both Apis mellifera and stingless bee colonies.Some studies have indicated that the presence of ABPV klepto virus and chronic bee paralysis virus, observed by the beekeepers in both groups of bees studied here, are responsible for worker deaths (Ueira-Vieira et al., 2015), although the symptoms of these diseases can be easily mistaken for poisoning when laboratory analyses are not performed (Kiljanek et al., 2016).Infections with these viruses are not always apparent, but they can spread rapidly and cause significant colony loss (Chen & Siede, 2007).Periodic evaluations of colonies are essential to control and minimize the incidences of these viral diseases (Chen & Siede, 2007).We also observed that frequent monitoring (weekly and monthly) of disease or pest infested colonies can help avoid losses, particularly in stingless bees.However, colony opening for monitoring may result in increased exposure to robber bees and other enemies (i.e.phorid fly and the robber bee Lestrimelitta limao) and lead to an increase in the use of resources needed to seal the colonies (Villas-Bôas, 2012) The vegetation surrounding colonies' locations is important as it determines the floral resources available to bees; natural vegetation normally provides a wide array of floral and non-floral resources contributing to better nutrition and increased resistance to pests and diseases (Di Pasquale et al., 2016;Dolezal & Toth, 2018;Kaluza et al., 2018;Requier & Leonhardt, 2020;Ricigliano et al., 2019;Steinhauer et al., 2018).Nevertheless, no significant differences were found in the incidence of pests and disease symptoms as a function of vegetation, silviculture, agriculture, and pasture surrounding colonies.Bee communities are strongly conditioned by local effects and may exhibit patterns of spatial heterogeneity at a scale as low as 500-1000 m in patches of homogeneous habitat (Torn e-Noguera et al., 2014).These results have important implications for local pollination dynamics and spatial variation of plant-pollinator networks.Kremen et al. (2002) and Kim et al. (2006) found that bees are negatively influenced by agricultural activity located near areas of natural vegetation.In addition, a major current problem facing beekeeping is the use of pesticides near colonies, which can reduce bee immunity, thus increasing susceptibility to pests and pathogens (Doublet et al., 2015;James & Xu, 2012;S anchez-Bayo et al., 2016).Our findings showed a lower incidence of pests and symptoms in urban environments for stingless bees, which is in accordance with a trend that indicates that social bees adapt well to urban environments when they are correctly managed (Banaszak-Cibicka & _ Zmihorski, 2012; Theodorou et al., 2020).The correct use of complementary feeding or the creation of suburban gardens makes it possible to raise stingless bees in cities.In fact, 47.2% of the beekeepers who responded to the survey raised bee colonies in cities. Kaluza et al. (2016) and Leonhardt et al. (2016) found results similar to those of our study, indicating greater foraging activity of the Australian stingless bee Tetragonula carbonaria in suburban gardens than in farmland.Gardens in urban areas may offer a greater supply of food resources for bees (Leonhardt et al., 2016;Mouga et al., 2015).
Colonies are often affected by a lack of floral resources, which can be mitigated if the beekeeper provides complementary food (Kulhanek et al., 2021).However, we observed that Sb-beekeepers who have informal training and provide complementary feeding informed more problems with bee diseases than those who do not provide such diets.In addition, a lack of knowledge incites some Sb-beekeepers to feed their colonies with A. mellifera honey and pollen (Teixeira et al., 2020).This complementary food may contain bee pathogens (e.g., Paenibacillus larvae, Ascosphaera apis, Nosema ceranae and Nosema apis) that cause diseases, including American foulbrood chalkbrood disease (brood diseases), and adult bee disease, nosemosis (Teixeira et al., 2018).Complementary feeding without adequate sterilization can increase the incidence of diseases, which would explain the problems observed.The absence of a relationship between complementary feeding and these symptoms for A. mellifera may be related to the inability of beekeepers to properly identify disease problems.Some health problems may be overcome by A. mellifera hygienic behavior (Guimarães-Cestaro et al., 2020;Palmer & Oldroyd, 2003;Spivak & Danka, 2021).Nosemosis, which was identified as diarrhea in the questionnaire, was rarely observed for both groups of bees.It is possible that this disease may be less impactful in tropical countries (Guimarães-Cestaro et al., 2020).On the other hand, further investigations are needed to establish a causal link between complementary feeding and the incidence of pests and pathogens.
Beekeeper experience and qualification apparently help reduce the frequency of diseases and pests.Beekeeper education affects their ability to identify symptoms of bee pests and diseases (Jacques et al., 2017), as well as timely sample collection and knowledge of control measures.Beekeepers should maintain records of problems encountered, control measures implemented, and information that can help determine the sources of problems, including artificial feed options used.Brazil has recently implemented sanitary guidelines for apiculture (Brazil, 2021) and so it will be interesting to follow how well these guidelines are implemented.
Though Brazil has favorable climatic conditions for beekeeping and abundant floral resources year-round (Ara ujo et al., 2018), it is still generally a rustic activity in the country (Ponciano et al., 2013).Also, although there has been considerable research, beekeepers tend to not use the best technological options and the bee health situation in the field has been little studied (Ara ujo et al., 2018).Meliponiculture faces an even bigger challenge since colonies and management techniques are not standardized as they are for A. mellifera (Jaff e et al., 2015).Training for meliponiculture is generally experience based and not objectively organized (Jaff e et al., 2015).This trial and error approach can lead to unnecessary losses; management practices based on research would be more efficient (Kulhanek et al., 2021).
We conclude that the occurrence of pests and pathogens contributes to the loss of managed bee colonies in Brazil and that such losses can be aggravated by inadequate management practices.Greater impacts on bee production can be avoided by taking measures to control and remedy the incidence of pests and pathogens in colonies.Guidance centers and advisors that can best instruct beekeepers on practices can help with colony health.Other recommendations for pest and pathogen control include the inspection and monitoring of pests and pathogens in colonies over space and time, with surveillance being continuous (Brodschneider et al., 2016;Pires et al., 2016).The aforementioned studies, although few and restricted to certain regions of Brazil, have not indicated the isolated involvement of pathogens causing such problems.Until public policies are instituted to promote broader assessments of bee health in the country, both by animal health regulatory agencies and by research funding agencies, little progress will be made in this area of knowledge (Pires et al., 2016).In addition, we believe that colony weakening, as evidenced by the simultaneous occurrence of pests and pathogens, may have been favored by factors not evaluated in this study, such as sublethal effects of pesticides (S anchez-Bayo et al., 2016) or breeding of stingless bee species in areas outside their natural range (Vollet-Neto & Menezes, 2018).Considering the current scenario of facilitated pesticide registrations, the vulnerability of colonies to pests and pathogens can result in even greater losses that may be reflected in the performance of production and export of bee products from Brazil via the international market.Therefore, more in depth investigations are needed into colony loss in Brazil, including the causes, as well as more extensive and continuous monitoring.Reliable and complete inventories made publicly available on virtual platforms can be used to help estimate trends and foster the support of decision makers for actions aimed at investigating and avoiding the decline of bee diversity.
In general, this work raises relevant points that can help in understanding the causes and consequences of the presence of different types of pests and diseases in the disappearance of bees that go beyond the geographic limits of the country.The results of this work can help to reduce the vulnerability of bee colonies to pests and diseases worldwide.

Figure 1 .
Figure 1.Distribution of questionnaire respondents.Dots represent the centroid of the locality mentioned by the beekeepers in reference to their apiaries/melinoparies.Dots in the same location are superimposed.Gray lines represent national and state boundaries.

Figure 2 .
Figure 2. Occurrences of bee pests (A) and disease symptoms (B) based on beekeepers responses.Colony loss percentages due to (C) bee pests and (D) disease symptoms for Apis mellifera and stingless bees.Values followed by the same letter are not significantly different a ¼ 0.05.Lower case letters compare results for Apis mellifera on the left and capital letters compare results for stingless bees on the right."Others" in pests Figure (C): predators, e.g., spiders, geckos, armadillos; "Others" in symptoms Figure (D): deformed wing virus, Brazilian sacbrood virus, among others.Ã The wax moth option and abandoned colony was only included in the Ap-beekeeper questionnaire.FWCF ¼ few combs, food and brood.

Figure 3 .
Figure 3.The influence of land use and occupation around colonies on the average number of bee pests (A) and disease symptoms (B) found for Apis mellifera and stingless bees.Responses from Ap-beekeepers and Sb-beekeepers are shown left and right, respectively.Vegetation = Natural vegetation.Bars represent standard error.Average number of types of pests or disease symptoms of A. mellifera followed by the same small-case letter do not differ significantly (a ¼ 0.05).Average number of types of pests or disease symptoms of stingless bees followed by the same capital letter do not differ significantly a ¼ 0.05.

Figure 4 .
Figure 4.The interaction between colony losses when pests were present/absent as a function of monitoring frequency (A) Apis mellifera (B) Stingless bees.The interaction between colony losses with and without disease symptoms as a function of monitoring frequency (C) A. mellifera (D) Stingless bees.Colony losses as a function of the type of complementary feeding when pests observed in (E) A. mellifera (F) Stingless bees.Colony losses as a function of the type of complementary feed when disease symptoms are observed in (G) A. mellifera (H) Stingless bees.Uppercase letters compare colonies with pests or with disease symptoms, yellow for A. mellifera and orange for stingless bees.Lowercase letters compare colonies without pests or without disease symptoms, in black for both A. mellifera and stingless bees.Roman numbers compare colonies with and without pests/diseases that received the same monitoring frequency (ABCD) or the same complementary feeding type (EFGH).