Search

The study focusses on handling of multiclass imbalanced data on classification of stingless bee samples by employing data balancing techniques, namely Synthetic Minority Oversampling Technique (SMOTE) and Adaptive Synthetic (ADASYN) approach. These techniques are applied in combination with machine learning (ML) algorithms; specifically Random Forest (RF), and Support Vector Machine (SVM), to assess the models’ predictive performance to infer stingless bee samples identities. We studied ML classifier models: RF, RF + SMOTE, RF + ADASYN, SVM, SVM + SMOTE and SVM + ADASYN on the six-class imbalanced dataset of stingless bees morphometrics. Multi-class area under curve (AUC), F1-score, G-mean, balanced accuracy, sensitivity and “No information rate” were used to assess model performance. SMOTE and ADASYN marginally improved the performance of RF and SVM classifiers. SVM outperformed RF, with SVM using SMOTE performing better than with ADASYN. SVM with ADASYN had a lower multi-class AUC (0.9898) and sensitivity (0.956) but a higher F1-score (0.939) compared to SVM with SMOTE (AUC = 0.9918, sensitivity = 0.959, F1-score = 0.934). Overall, SVM with SMOTE was superior to RF with SMOTE. All models except SVM with ADASYN, correctly classified four of the six species, M. (Meliponula) bocandei, M. (Meliplebeia) lendliana, D. schmidti and P. armata but not the two morphs, Meliponula (Axestotrigona) togoensis and Meliponula (Axestotrigona) ferruginea. This study therefore confirms that the impact of imbalanced learning is minimal when classes are separable. Random forest recursive feature elimination technique was used to assess variable importance, guiding future studies on key morphometric measurements to save time and cost while maintaining high classification performance. Our results pave the way for the development of smart and automated machine learning applications to complement the existing methods for the identification of stingless bee species.

Abstract Background Stingless bee honey (SBH) is a natural remedy and therapeutic agent traditionally used by local communities across the (sub-)tropics. Forest SBH represents a prime non-timber forest product (NTFP) with a potential to revitalize indigenous foodways and to generate income in rural areas, yet it is also used in a variety of non-food contexts that are poorly documented in sub-Saharan Africa and that collectively represent a significant part of the local traditional ecological knowledge (TEK) passed on across generations. Documenting TEK of local communities in African tropical forests facing global change is a pressing issue to recognize the value of their insights, to evaluate their sustainability, to determine how they contribute to enhancing conservation efforts, and how TEK generally contributes to the well-being of both the natural environment and the communities that rely on it. This is particularly important to achieve in Kenya’s only tropical rainforest at Kakamega where SBH production and non-food uses have evolved and diversified to a remarkable extent. Methods We used ethnographic techniques and methods, including semi-structured questionnaires and recorded interviews. We used snowball sampling, a non-probability sampling method where new interviewees were recruited by other respondents, to collectively form a sample consisting of 36 interviewees (including only one woman). Results Our results indicate that local communities in Kakamega were able to discriminate between six different and scientifically recognized stingless bee species, and they provided detailed accounts on the species-specific non-food uses of these SBH. Collectively, we recorded an array of 26 different non-food uses that are all passed on orally across generations in the Kakamega community. Conclusion Our results uncover the vast and hitherto unexpected diversity of TEK associated with SBH and pave the way for a systematic survey of SBH and their non-food uses across a network of communities in different environments and with different cultural backgrounds in the Afrotropics. This, along with parallel and more in-depth investigations into honey chemistry, will help develop a comprehensive understanding of SBH, offering insights into holistic ecosystem management, resilience and adaptation while in the mid- to long-term promoting cross-cultural exchanges and pathways for the revitalization of cultural practices and traditions.

Abstract Honey collection evolved from simple honey hunting to the parallel and independent domestication of different species of bees in various parts of the world. In this study, we investigate the extent to which the composition of Apis and stingless bee honeys has been a driver in the selection of different bee species for domestication in Mesoamerica (Mexico) and Asia (Thailand) using a sampling design that combines peak honey profiling by H1 NMR spectroscopy with the collection of honeys from domesticated and undomesticated bee species. Our results show that, independently of the region of the world considered, domesticated stingless bees produce honey whose compositional profiles differ from those of the non-domesticated species and exhibit more similarities towards honeys produced by the domesticated Apis species. Our results provide evidence for the first time that the search for natural sweeteners in the environment by our ancestors led to the parallel and independent domestication of social bees producing honeys with similar compositional profiles.

Worldwide, French beans (Phaseolus vulgaris L.), tomato (Solanum lycopersicum L.), and kales (Brassica oleracea L. var. acephala) are considered economically important food crops. There is a rapid decline in their yield due to severe soil degradation. Thus, high commercial fertilizer inputs are crucial, though they remain expensive and inaccessible to resource poor farmers. We investigated the comparative performance of composted black soldier fly frass fertilizer (BSFFF), conventionally composted brewer's spent grain (BSG), commercial organic fertilizer (Evergrow), and mineral [nitrogen, phosphorus, and potassium (NPK)] fertilizer on growth, yield, N use efficiency, and nutritional quality (crude protein, crude fiber, crude fats, ash, and carbohydrate concentrations) of tomatoes, kales, and French beans under greenhouse and open-field conditions for two seasons. The fertilizers were applied at rates equivalent to 371 kg of N ha−1. For each crop, the plots were treated with sole rates of BSFFF, BSG, Evergrow, and NPK to supply 100% of the N required. Additional treatments included a combination of BSFFF and NPK, and BSG and NPK so that each fertilizer supplies 50% of the N required. The control treatment consisted of unfertilized soil. Results show that vegetable yields achieved using a combination of BSFFF and NPK were 4.5, 2.4, and 5.4-folds higher than the yield from the control treatment for tomatoes, kales, and French beans, respectively. The combined application of BSFFF and NPK produced 22–135%, 20–27%, and 38–50% higher yields than sole NPK for tomatoes, kales, and French beans, respectively, under both greenhouse and open-field conditions. The highest agronomic N use efficiency was achieved in sole BSFFF-treated plots compared to sole BSG and Evergrow. The N taken up by the vegetables was significantly higher when BSFFF and NPK were integrated. Vegetables grown using a combination of BSFFF and NPK had the highest crude protein and ash concentrations. Our findings demonstrate that the integration of BSFFF and NPK in vegetable cropping systems at the recommended rate of 1.24 t ha−1 BSFFF and 322 kg ha−1 NPK would improve soil health, boost yield, and nutritional quality of vegetable crops.

Stingless bees (Apidae: Meliponini) are the most diverse group of corbiculate bees and are important managed and wild pollinators distributed in the tropical and subtropical regions of the globe. However, little is known about their associated beneficial microbes that play major roles in host nutrition, detoxification, growth, activation of immune responses, and protection against pathogens in their sister groups, honeybees and bumble bees. Here, we provide an initial characterization of the gut bacterial microbiota of eight stingless bee species from sub-Saharan Africa using 16S rRNA amplicon sequencing. Our findings revealed that Firmicutes, Actinobacteria, and Proteobacteria were the dominant and conserved phyla across the eight stingless bee species. Additionally, we found significant geographical and host intra-species-specific bacterial diversity. Notably, African strains showed significant phylogenetic clustering when compared with strains from other continents, and each stingless bee species has its own microbial composition with its own dominant bacterial genus. Our results suggest host selective mechanisms maintain distinct gut communities among sympatric species and thus constitute an important resource for future studies on bee health management and host-microbe co-evolution and adaptation.

Honeybees (Apis mellifera) are principal insect pollinators, whose worldwide distribution and abundance is known to largely depend on climatic conditions. However, the presence records dataset on potential distribution of honeybees in Indian Ocean Islands remain less documented. Presence records in shape format and probability of occurrence of honeybees with different temperature change scenarios is provided in this article across Zanzibar Island. Maximum entropy (Maxent) package was used to analyse the potential distribution of honeybees. The dataset provides information on the current and future distribution of the honey bees in Zanzibar Island. The dataset is of great importance for improving stakeholders understanding of the role of temperature change on the spatial distribution of honeybees.

Bee keeping is indispensable to global food production. It is an alternate income source, especially in rural underdeveloped African settlements, and an important forest conservation incentive. However, dwindling honeybee colonies around the world are attributed to pests and diseases whose spatial distribution and influences are not well established. In this study, we used remotely sensed data to improve the reliability of pest ecological niche (EN) models to attain reliable pest distribution maps. Occurrence data on four pests (Aethina tumida, Galleria mellonella, Oplostomus haroldi and Varroa destructor) were collected from apiaries within four main agro-ecological regions responsible for over 80% of Kenya’s bee keeping. Africlim bioclimatic and derived normalized difference vegetation index (NDVI) variables were used to model their ecological niches using Maximum Entropy (MaxEnt). Combined precipitation variables had a high positive logit influence on all remotely sensed and biotic models’ performance. Remotely sensed vegetation variables had a substantial effect on the model, contributing up to 40.8% for G. mellonella and regions with high rainfall seasonality were predicted to be high-risk areas. Projections (to 2055) indicated that, with the current climate change trend, these regions will experience increased honeybee pest risk. We conclude that honeybee pests could be modelled using bioclimatic data and remotely sensed variables in MaxEnt. Although the bioclimatic data were most relevant in all model results, incorporating vegetation seasonality variables to improve mapping the ‘actual’ habitat of key honeybee pests and to identify risk and containment zones needs to be further investigated.

Background: Bee products are gaining interest in the field of research due to their biochemical and nutritive properties. Honey bee products have been researched extensively but little has been done in regards to stingless bees. There are many species of stingless bees including Plebeina armata, which are found in the Afrotropics. They are underground nesting and produce honey, propolis, wax, pollen and bee bread. These products are known to be rich in polyphenols that comprise of flavonoids and phenolic. In our study, we analysed colorimetrically the total flavonoid, phenolic content and radical scavenging activity of honey (n = 22) and propolis (n = 25) from Bomet, Kisii and Maralal in Kenya. Results: Honey and propolis had total flavonoid content of 12.00–22.67 mg QE/100 g and 288.15–944.76 mg QE/100 g while total phenolic content was 87.01–239.93 mg GAE/100 g and 524.14–1225.01 mg GAE/100 g, respectively. In considerations to the regions, Maralal had the highest phenolic and flavanoid content followed by Bomet and Kisii was the least. The same trend was observed in the radical scavenging activity. Except for the total flavonoid content in honey, the difference was significant (p < 0.05). Conclusion: The polyphenol content of both honey and propolis of P. armata are equally affected by geographical location as a result of different vegetation. They are good source of antioxidants, which can be utilized in diet due to their radical scavenging properties. [ABSTRACT FROM AUTHOR]

An increasing demand for pollination services highlights the need for research on alternative pollinators for greenhouse and open field food crops. We compared the foraging behaviour and effectiveness of seven endemic African Meliponinae (Meliponula bocandei, Dactylurina schmidti, Plebeina hildebrandti, Meliponula lendliana, Hypotrigona gribodoi, Meliponula ferruginea, Meliponula togoensis) as alternative pollinators to honey bees of greenhouse cultivated Cantaloupe melons. It was observed that honey bees started foraging sooner after introduction in the greenhouse (av. 4 days) than the stingless bee species (8 to 16 days). Stingless bees were mainly harvesting nectar (ca. 2/3 of all flower visits) and M. ferruginea, M. togoensis, H. gribodoi and P. hildebrandti spent more time collecting nectar than other stingless bee species. Fruit maturation was significantly faster when flowers were pollinated by stingless bees compared to honey bees, with the shortest fruit maturation time found when flowers were pollinated by M. bocandei and H. gribodoi. All treatments resulted in fruits of similar roundness and ellipsoid ratios. The highest fruit weight and fruit volume were obtained from flowers pollinated by either manual pollination, H. gribodoi, M. bocandei, M. lendliana and P. hildebrandti compared to honey bees. Correspondingly, pollination by the latter 5 species resulted in the highest seed counts per fruit, although these differences did not affect seed weight or volume. Altogether, our results indicate that stingless bees such as H. gribodoi, M. bocandei, M. lendliana and P. hildebrandti are more efficient pollinators of sweet melon than A. m. scutellata and can be recommended for use in greenhouse crops. [ABSTRACT FROM AUTHOR]

This study assessed the nontarget effect of entomopathogenic fungi on the Western honey bee Apis mellifera L. and the African stingless bee Meliponula ferruginea Cockrell (Hymenoptera: Apidae). Pathogenicity of five Metarhizium anisopliae (ICIPE 7, ICIPE 20, ICIPE 62, ICIPE 69, and ICIPE 78) (Metschnikoff) Sorokin (Hypocreales: Clavicipitaceae) and one of Beauveria bassiana (ICIPE 284) (Balsamo) Vuillemin (Hypocreales: Cordicipitaceae) isolates were evaluated on bees at 108 conidia/ml. Conidial acquisition was evaluated immediately after exposure. Apis mellifera acquired more conidia (2.8 × 104–1.3 × 105 conidia per bee) compared to M. ferruginea (1.1 × 104–2.3 × 104 conidia per bee). In the bioassay with A. mellifera, ICIPE 7, ICIPE 20, and ICIPE 69 moderately reduced the survival by 16.9, 17.4, 15.3%, with lethal times LT10 = 7.4, 7.6, 8.1 d and LT25 = 8.7, 10.0, 9.9 d, respectively. The three isolates caused A. mellifera mycosis of 11.6–18.5%. None of the isolates had a significant effect on M. ferruginea. The tested isolates are nontoxic to bees according to the International Organization of Biological Control (IOBC) classification. However, the effect of ICIPE 7, ICIPE 20, and ICIPE 69 merits further studies on bee colonies, especially those of A. mellifera, under field conditions. Graphical Abstract [ABSTRACT FROM AUTHOR]

A study was carried out in Kakamega forest, in the western region of Kenya, to evaluate the effectiveness of the stingless bee Hypotrigona gribodoi ( Magretti, 1884) on the pollination of green pepper. Three treatments were applied and consisted of self-pollination, pollination by feral pollinators in the open field and pollination by H. gribodoi in a net cage. The differences in fruit yield and seed quality were compared among treatments. Flowers pollinated by H. gribodoi produced the heaviest fruits with the highest seed numbers followed by feral pollinators and lastly self-pollinated flowers. Moreover, seeds were significantly bigger in size in fruits resulting from flowers pollinated by H. gribodoi compared to fruits obtained from self-pollinated flowers or flowers pollinated by feral insects. We, therefore, conclude that H. gribodoi is an efficient pollinator of green pepper in the tropical region of East Africa. [ABSTRACT FROM AUTHOR]