Traps baited with attractive lures are increasingly used at entry-points and surrounding natural areas to intercept exotic wood-boring beetles accidentally introduced via international trade. Several trapping variables can affect the efficacy of this activity, including trap color. In this study, we tested whether species richness and abundance of jewel beetles (Buprestidae), bark and ambrosia beetles (Scolytinae), and their common predators (i.e., checkered beetles, Cleridae) can be modified using trap colors different to those currently used for surveillance of jewel beetles and bark and ambrosia beetles (i.e., green or black). We show that green and black traps are generally efficient, but also that many flower-visiting or dark-metallic colored jewel beetles and certain bark beetles are more attracted by other colors. In addition, we show that checkered beetles have color preferences similar to those of their Scolytinae preys, which limits using trap color to minimize their inadvertent removal. Overall, this study confirmed that understanding the color perception mechanisms in wood-boring beetles can lead to important improvements in trapping techniques and thereby increase the efficacy of surveillance programs.

Along environmental gradients, communities are expected to be filtered from the regional species pool by physical constraints, resource availability, and biotic interactions. This should be reflected in species trait composition. Using data on species-rich moth assemblages sampled by light traps in a lowland rainforest landscape in Costa Rica, we show that moths in two unrelated clades (Erebidae-Arctiinae; Geometridae) are much smaller-sized in oil palm plantations than in nearby old-growth forest, with intermediate values at disturbed forest sites. In old-growth forest, Arctiinae predominantly show aposematic coloration as a means of anti-predator defense, whereas this trait is much reduced in the prevalence in plantations. Similarly, participation in Müllerian mimicry rings with Hymenoptera and Lycidae beetles, respectively, is rare in plantations. Across three topographic types of old-growth forests, community-weighted means of moth traits showed little variation, but in creek forest, both types of mimicry were surprisingly rare. Our results emphasize that despite their mobility, moth assemblages are strongly shaped by local environmental conditions through the interplay of bottom–up and top–down processes. Assemblages in oil palm plantations are highly degraded not only in their biodiversity, but also in terms of trait expression.

Dietary macro-nutrients (i.e., carbohydrates, protein, and fat) are important for bee larval development and, thus, colony health and fitness. To which extent different diets (varying in macro-nutrient composition) affect adult bees and whether they can thrive on nectar as the sole amino acid source has, however, been little investigated. We investigated how diets varying in protein concentration and overall nutrient composition affected consumption, longevity, and breeding behavior of the buff-tailed bumble bee, Bombus terrestris (Hymenoptera: Apidae). Queenless micro-colonies were fed either natural nutrient sources (pollen), nearly pure protein (i.e., the milk protein casein), or sucrose solutions with low and with high essential amino acid content in concentrations as can be found in nectar. We observed micro-colonies for 110 days. We found that longevity was highest for pure pollen and lowest for pure sucrose solution and sucrose solution supplemented with amino acids in concentrations as found in the nectar of several plant species. Adding higher concentrations of amino acids to sucrose solution did only slightly increase longevity compared to sucrose alone. Consequently, sucrose solution with the applied concentrations and proportions of amino acids or other protein sources (e.g., casein) alone did not meet the nutritional needs of healthy adult bumble bees. In fact, longevity was highest and reproduction only successful in micro-colonies fed pollen. These results indicate that, in addition to carbohydrates and protein, adult bumble bees, like larvae, need further nutrients (e.g., lipids and micro-nutrients) for their well-being. An appropriate nutritional composition seemed to be best provided by floral pollen, suggesting that pollen is an essential dietary component not only for larvae but also for adult bees.

Insect brood parasites have evolved a variety of strategies to avoid being detected by their hosts. Few previous studies on cuckoo wasps (Hymenoptera: Chrysididae), which are natural enemies of solitary wasps and bees, have shown that chemical mimicry, i.e., the biosynthesis of cuticular hydrocarbons (CHC) that match the host profile, evolved in several species. However, mimicry was not detected in all investigated host-parasite pairs. The effect of host range as a second factor that may play a role in evolution of mimicry has been neglected, since all previous studies were carried out on host specialists and at nesting sites where only one host species occurred. Here we studied the cuckoo wasp Parnopes grandior, which attacks many digger wasp species of the genus Bembix (Hymenoptera: Crabronidae). Given its weak host specialization, P. grandior may either locally adapt by increasing mimicry precision to only one of the sympatric hosts or it may evolve chemical insignificance by reducing the CHC profile complexity and/or CHCs amounts. At a study site harbouring three host species, we found evidence for a weak but appreciable chemical deception strategy in P. grandior. Indeed, the CHC profile of P. grandior was more similar to all sympatric Bembix species than to a non-host wasp species belonging to the same tribe as Bembix. Furthermore, P. grandior CHC profile was equally distant to all the hosts' CHC profiles, thus not pointing towards local adaptation of the CHC profile to one of the hosts' profile. We conducted behavioural assays suggesting that such weak mimicry is sufficient to reduce host aggression, even in absence of an insignificance strategy, which was not detected. Hence, we finally concluded that host range may indeed play a role in shaping the level of chemical mimicry in cuckoo wasps.

One of the pronounced global challenges facing ecologists is how to feed the current growing human population while sustaining biodiversity and ecosystem services. To shed light on this, I investigated the impact of human land use on bee diversity and plant-pollinator interactions in Tanzania Savannah ecosystems. The thesis comprises the following chapters: Chapter I: General Introduction This chapter provides the background information including the study objectives and hypotheses. It highlights the ecological importance of bees and the main threats facing bee pollinators with a focus on two land-use practices namely livestock grazing and agriculture. It also highlights the diversity and global distribution of bees. It further introduces the tropical savannah ecosystem, its climate, and vegetation characteristics and explains spectacular megafauna species of the system that form centers of wildlife tourism and inadequacy knowledge on pollinators diversity of the system. Finally, this chapter describes the study methodology including, the description of the study area, study design, and data collection. Chapter II: Positive effects of low livestock grazing intensity on East African bee assemblages mediated by increases in floral resources The impact of livestock grazing intensity on bee assemblage has been subjected to research over decades. Moreover, most of these studies have been conducted in temperate Europe and America leaving the huge tropical savannah of East Africa less studied. Using sweep netting and pan traps, a total of 183 species (from 2,691 individuals) representing 55 genera and five families were collected from 24 study sites representing three levels of livestock grazing intensity in savannah ecosystem of northern Tanzania. Results have shown that moderate livestock grazing slightly increased bee species richness. However, high livestock grazing intensity led to a strong decline. Besides, results revealed a unimodal distribution pattern of bee species richness and mean annual temperature. It was also found that the effect of livestock grazing and environmental temperature on bee species richness was mediated by a positive effect of moderate grazing on floral resource richness. The study, therefore, reveals that bee communities of the African savannah zone may benefit from low levels of livestock grazing as this favors the growth of flowering plant species. A high level of livestock grazing intensity will cause significant species losses, an effect that may increase with climatic warming. Chapter III: Agricultural intensification with seasonal fallow land promotes high bee diversity in Afrotropical drylands This study investigated the impact of local agriculture intensification on bee diversity in the Afro tropical drylands of northern Tanzania. Using sweep netting and pan traps, a total of 219 species (from 3,428 individuals) representing 58 genera and six families were collected from 24 study sites (distributed from 702 to 1708 m. asl) representing three levels of agriculture intensity spanning an extensive gradient of mean annual temperature. Results showed that bee species richness increased with agricultural intensity and with increasing temperature. However, the effects of agriculture intensity and temperature on bee species richness were mediated by the positive effects of agriculture and temperature on floral resource richness used by bee pollinators. Moreover, results showed that variation of bee body sizes increases with agricultural intensification, “that effect”, however, diminished in environments with higher temperatures. This study reveals that bee assemblages in Afrotropical drylands benefit from agriculture intensification in the way it is currently practiced. Further intensification, including year-round irrigated crop monocultures and extensive use of agrochemicals, is likely to exert a negative impact on bee diversity and pollination services, as reported in temperate regions. Moreover, several bee species were restricted to natural savannah habitats. Therefore, to conserve bee communities in Afro tropical drylands and guarantee pollination services, a mixture of savannah and agriculture, with long periods of fallow land should be maintained. Chapter IV: Impact of land use intensification and local features on plants and pollinators in Sub-Saharan smallholder farms For the first time in the region, this study explores the impact of land-use intensification on plants and pollinators in Sub-Saharan smallholder farms. The study complemented field surveys of bees with a modern DNA metabarcoding approach to characterize the foraged plants and thus built networks describing plant-pollinator interactions at the individual insect level. This information was coupled with quantitative traits of landscape composition and floral availability surrounding each farm. The study found that pollinator richness decreased with increasing impervious and agricultural cover in the landscape, whereas the flower density at each farm correlated with pollinator richness. The intensification of agricultural land use and urbanization correlated with a higher foraging niche overlap among pollinators due to the convergence of individuals' flower-visiting strategies. Furthermore, within farms, the higher availability of floral resources drove lower niche overlap among individuals, greater abundance of flower visitors shaped higher generalization at the networks level (H2I), possibly due to increased competition. These mechanistic understandings leading to individuals’ foraging niche overlap and generalism at the network level, could imply stability of interactions and the pollination ecosystem service. The integrative survey proved that plant-pollinator systems are largely affected by land use intensification and by local factors in smallholder farms of Sub-Saharan Africa. Thus, policies promoting nature-based solutions, among which the introduction of more pollinator-friendly practices by smallholder farmers, could be effective in mitigating the intensification of both urban and rural landscapes in this region, as well as in similar Sub-Saharan contexts. Chapter V: A synopsis of the Bee occurrence data of northern Tanzania This study represents a synopsis of the bee occurrence data of northern Tanzania obtained from a survey in the Kilimanjaro, Arusha, and Manyara regions. Bees were sampled using two standardized methods, sweep netting and colored pan traps. The study summed up 953 species occurrences of 45 species belonging to 20 genera and four families (Halictidae, Apidae, Megachilidae, and andrenidae) A. This study serves as the baseline information in understanding the diversity and distribution of bees in the northern parts of the country. Understanding the richness and distribution of bees is a critical step in devising robust conservation and monitoring strategies for their populations since limited taxonomic information of the existing and unidentified bee species makes their conservation haphazard. Chapter VI: General discussion In general, findings obtained in these studies suggest that livestock grazing and agriculture intensification affects bee assemblages and floral resources used by bee pollinators. Results have shown that moderate livestock grazing intensity may be important in preserving bee diversity. However, high level of livestock grazing intensity may result in a strong decline in bee species richness and abundance. Moreover, findings indicate that agriculture intensification with seasonal fallow lands supports high floral resource richness promoting high bee diversity in Afrotropical drylands. Nonetheless, natural savannahs were found to contain unique bee species. Therefore, agriculture intensification with seasonal fallow should go in hand with conserving remnant savannah in the landscapes to increase bee diversity and ensure pollination services. Likewise, findings suggest that increasing urbanization and agriculture cover at the landscape level reduce plant and pollinator biodiversity with negative impacts on their complex interactions with plants. Conversely, local scale availability of floral resources has shown the positive effects in buffering pollinators decline and mitigating all detrimental effects induced by land-use intensification. Moreover, findings suggest that the impact of human land use (livestock grazing and agriculture) do not act in isolation but synergistically interacts with climatic factors such as mean annual temperature, MAT. The impact of MAT on bee species richness in grazing gradient showed to be more detrimental than in agriculture habitats. This could probably be explained by the remaining vegetation cover following anthropogenic disturbance. Meaning that the remaining vegetation cover in the agricultural gradient probably absorbs the solar radiations hence reducing detrimental effect of mean annual temperature on bee species richness. This one is not the case in grazing gradient since the impact of livestock grazing is severe, leaving the bare land with no vegetation cover. Finally, our findings conclude that understanding the interplay of multiple anthropogenic activities and their interaction with MAT as a consequence of ongoing climate change is necessary for mitigating their potential consequences on bee assemblages and the provision of ecosystem services. Morever, future increases in livestock grazing and agriculture intensification (including year-round crop irrigated monocultures and excessive use of agrochemicals) may lead to undesirable consequences such as species loss and impair provision of pollination services.