There are a multitude of reproductive strategies utilized across the animal kingdom [6]. However, most fit into one of three categories: oviparity (development and deposition of eggs in early embryonic development); ovoviviparity (deposition of eggs at later stages of embryonic development or live offspring early in their development); or true viviparity. True viviparous reproduction is characterized by the obligate retention and nourishment of developing offspring in the maternal reproductive tract. This is often associated with reproductive tract modifications that facilitate gestation for immature development and birth. An important distinction between true viviparity and ovoviviparity is the general inability of offspring of viviparous species to survive premature birth or egg deposition. There exist many diverse implementations of viviparity for animals, where the provision of extra-vitelline nutrition to developing progeny usually represents the most advanced states [6–8].

While pregnancy is most commonly associated with mammals, viviparous reproduction represents a striking example of convergent evolution in the animal kingdom [6]. This type of reproduction is well represented in arthropods and even organisms as basal as sea sponges (Porifera). Independent evolutionary origins of viviparous reproduction have been identified in over 150 invertebrate lineages alone [6–9]. Present in at least 13 orders of insects, viviparity has at least 62 independent origins within the 22 families of Dipterans [6,9].

Matrotrophic viviparity, whereby mothers provision developing offspring with nutrients, is also widespread across the animal kingdom (21 of the 34 animal phyla). It is estimated that 213 chordate and 45 arthropod families have member species that are matrotrophic viviparous [8]. Despite the widespread incidence of diverse forms of viviparous reproduction, the current understanding of matrotrophic viviparity in insects is sparse outside a few systems [7,8,10,11].

The lack of information on viviparous reproduction outside of mammals, specifically in ectotherms, hinders our ability to predict how these species will respond to climate change. The dramatic physiological differences between oviparous and viviparous species mean that broad comparisons between these groups to predict the impacts of climate on viviparous species could be misleading due to incorrect assumptions about the underlying biology [11].

Tsetse flies have three significant morphological adaptations that facilitate adenotrophic viviparity. Tsetse ovaries only produce a single oocyte per gonotrophic cycle; the reproductive tract is expanded and lined with tracheated muscle-lined epithelial cells that form the uterus [7,13]; and the female accessory gland is expanded into a large, branched organ [7,10,13] attached directly to to the uterus. The tsetse uterus can accommodate a single larva each gonotrophic cycle [7,13]. A unique aspect of tsetse biology is the provisioning of nutritive secretions (fly milk) to the developing larvae that allow a 100-fold (dry mass) growth in utero [14,15]. Milk protein components consist of a lipocalin, Hippoboscoidea-specific milk proteins, transferrin, acid sphingomyelinase, and other minor constituents [1,16]. In addition to essential proteins, the milk secretions contain bacterial symbionts transmitted from the mother to the larva [1,7,17].

These adaptations to viviparity come with significant nutritional and physiological costs. An oviparous organism under nutritional stress can modulate its reproductive output via oocyte resorption to balance the output relative to acquired nutrients. Glossina do not have this option, and if adequate nutrients are not obtained during pregnancy, they must abort the developing larvae and start the cycle again [7,16].

The intense nutritional demands of pregnancy have driven distinct adaptations for diet and nutrient processing. Across all tsetse species, both sexes feed exclusively on vertebrate blood, which contains abundant proteins and lipids but few carbohydrates and specific micronutrients. After locating a host, feeding requires a complex suite of salivary proteins unique to Glossina species [18,19]. Immediately after feeding, blood moves into the crop and passes through the proventriculus, where the peritrophic matrix (PM) envelops the blood. Rapid secretion of the excess water in the blood through diuresis is critical to digestion [20]. A suite of specific enzymes digest the lipids and proteins [21–23]. The digestion process involves the secretion of multiple peptidases, proteinases, and lipases, which induce rapid breakdown of blood components. Free amino acids and lipids generated from the breakdown of blood are metabolized into lipid reserves stored in the fat body tissue [7,24]. The digestion process is temperature-dependent, where the process of the blood meal occurs more rapidly at higher temperatures [25]. Due to their dependence on a high protein diet, Glossina relies on protein-based metabolism. As a result, tsetse have likely evolved to use proline as an energy source to fuel their biological processes [26,27].

Key for metabolism, tsetse flies harbor the obligate mutualist Wigglesworthia, a symbiotic association that dates 50–80 million years [28]. This bacteria is found both intracellularly within bacteriocytes that form tsetse’s bacteriome organ, located adjacent to the anterior midgut, and extracellularly within maternal milk. Wigglesworthia is capable of B vitamin synthesis, which is critical to supplement tsetse diet of vertebrate blood [27,29]. Fly metabolism is drastically impaired without B vitamin provisioning from Wigglesworthia [1,30]. Tsetse flies have other known bacterial associates, primarily Sodalis and Wolbachia, that impact their biology. However, these bacteria are not found in all species and populations [1,17,30]. Notably, the elimination of the symbionts leads to suppression of reproduction that can be recovered by supplementing a cocktail of B vitamins [27,29].

African trypanosomes must complete 14–28 days of developing in tsetse flies to be transmitted to vertebrate hosts [31]. At the ingestion of infectious blood, the parasites must undergo multiple stages of differentiation and replication to move from the midgut through the cardia organ (proventriculus) and then into the foregut, salivary glands, and mouthparts, where they undergo the last differentiation into vertebrate-infectious metacyclics. Despite likely ingesting many trypanosome-infected blood meals, only a low percentage of challenged tsetse (1–10%) become capable of onward transmission [32]. This low infection prevalence is reflective of robust passive and active gut-associated innate immune mechanisms, which include reactive oxygen intermediates (ROIs) [33,34], antimicrobial factors [34–36] and their type II peritrophic matrix (PM). The PM is a sleeve-like structure secreted by the cardia organ [37] lines the midgut and separates epithelial cells from ingested products and pathogens [38]. Young, unfed (‘teneral’) flies have an immature PM structure – which does not fully line the midgut – and this is one potential mechanism explaining the higher susceptibility of becoming infected among young flies taking their first blood meal [39]. Mature adult tsetse, in contrast, have a fully formed PM and are more refractory to trypanosome infection than young adults[39,40].

Being ectotherms, tsetse body temperatures are primarily dictated by environmental conditions [41]. Behaviorally, tsetse flies can buffer changes by moving into protected microhabitats, such as warthog dens, that remain cooler and more humid, protecting both the mother and deposited larvae [42]. Warmer ambient environmental temperature will exponentially increase resting and active metabolism, resulting in more rapid blood meal digestion [25], and likely leading to increased blood feeding frequency (summarized in Figure 1). These physiological and behavioral changes may cause increased disease transmission risk, given that blood meals are tsetse’s only source of nutrients, and thus increase R0 in classic disease transmission equations. However, transmission risk may be lower as vertebrate host populations become scarce or patchy (due to direct climatic factors or changes in suitable habitat), negatively impacting tsetse population sizes.

The effects of thermal stress on the interactions between tsetse flies and their symbionts have been limited. However, continual exposure to temperatures above 37°C will lead to death in Sodalis [43]. While direct studies are lacking on Wigglesworthia, prolonged exposure to high temperatures will likely lead to symbiont elimination resulting in reduced fly health and female sterility [44]. There are few studies on thermal stress and parasite dynamics in tsetse [45], but both direct and indirect impacts of climate change are anticipated. The parasites are impacted by several factors, including environmental stress, tsetse nutrition, and host biology (e.g., reproductive status and age) interactions. Moreover, parasite dynamics within tsetse are impacted by symbionts, as these influence tsetse immune and nutritional state. Thus, climate change directly and indirectly impacts all facets of the host-microbe-parasite system and the interactions therein.

Given that climatic factors likely played a significant role in the evolution of tsetse viviparity, climate change conditions will impact tsetse viviparity in myriad ways (summarized in Figure 2). On the one hand, producing so few offspring (~10 per lifetime), each requiring high maternal investment, renders tsetse particularly vulnerable to climate change. On the other hand, the benefits of offspring retention, allowing females to reproduce across the year and thermoregulate via behavior and without the need to select suitable larval habitats, may provide some buffer. Most likely, tsetse evolved this costly reproductive strategy under optimal thermal conditions with abundant hosts [7,46], suggesting that climate change is likely to have a substantial negative impact on tsetse viviparous processes. Detailed population dynamics modeling is required to manage this complexity (e.g. [2,3,5]), incorporating how aspects of tsetse reproduction vary with climatic factors, not all pathways of which are fully understood.

We anticipate females will produce smaller and fewer offspring under warmer temperatures. This is because warmer temperature increases the developmental rate [47], with less time to develop into a larger pupa. These smaller offspring, in turn, will have reduced survival in hotter months due to their limited reserves for completion of metamorphosis and locating their first blood meal [48]. Moreover, females have higher rates of abortion when it is warmer [49], a potential mechanism being ovarian damage due to heat stress that prevents subsequent investment in nonviable progeny [50]. Recent experimental work simulating heatwaves has shown that tsetse females take longer to produce their first offspring after a heatwave, possibly due to early egg-stage abortion or delayed ovulation [51].

Tsetse survival and reproduction could be significantly impacted by nighttime temperatures, which appear to be rising disproportionately [52]. Tests focusing on the effects of nighttime temperatures on tsetse have yet to be conducted, representing a promising avenue of research to better understand potential outcomes. Importantly, as tsetse flies show very little activity during the night [53] behavioral regulation of thermal status may be reduced. While slight shifts in timing of activity may be possible under strong selection (as per Anopheles mosquitoes shifting feeding behaviour in response to selection pressure of widespread use of bednets [54]). It is unlikely that tsetse – with a visual system highly adapted to being diurnal – would shift to being more active when there is not sufficient light for movement. Direct heating may also lead to cumulative cellular heat injury, which, once critical environmental temperature thresholds are surpassed, can result in exponential organismal failure and inability to recover from high temperatures [55].

Climate change produces increased environmental temperatures and more frequent and unpredictable droughts [56]. Drought-induced changes in vertebrate host behavior and abundance may indirectly impact tsetse flies, given their reliance on blood for water requirements [20]. Locating vertebrates has been suggested as a more reliable water source during scarcity of free water or water from plants, potentially driving increased blood feeding during drought, as in other vectors [57]. Furthermore, as blood is the sole water source for tsetse, teneral flies may be highly susceptible to dehydration-induced mortality if emergence occurs under dry conditions with few hosts available [58]. There are species-specific differences in water balance parameters among tsetse flies [59], suggesting that different ecotypes will vary in drought tolerance. As such, reduced host availability coupled with warmer and dry conditions reducing suitable flight time for tsetse and increasing metabolic rate can impose substantial nutritional and water balance stress. Such stress produces smaller offspring [60] and may lead to increased starvation and dehydration of teneral flies if they run out of lipid or water reserves before their first blood meal.