Introduction
Species respond differently to anthropogenic habitats, such as villages and urban areas (McKinney, 2008; Otto, 2018; Szulkin et al., 2020). While generally these ecological changes have negative consequences, a handful of species have successfully exploited these novel human environments (Johnson & Munshi-South, 2017). Among them, one of the most notorious examples is the yellow fever mosquito, Aedes aegypti , a major vector of several arboviral diseases that cause millions of infections each year, including yellow fever, dengue, chikungunya, and Zika (Aubry et al., 2018; World Health Organization, 2014). The mosquitoes’ high efficacy at transmitting diseases stems partly from their adaptation to human-made domestic habitats, assuring close contact with humans (Carvalho & Moreira, 2017; Fontenille & Powell, 2020).
The recent evolutionary history of Ae. aegypti is strongly linked to human activities (Powell et al., 2018; Powell & Tabachnick, 2013).Ae. aegypti originated on southwestern Indian Ocean islands and moved to continental Africa around 85,000 years ago before spreading across the continent in tropical forests (Soghigian et al., 2020). Probably five to ten thousand years ago, this species invaded human settlements and likely evolved domestic adaptations (Crawford et al., 2017; Kotsakiozi et al., 2018). This domestication prepared them for later spread to the rest of the world a few hundred years ago, likely from West Africa and facilitated by human movements (Brown et al., 2014; Gloria‐Soria et al., 2016; Powell & Tabachnick, 2013). Extant mosquitoes in and out of Africa show a clear genetic distinction (Gloria‐Soria et al., 2016, but see exceptions in Kotsakiozi et al., 2018 and Rose et al., 2020), which roughly matches the two classical subspecies: Ae. aegypti formosus (Aaf) and Ae. aegypti aegypti (Aaa) , respectively. Complexities exist in this subspecies definition (Powell & Tabachnick, 2013), but in this paper, we refer to them simply based on their geographic range (in or out of Africa). Non-African Aaa is mostly human specialized and lives only in domestic settings such as urban area (McBride, 2016; Powell & Tabachnick, 2013), except for a few forest-living populations in the Caribbean and Argentina (Chadee et al., 1998; Mangudo et al., 2015). On the other hand, African Aaf inhabits both forest and domestic habitats (Kotsakiozi et al., 2018; Paupy et al., 2014; Sylla et al., 2009), with the latter likely representing an intermediate step towards true human specialization outside Africa (i.e., Aaa ).
Despite a well-characterized evolutionary path of Ae. aegyptideduced from genetic data, how this species initially moved into domestic habitats in Africa is not fully understood. In addition, most mosquito species (over 3,500 species) have not colonized domestic habitats, including many African mosquitoes that share the same forest habitats as Ae. aegypti (Clements, 1999). This raises the question of why Ae. aegypti , among only a few mosquito species, was able to invade domestic habitats (Carvalho & Moreira, 2017; Fontenille & Powell, 2020). Addressing these questions could help us further understand the unique evolutionary history of this epidemiologically important species, provide insights into mosquito control, and possibly predict other emerging disease vectors.
For Ae. aegypti , the transition from the ancestral forest habitat to human settlement involved two major behavioral changes: a preference for humans as a blood source (McBride, 2016) and using human-made containers for larval breeding (Day, 2016). Ancestral forest-livingAaf in Africa is a feeding generalist and bites wildlife for blood, while Aaa out of Africa specializes in biting humans (Powell et al., 2018). Recent studies have demonstrated the variations of blood-feeding preference across Africa in different habitats and between the two subspecies (McBride, 2016; McBride et al., 2014; Rose et al., 2020). They also identified dry season intensity and human population density as two main ecological drivers for the evolution of feeding preference for humans in Africa.
In comparison, larval breeding sites are relatively understudied, especially in Africa. Aedes aegypti lay eggs at the edge of small water containers, i.e., oviposition sites that becomes larval breeding sites (Christophers, 1960). Non-African Aaa uses various artificial containers, consistent with being a human specialist (Day, 2016; Swan et al., 2018; Vezzani, 2007; Yee, 2008). In Africa,Aaf in the forest and domestic habitats utilize different larval breeding sites: the former uses natural containers like rainwater-filled tree holes and rock pools (Lounibos, 1981), while the latter resemblingAaa , relies mostly on artificial containers, such as plastic buckets, tires, and clay pots (Leahy et al., 1978; McBride et al., 2014; Petersen, 1977). Some artificial containers hold human-stored water year-round and could provide valuable or even the only available larval habitats during the dry season when natural containers dry out. Therefore, it is hypothesized that seeking human water storage for oviposition during the dry season likely drove Ae. aegypti to enter domestic habitats, leading to the evolution of feeding preference for humans (Petersen, 1977; Powell et al., 2018; Rose et al., 2020). Despite the presumed key role of larval breeding habitats in the domestic adaptation of Ae. aegypti , few studies have characterized natural versus artificial larval breeding sites in Africa (Dickson et al., 2017).
If a substantial difference exists between natural and artificial containers, it could pose challenges to both female oviposition and larvae development. Oviposition preference and larvae performance are likely associated (Wong et al., 2012) but not always aligned (Albeny-Simoes et al., 2014; Refsnider & Janzen, 2010). Given the large variety of larval habitats, it is possible that ancestral Ae. aegypti were generalist egg-layers, and the larvae can tolerate a wide range of conditions, which allows them to take advantage of artificial containers in harsh environments (Powell et al., 2018; Rose et al., 2020). Prolonged breeding in distinct containers could then lead to ecological divergence (Gimonneau et al., 2010; Schluter, 2000; Shafer & Wolf, 2013), resulting in adaptations and specializations to each container type. Such behavioral differentiation, in turn, could facilitate population segregation (Ayala et al., 2011; Nosil et al., 2009; Servedio et al., 2011). Indeed, a few studies have implied that oviposition divergence may have emerged between Aaa andAaf (Leahy et al., 1978; Petersen, 1977). Larvae of the two subspecies also showed higher fitness in containers representing their respective preferred habitats (Saul et al., 1980). However, previous work only compared the two subspecies, i.e., two ends of the domestication history of Ae. aegypti . Whether divergence in oviposition or larval performance already exist within AfricanAaf living in forest versus domestic habitats remains largely unknown. Examining this potential incipient divergence could provide valuable insights on when, where, and how Ae. aegypti adapted to domestic habitats.
In this study, we characterized the environmental conditions ofAe. aegypti larval breeding sites in forest and domestic habitats and examined whether oviposition divergence has evolved. We focused on two locations in Africa, La Lopé in Gabon and Rabai in Kenya. Mosquitoes in both locations are Aaf based on their morphology and broad genetic pattern (Kotsakiozi et al., 2018; Xia et al., 2020), but can be found in forest and village environments in close proximity. In each location, forest and village populations showed little genetic differentiation (Xia et al., 2020), suggesting local habitat expansion instead of external introduction. Therefore, Ae. aegypti in these two locations possibly exemplify the initial colonization step of domestic habitats. We first compare the physical characteristics, competition and predation, bacterial profiles, and chemical volatiles of natural and artificial containers used as larval breeding sites between habitats (forest and village). Many of these environmental variables have been shown to affect Ae. aegypti oviposition (Afify & Galizia, 2015; Harrington et al., 2008; Ponnusamy et al., 2008; Reiskind & Zarrabi, 2012; Zahiri & Rau, 1998). Therefore, we also investigated the oviposition choices of forest and domestic Aaf towards some variables that showed the greatest differences between natural and artificial containers. This allowed us to examine whether the mosquitoes in different habitats remained oviposition generalist or have developed behavioral specialization.