Introduction

Wolbachia is one of the best-known groups of heritable endosymbionts, widely distributed in arthropods and some nematodes (Hertig 1936; Sironi et al. 1995; Werren 1997). These bacteria form one of the most abundant and diverse groups of symbionts on earth: an estimated 40-60% of arthropod species are infected with Wolbachia strains (Zug & Hammerstein 2012; Weinert et al. 2015). Their ability to induce reproduction manipulations (Rousset et al. 1992; Werren et al. 2008) and their application in vector born disease control (Kambris et al. 2009; Hoffmann et al. 2015; Ross et al. 2019) are key aspects of Wolbachia that have been studied extensively in the last two decades.
Similar to many other symbionts, the current distribution of Wolbachia results from three major processes: co-diversification with the host clade, shifting between host species, and symbiont loss (Thompson 1987; Charleston & Perkins 2006). Although co-speciation is common among Wolbachia strains belonging to supergroups C and D in their nematode hosts (Bandi et al. 1998; Fenn & Blaxter 2004) and certain strains of supergroup F infecting bed bugs (Balvín et al. 2018), many studies have failed to find evidence of codiversification between Wolbachia strains of supergroups A/B and arthropods (e.g., in fig wasps (Shoemaker et al. 2002), ants (Frost et al. 2010), butterflies (Ahmed et al. 2016), bees (Gerth et al. 2013), and collembolans (Ma et al. 2017)). In the absence of co-diversification, host-shifting is the alternative hypothesis to explain the current distribution of Wolbachia (reviewed in Sanaei et al. (2021a)).Wolbachia shift hosts when a given strain infects a novel arthropod species, mostly through horizontal transfer (Boyle et al. 1993; Heath et al. 1999) and possibly occasionally through hybridisation (Turelli et al. 2018; Cooper et al. 2019). The possibility of host-shift events in Wolbachia has been confirmed through numerous transinfection studies when a strain is artificially introduced to an uninfected species (reviewed in Hughes & Rasgon, 2014), and the existence of “super spreader strains” that infect host species which are phylogenetically distantly related (e.g., ST41 strain type in Lepidoptera (Ilinsky & Kosterin 2017)). Physical transfer ofWolbachia from donor to recipient species is the first step of host-shifting, achieved via various ”routes of transfer” and usually facilitated by a biological vector or a suitable environmental medium (Vavre et al. 2003; Riegler et al. 2004). Routes of transfer reported so far include prey-predator interactions (Le Clec’h et al. 2013), host-parasite interactions (Cook & Butcher 1999; Vavre et al. 1999; Ahmed et al. 2015) and sharing a common food resource (Li et al. 2017).
Host phylogeny and ecological connectivity are thought to be the two main factors determining Wolbachia host shifting. As phylogenetically closely related species are similar in many respects, including their intercellular environment and immunology (Perlman & Jaenike 2003), it is expected that a given symbiont will shift more easily between them than between distantly related species (Charleston & Robertson 2002). This assumption, referred to as the “phylogenetic distance effect” (PDE) (Longdon et al. 2011; Engelstädter & Fortuna 2019), may partly explain host shifts of Wolbachia across closely related species. In spite of limited case studies which indicated the presence of PDE in part of the host phylogeny (e.g., in fig wasps (Shoemaker et al. 2002), fungus growing ants (Frost et al. 2010), bees (Gerth et al. 2013) and Collembolans (Ma et al. 2017)), the influence of PDE on Wolbachiahost shifting is not clear. Overlapping geographic distributions of host species is another possible explanatory factor. Sharing a common habitat and consequently potential ecological interactions may lead to several direct and indirect physical contacts between a given donor and recipient host and, therefore, also increase the probability ofWolbachia host shifting. Indeed, several case studies documented host-shift events between host species that share the same habitat, e.g., in a rice field community (Kittayapong et al. 2003) and a mushroom habitat (Stahlhut et al. 2010)).
Here, we use scale insects as a model system to gain a better understanding of Wolbachia host shifting. With more than 8200 described species and 24 families, the superfamily of scale insects (Coccoidea) is globally distributed (Gullan & Cook 2007; García Morales et al. 2016). Like many other members of the suborder Sternorrhyncha, such as aphids, whiteflies and psyllids, scale insects exclusively feed on plants and some are considered serious agricultural pests (Kondo et al. 2008). Scale insects have been recorded in ecological associations with a range of arthropods species. In particular, many are usually observed in close interactions with ants through trophallaxis (where the honeydew produced by the scale insects is consumed by ants) (Hölldobler et al. 1990; Buckley & Gullan 1991; Gullan et al. 1993). Despite several similarities with other hemipterans, Sanaei et al. (2021b) found that most species are predicted to have low to intermediateWolbachia prevalence, in contrast to a u-shaped distribution (most species have a very low or very high prevalence) predicted for other groups (Hilgenboecker et al. 2008). Also, a positive correlation between Wolbachia infection in scale insects and their associate ants indirectly points to a plausible route of transfer (Sanaei et al. 2021b). These preliminary results provided a broad view of theWolbachia infection dynamic in scale insects and thus inspired us to investigate Wolbachia strain diversity and consequently host-shifting in scale insects.
Studying of Wolbachia host-shifting requires using and developing new methodologies. To overcome technical problems associated with Sanger sequencing (see Discussion), we adopted Illumina multi-target amplicon sequencing techniques to determine the Wolbachia strains in scale insects and their associate species. Using this effective methodology, we revealed the strain diversity and composition (including both single and multiple infections) in scale insects. Using phylogenetic trees of both scale insects and their Wolbachia strains, and the geographic distribution range of each scale insect species, we assessed which factors (phylogeny or geography) best explain host-shifting events. Finally, by determining Wolbachia strains in individual scale insects and directly associated individuals of other species, we identified plausible routes of horizontal transfer.