The Biodiversity of Water Mites That Prey on and Parasitize Mosquitoes

: Water mites form one of the most biodiverse groups within the aquatic arachnid class. These freshwater macroinvertebrates are predators and parasites of the equally diverse nematocerous Dipterans, such as mosquitoes, and water mites are believed to have diversiﬁed as a result of these predatory and parasitic relationships. Through these two major biotic interactions, water mites have been found to greatly impact a variety of mosquito species. Although these predatory and parasitic interactions are important in aquatic ecology, very little is known about the diversity of water mites that interact with mosquitoes. In this paper, we review and update the past literature on the predatory and parasitic mite–mosquito relationships, update past records, discuss the biogeographic range of these interactions, and add our own recent ﬁndings on this topic conducted in habitats around the Laurentian Great Lakes. The possible impact on human health, along with the importance of water mite predator–prey dynamics in aquatic food webs, motivates an increase in future research on this aquatic predator and parasite and may reveal novel ecological functions that these parasitic and predator–prey relationships mediate.


Introduction
Water mites are both abundant and ubiquitous aquatic arachnids that are found globally in freshwater habitats, except in Antarctica. Water mites have high species richness and biomass and can easily be collected in the many habitats they occupy. Reports of over 600 specimens that represent up to 13 genera can be collected in under three hours in typical freshwater habitats in the Great Lakes region by using a basic dip net to collect aquatic debris, which, when placed on a white enamel pan, allows the mites to be easily siphoned by using a pipette, as they scurry out of the debris [1]. More recently, we reported 17 genera occupying one location in this region within the Detroit River [2]. A species accumulation curve from the Palearctic shows that a plateau in the curve has not been reached, indicating that many more species remain to be discovered [3]. Previous studies report that perhaps only half of water mite species in North America have been named, constituting around 6000 water mite species with potentially 10,000 or more species found globally [3]. In other regions of sheds its skin, lasting for one to three weeks, depending on species type, water temperature, and food availability. The mosquito larval stage suffers the greatest threat of predation from aquatic species such as water mites. The pupa is a resting stage that is solely aquatic, with no feeding, and lasts from one to three days, during which the pupa metamorphoses into a flying adult that lives its life in both a terrestrial and aerial environment. The switching of adult mosquito hosts from animal to human can occur seasonally, enabling zoonotic disease transmission [16].
During the larval aquatic stage in the life history of mosquitoes, they are preyed upon by water mites. Water mites are true aquatic organisms, but many species (which are reviewed here) have an ectoparasitic larval stage that parasitizes organisms that may become airborne, such as mosquitoes. Biotic interactions, such as those of water mites and mosquitoes, contribute to functional biodiversity, which might be critical in sustaining ecosystems [17]. The impact of contemporary global biodiversity decline has prompted the United Nations (UN) to declare 2011 to 2020 as the "UN decade of biodiversity" [18]. Freshwater biodiversity is the most threatened form of biodiversity, and experts implore an increased investment in research and documentation of freshwater biodiversity [19].
Through the highly complex web of interactions among species, such as parasitism, ecosystem functioning and biodiversity can be altered [20]. One example of an ecosystem service provided by water mites includes lowering mosquito fecundity, and thus reducing mosquito prevalence [21][22][23]. These positive effects can be attributed to the presence of high diversity, because the likelihood of selection effects, facilitation from long-term coexistence, and niche complementarity are greater as diversity increases [20]. The ecosystem-level consequences from biodiversity loss are significant, being of the same magnitude as the effects on environments from other anthropogenic global-change stressors [24]. This reinforces the urgency of the conservation and restoration of biodiversity worldwide [25]. The importance of biodiversity can be demonstrated through several theories that link higher diversity to increased productivity, ecosystem stability, and resistance to invasion from exotic species [25].
Through the loss of biodiversity, we not only lose the species themselves, but we harm the direct and indirect community and ecosystem-level biotic interactions that they are embedded in, as well. Beyond just consumer diversity, the role of parasite diversity on ecosystem functions has rarely been regarded. Parasites are ubiquitous organisms that are capable of regulating host abundance and community assemblages, which in turn can impact host biodiversity and the ecosystem processes those hosts influence [20]. For example, a parasite that uses a herbivore host can reduce herbivore abundance, which can have a trophic cascade that increases plant primary productivity through reduced grazing pressure [26]. Parasites are capable of increasing or decreasing biodiversity through facilitating or removing novel traits, as well as increasing or decreasing trait diversity [20].
Water mites are globally diverse aquatic arachnids and increase the complexity of trophic networks by being both predators and parasites. This review updates our current knowledge of the diversity of water mites that interact with mosquitoes and updates past records. We summarize the literature on biogeography and discuss possible life history strategies of water mites. This work aims to advance water mite research by exploring new avenues of research revealed by preliminary data from mesocosm experiments regarding water mite predation on mosquito larvae in urban parks. Our mesocosm experiments reported here can be used to identify other water mite mosquito predators and could also be modified to study parasitism. This review also provides a platform to advance important aquatic ecological topics, such as predator-prey interactions and parasitism. The research on water mite parasitism and predation on mosquitoes remains a relevant area of investigation, given the many unknowns of the diversity of these biotic interactions and the continuing and ever-expanding threat from mosquitoes.

Literature Review
A literature review of all water mite associations with mosquitoes as predators and parasites, using the Wayne State University Web of Science© portal (Clarivate Analytics), resulted in 186 records from a total of 148,858,601 records. The search terms were "water mites parasite mosquito". A second search of "water mites predator mosquito" returned 24 records, of which none was relevant to the present review. A Google Scholar Publication search identified 41 records, comprising 2 books and 39 articles. Other articles of interest were obtained from primary authors themselves.

Field Experiments to Identify Water Mite Mosquito Predators
Mosquito-attracting mesocosms were deployed at 6 sites in Lake St. Clair Metropark (LSCMP), located in Harrison Township (42. Figure 1). The mesocosms, consisting of buckets with a volume of 5 L suspended from wooden frames (Figure 2), were set up in wet-mesic flatwoods forest and marsh wetland habitats and monitored approximately every two weeks, from April through November 2018. The buckets filled naturally with rainwater, to varying depths, and by July, mosquitoes laid eggs in the buckets, and mosquito larvae were observed through October. After mosquito larvae were detected in the mesocosms, water mites (Lebertia quinquemaculosa, Hydrachna, or Arrenurus) were added to some mesocosms, at various intervals, while noting the presence or absence of mosquito eggs, larvae, and pupae at each interval (see Figure 2). to the present review. A Google Scholar Publication search identified 41 records, comprising 2 books and 39 articles. Other articles of interest were obtained from primary authors themselves.

Field Experiments to Identify Water Mite Mosquito Predators
Mosquito-attracting mesocosms were deployed at 6 sites in Lake St. Clair Metropark (LSCMP), located in Harrison Township (42.5818° N, 82.8093° W), adjacent to Lake St. Clair, and at 6 sites in Belle Isle State Park (BI) (42.3433° N, 82.9743° W), a 400 acre urban island park in Detroit, MI, in the Detroit River, which forms the border between the United States and Canada ( Figure 1). The mesocosms, consisting of buckets with a volume of 5 liters suspended from wooden frames ( Figure  2), were set up in wet-mesic flatwoods forest and marsh wetland habitats and monitored approximately every two weeks, from April through November 2018. The buckets filled naturally with rainwater, to varying depths, and by July, mosquitoes laid eggs in the buckets, and mosquito larvae were observed through October. After mosquito larvae were detected in the mesocosms, water mites (Lebertia quinquemaculosa, Hydrachna, or Arrenurus) were added to some mesocosms, at various intervals, while noting the presence or absence of mosquito eggs, larvae, and pupae at each interval (see Figure 2).

Predation
Since the 1700s, water mites have been studied by classical taxonomists such as Linnaeus and DeGeer. DeGeer (1778) reported water mite parasitism in his renowned work "Mémoires pour servir à l'histoire des insectes" (eight volumes, 1752-1778) [27]. However, observations of water mite predation on mosquito larvae were only reported much later, at the beginning of the 20th century. Water mites have been observed as predators of mosquito larvae under both natural conditions and controlled laboratory experiments. However, in comparison to the many studies of the diversity of water mites that parasitize mosquitoes, very few studies have been reported on the diversity of water mites as predators of mosquito larvae, eggs, or pupae. So far, there is only evidence of water mite predation on mosquito larvae and eggs, while there are no reports of predation on pupae. Smith [27] summarized the known material regarding mite predation of mosquitoes in his review, but there has been no update since then.
Here we discuss these previous observations and add six additional water mite records, since Smith [27], to the list of water mites that prey upon mosquito life stages (see Table 1). Mullen [15] observed Thyas barbigera and T. stolli preying on Aedes eggs in the laboratory. In that same work, he reported Hydryphantes ruber preying on Aedes stimulans larvae in the laboratory, and Piona feeding on mosquito larvae in woodland ponds. An earlier work by Laird observed Limnesia jamurensis feeding on Culex pullus and Anopheles farauti eggs and small larvae. They observed ponds devoid of mosquito larvae but filled with water mites and thus deduced that the mites might be the predators. They also conducted feeding experiments in the laboratory. Smith [27] also reported field observations of another water mite, Piona, feeding on mosquito larvae. Smith [27] suggested that adult Arrenurus mites feed on ostracods, while larval Arrenurus parasitize mosquitoes. Rajendran and Prasad [28] added a new taxon preying on mosquito larvae, Encentridophorus similis from the Unionicolidae family. Rajendran and Prasad [28] collected water mites of this species from adult mosquitoes and fed them Aedes albopictus larvae, which they preferred over copepods and ostracods. A subsequent study Rajendran and Prasad [29] provided the sole example of adult Arrenurus feeding on mosquito larvae. In experimental studies, Arrenurus madaraszi were fed larvae from Aedes albopictus, A. hyrcanus, and A. vagus. He noted that mosquito larvae became paralyzed when water mites attached themselves to the larvae. This suggests that water mites may be injecting venom that paralyzes the larvae, certainly a potential avenue for future research.
While Smith [27] mentions the work of Hearle [30], in which red water mites were observed to feed voraciously on mosquito "wrigglers" (larvae), known elsewhere as "wiggle waggles" (Pers comm. Belize colloquial use), he did not include it in his list. Perhaps this is because Hearle [30] did not identify the mite, although he wrote extensively on some life history characteristics where he kept mites and fed them. Aedes vexans larvae were provided upon which the mites then laid eggs that hatched after the season passed. Hearle [30] deduced that, in nature, the mites would lay eggs on leaves and debris and these would remain dormant until the following spring, at which point they would hatch. This observation has been confirmed in other mite species where seasonality is important in their life history. Smith [27] suggests that the mites observed by Hearle [30] were most likely Piona, which Smith [27] has also reported in his work as being predators of mosquito larvae. Bottger [31] had reported observations of Teutonia cometes, Limnesia koenikei, and Hygrobates calliger as preying on mosquito larvae, and although these were not included in the Smith [27] review of water mite predators of mosquitoes, these observations by Bottger were cited in the review by Proctor and Pritchard [7] on the scope of prey that water mites feed on.
Despite an estimated 57 families of water mites with 428 genera currently described [3], our current review observed only nine genera preying upon mosquito larvae or eggs (see Table 2). These observations include a new record, Lebertia quinquemaculosa, from our own study, described in this review (see Table 1). In this limited dataset, the addition of Lebertia quinquemaculosa to two mesocosms with mosquito larvae reduced the number of mosquito larvae to zero, whereas mosquito larvae continued to be present where Hydrachna and Arrenurus (see Table 1 and Figure 3) had been added. Interestingly, L. quinquemaculosa added to a mesocosm with only mosquito pupae present still had live pupae remaining when next inspected. likely Piona, which Smith [27] has also reported in his work as being predators of mosquito larvae. Bottger [31] had reported observations of Teutonia cometes, Limnesia koenikei, and Hygrobates calliger as preying on mosquito larvae, and although these were not included in the Smith [27] review of water mite predators of mosquitoes, these observations by Bottger were cited in the review by Proctor and Pritchard [7] on the scope of prey that water mites feed on. Despite an estimated 57 families of water mites with 428 genera currently described [3], our current review observed only nine genera preying upon mosquito larvae or eggs (see Table 2). These observations include a new record, Lebertia quinquemaculosa, from our own study, described in this review (see Table 1). In this limited dataset, the addition of Lebertia quinquemaculosa to two mesocosms with mosquito larvae reduced the number of mosquito larvae to zero, whereas mosquito larvae continued to be present where Hydrachna and Arrenurus (see Table 1 and Figure 3) had been added. Interestingly, L. quinquemaculosa added to a mesocosm with only mosquito pupae present still had live pupae remaining when next inspected.

Parasitism
Parasitism of mosquitoes by water mites was first recorded by DeGeer in 1778 and has since been an interesting focus of research for water mite investigators and others [27,33]. Comprehensive reviews of water mite parasitism on mosquitoes by Mullen [33] and Simmons and Hutchinson [34] revealed a global biodiversity of water mites that parasitize mosquitoes. The early work by Mullen [33] reported 15 genera of water mites that parasitize mosquitoes, but he disqualified five based on what he thinks were misidentifications or other inconsistencies. Smith [27] reported 10 genera mainly based on Mullen's work but not including the genera Mullen disqualified. Smith and Oliver [35] compiled an excellent review of parasitic hosts of larval water mites and their work agrees with the water mite-mosquito associations reported here. A newer study by Simmons and Hutchinson [34] reported seven genera and two families of water mites that parasitize mosquitoes. Water mites that

Parasitism
Parasitism of mosquitoes by water mites was first recorded by DeGeer in 1778 and has since been an interesting focus of research for water mite investigators and others [27,33]. Comprehensive reviews of water mite parasitism on mosquitoes by Mullen [33] and Simmons and Hutchinson [34] revealed a global biodiversity of water mites that parasitize mosquitoes. The early work by Mullen [33] reported 15 genera of water mites that parasitize mosquitoes, but he disqualified five based on what he thinks were misidentifications or other inconsistencies. Smith [27] reported 10 genera mainly based on Mullen's work but not including the genera Mullen disqualified. Smith and Oliver [35] compiled an excellent review of parasitic hosts of larval water mites and their work agrees with the water mite-mosquito associations reported here. A newer study by Simmons and Hutchinson [34] reported seven genera and two families of water mites that parasitize mosquitoes. Water mites that parasitize mosquito adults overlap with six water mite genera that prey upon mosquito larvae and eggs (see Discussion). The mosquito hosts are similarly diverse, with 12 genera of mosquitoes having been identified by Simmons and Hutchinson [34]. In the present review, we add newer studies of water mites that parasitize mosquitoes; Table 3 lists the newer cases of water mite parasitism on mosquitoes that update the work since Simmons and Hutchinson [34].  Especially notable updates in Table 3 include observations by Atwa et al. [36], Manges et al. [37], and others. Atwa, Bilgrami and Al-Saggaf [36] reported new studies of water mites and their parasitic associations with mosquitoes, based on collections sourced in North India. Novel associations noted in this work included Culex infula associated with Arrenurus danbyensis and Parathyas barbigera [36]. Other Culex species, including C. nigropuntatus, C. fatigans, C. malayi, and C. bitritaeniorhynchus were reported associating with multiple Arrenurus genera, including A. acuminatus, A. gibberifrons, A. madaraszi, A. kenki, A. danbyensis, and Parathyas barbigera [36]. Anopheles genera also had new associations with A. culcifactes, A. quinquefaciatis, A. stephensi, A. minimus, and A. barbarostris with Arrenurus genera and Parathyas barbigera [36]. Aedes genera were also reported with new associations, including A. albopictus, A. aegypti, A. pallidostriatus, A. pipersalatus, A. novalbopictus, A. vittatus, and A. ramachandarat with Arrenurus genera. Likewise, Parathyas barbigera associated with A. albopictus, A. aegypti, A. vittatus, and A. ramachandarat [36]. Manges, Simmons, and Hutchinson [37] reported several new mosquito mite associations in North America, with mosquitoes that are considered invasive. Aedes genera, including A. albopictus and A. japonicus, were associated with Parathyas and Culex restuands, and C. pipiens were also associated with Parathyas [37]. Other interesting cases of parasitism, such as Arrenurus seen parasitizing a Culex pipiens pupae and Unionicola seen parasitizing a Cladoceran (Bosmina tubicen), are notable observations [38,39].
Our updated lists also include previously excluded data that should be considered, such as Lebertia tauinsignata reported by Marshall [40] but disqualified by Mullen [33]. We urge this reconsideration as we think Lebertia tauinsignata could possibly parasitize mosquitoes, as our own research shows Lebertia feeding on mosquito larvae and parasitizing chironomids, which are related to mosquitoes [10]. Our critical assessment of the work done by Marshall [40] did not find any reason to disqualify the observation. Another study that was disqualified by Mullen was the study by Mira [41] that identified Unionicola mites parasitizing Anopheles mosquitoes in what was Italian East Africa. Newer studies in the Arabian Peninsula adjacent to Ethiopia have identified Unionicola mites parasitizing mosquito pupae [38]. Other associations reported by Mullen [33] might need further assessment to determine why they were disqualified and if they should be considered again, given new research insights.
The Arrenurus genus commanded 61.67% of the parasitic associations, with 111 species of mosquitoes being parasitized (see Figure 4). The Parathyas genus was second highest, with 25.55% parasitic associations and 46 species of mosquitoes being parasitized. Further discussions on these two groups will be presented later, but it must be noted that the Arrenurus genera included several species of Arrenurus, but Parathyas was represented primarily by one species: Parathyas barbigera. In sum, the Arrenuridae water mite family (especially species of the genus Arrenurus) parasitized, by far, the greatest number of genera (11) and species (111) of mosquitoes (Figures 4 and 5). The water mite species and the number of mosquito species they parasitize are summarized in Figure 5. Within the Arrenurus genus, 27 species were found to parasitize mosquito larvae, with the most frequently observed species being A. angustilimbatus, A. kenki, and A. madaraszi. Since Arrenurus has the highest species richness of all water mite genera, and their larvae are generally difficult to identify at the species level, the diversity of Arrenurus species parasitizing mosquitoes may be even greater. Worldwide, 950 Arrenurus species have been documented [42], with 400 in the Nearctic region to date [6]. Newly assigned genera of water mites that parasitize mosquitoes included in the present review are Lebertia and Unionicola. These associations are based on our literature review, unpublished and published observations, and reassessment of previously rejected literature observations. For brevity, we did not include associations where the water mites could only be identified to family or subfamily taxonomic level, which included Euthyasinae and Thyadinae [34]. Diversity 2020, 12, x FOR PEER REVIEW 10 of 24

Global Perspectives and Considerations of the Biodiversity of Water Mite Predation and Parasitism of Mosquitoes
Despite Hydrachnidia (water mites) being the most biodiverse taxonomic group of the Arachnids, our analysis suggests that only about 3.5% of the total known water mite genera preys on and parasitize mosquitoes (see Table 4). However, some of the genera that have been shown to

Global Perspectives and Considerations of the Biodiversity of Water Mite Predation and Parasitism of Mosquitoes
Despite Hydrachnidia (water mites) being the most biodiverse taxonomic group of the Arachnids, our analysis suggests that only about 3.5% of the total known water mite genera preys on and parasitize mosquitoes (see Table 4). However, some of the genera that have been shown to

Global Perspectives and Considerations of the Biodiversity of Water Mite Predation and Parasitism of Mosquitoes
Despite Hydrachnidia (water mites) being the most biodiverse taxonomic group of the Arachnids, our analysis suggests that only about 3.5% of the total known water mite genera preys on and parasitize mosquitoes (see Table 4). However, some of the genera that have been shown to parasitize and prey on mosquitoes are believed to be some of the most specious, with Piona having potentially more than 100 species and Arrenurus up to 400 species [6]. Table 4. Overlap of water mite parasites and predators of mosquito adult and larvae, respectively.
The aforementioned observations were based on both field and laboratory studies, comprising observations from the United States, Canada, Germany, Sweden, France, Denmark, New Zealand, Australia, Panama, Brazil, China, Japan, Uganda, Gambia, Madagascar, Nigeria, Angola, Indonesia, Malaysia, India, Sri Lanka, and Saudi Arabia (see Figure 6). While this demonstrates a broad global distribution of diverse water mite parasitism on mosquitoes, most reports originated from the United States and India, with over 100 records each.
Additionally, many biogeographical regions are left to be studied for water mite-mosquito associations, such as the Afrotropical and Neotropical regions, which are consequently known for mosquitoes and the diseases they cause (see Figure 6). parasitize and prey on mosquitoes are believed to be some of the most specious, with Piona having potentially more than 100 species and Arrenurus up to 400 species [6]. Table 4. Overlap of water mite parasites and predators of mosquito adult and larvae, respectively. Mullen, 1975 [33].

Mite Genus Mosquito Parasite Mosquito Predator
The aforementioned observations were based on both field and laboratory studies, comprising observations from the United States, Canada, Germany, Sweden, France, Denmark, New Zealand, Australia, Panama, Brazil, China, Japan, Uganda, Gambia, Madagascar, Nigeria, Angola, Indonesia, Malaysia, India, Sri Lanka, and Saudi Arabia (see Figure 6). While this demonstrates a broad global distribution of diverse water mite parasitism on mosquitoes, most reports originated from the United States and India, with over 100 records each.
Additionally, many biogeographical regions are left to be studied for water mite-mosquito associations, such as the Afrotropical and Neotropical regions, which are consequently known for mosquitoes and the diseases they cause (see Figure 6).

Discussion
Water mites are known to be predators of a variety of aquatic invertebrates, including copepods, cladocerans, ostracods, and Dipteran larvae, and the larvae of water mites are also known to parasitize a diverse selection of invertebrates, such as dragonflies, mayflies, mosquitoes, and water

Discussion
Water mites are known to be predators of a variety of aquatic invertebrates, including copepods, cladocerans, ostracods, and Dipteran larvae, and the larvae of water mites are also known to parasitize a diverse selection of invertebrates, such as dragonflies, mayflies, mosquitoes, and water beetles [7,35,44]. Given this diversity of biotic interactions they have as both predators and parasites, we consider their contribution to aquatic ecosystems to be very significant. The biodiversity of biotic interactions and the effect of parasitism on biodiversity are areas of research that are gaining renewed interest [17,20]. The specific aims of our work are to (i) update the known predatory and parasitic associations of water mites and mosquitoes (adults and larvae); (ii) update past records of water mite-mosquito associations; (iii) identify specific water mite genera and the biodiversity of their biotic interactions; and (iv) suggest future directions for studies with water mites, to increase our understanding of predatory biotic interactions in nature.

Identification of Major Water Mite Mosquito Parasites
Water mites are considered hyperparasites of mosquitoes, as they are a parasite that parasitizes another parasite, but despite these ecologically relevant biotic interactions, this area of research has been understudied [37,45]. Some mite species have been documented to parasitize multiple mosquito species throughout several genera, while other mite species have a very specific parasite-host association, with only a few mite species parasitizing a specific species of mosquito. Our review identified Arrenurus kenki as parasitizing 24 mosquito species of the genera Aedes, Anopheles, and Culex (see Table A1). Another water mite species that parasitizes a large number of mosquito species is Parathyas barbigera, which parasitizes 42 identified mosquito species in the genera Aedes, Anopheles, Culex, possibly Coquillettidia [46], Psorophora, Toxorhynchitis, and Uranotaenia (see Table A2).
The findings in our review suggest that the Arrenurus and Parathyas genus of water mites and their biotic interactions with mosquitos (as parasites of adult mosquitoes) might be an important area for future studies. Arrenurus kenki was listed as being "facultative tolerant" to organic wastes in a study that had a 0-5 range [47]. Parathyas barbigera (listed as Thyas in this reference) listed it as being facultative tolerant suggesting that both these species require an aquatic habitat that does not have excessive pollution [47]. This strengthens the idea that preserving biodiversity is important, especially in freshwater ecosystems that may contain these types of water mites which have a prominent role as mosquito parasites. It also emphasizes the loss of potential ecosystem services when the biodiversity of these biotic interactions is lost due to habitat loss or degradation. Work like this strengthens conservation efforts to improve freshwater habitats, since this is the habitat where biodiversity is disappearing at a faster rate than terrestrial systems [19].

The Potential Impact of Water Mite Life History Strategies on Their Biotic Interactions with Mosquitoes
Jalil and Mitchell [48] postulated that there are two types of water mites: the "thyasid-type", which belongs to the Thyas (=Parathyas) genus, and the "pionid-type", which includes those from the Arrenurus genus. In our review, the genera Parathyas and Arrenurus are those with the most significant parasitic associations with mosquito adult flies. The few studies focused on this topic have documented the possibility of water mite parasitism limiting the rate of survival and reproduction of mosquito hosts in natural environments to varying degrees [21,[48][49][50]. The differing life history and behavior of thyasid-type and pionid-type mite larvae has previously been argued as playing a role in the attachment site and rate on mosquito hosts, and thus the intensity (as defined by the number of parasites on a host [51]) and severity of effects on mosquito survival and reproduction [48,49]. Thyasid-type mite larvae are believed to be closer relatives of terrestrial mites, from whom they evolved, than pionid-type larvae, because of their generalized, semi-aquatic life history [52]. Thyasid-type larvae are able to break through the water surface film immediately after hatching and "walk" on the water surface, having left the water altogether [48]. The thyasid-type larvae can only attach to adult mosquito hosts that return to the water surface, giving them only a few minutes of attack time to parasitize an adult mosquito host [49]. This may result in sexual discrimination by the thyasid-type mite larvae on female mosquitoes that exhibit a higher likelihood of returning to the water surface for oviposition of eggs than male mosquitoes [49]. In contrast, the pionid-type larvae are fully aquatic specialized swimmers that cannot leave the water until after they attach to the host. Therefore, pionid-type larvae can only seek a host during the mosquito's aquatic pupa stage, in which the mite larvae rest until mosquito ecdysis. After ecdysis of the mosquito pupa, the water mite remains on the adult mosquito, on which it initiates parasitism of the adult. This life history strategy may relate to the success in high intensity of mite parasitism with pionid-type larvae, specifically with Arrenurus, where Arrenurus mite load is commonly seen to be 30 or more mites per host [46,49]. Our review also identified Arrenurus as being the genus with the most species of Arrenurus parasitizing a wide diversity of mosquito hosts (see Figure 4 and Table A1).
Parathyas barbigera and P. stolli present an interesting case, as they were the second highest water mites having parasitic interactions with adult mosquitoes, with 42 different species of mosquitoes being parasitized by P. barbigera (see Figure 5 and Table A2). It was also found to prey on mosquito larvae (see Table 2). It was one of the few mites that had overlap with its larvae being parasites on mosquito adults and its adult form preying on mosquito larvae. This comparison can be seen in Table 4. P. barbigera and P. stolli belong to those water mites classified by Jalil and Mitchell [48] as "thyasid-type" mites, but despite this, they are very successful as predators and parasites of mosquitoes.

Water Mite Parasitism Reduces Mosquito Fecundity and Survivorship
The impact of high mite loads is evident from the linear relationship of mite-induced mortality and decreased fecundity on mosquito hosts, with the slope relating to the ratio of mite weight to host weight and the mite load on the host [22,50]. The few laboratory and natural experiments that have focused on the impact of high mite loads have commonly used the mite genus Arrenurus with the mosquito genus Anopheles, where reduced survivorship and reproduction of mosquitoes by Arrenurus species have been documented [21,23,50]. An experimental study by Lanciani and Boyt found that unparasitized female Anopheles crucians had a survival time of 23.32 days, while heavily parasitized females with Arrenurus pseudotenuicollis (around 17-32 mites) had a survival time of 6.25 days [21]. They also found that the number of eggs produced by A. pseudotenuicollis significantly decreased as Arrenurus mite load increased for both field-engorged mosquitoes and laboratory-fed mosquitoes, regardless of mosquito blood meal size [21]. Another experiment by Smith and McIver discovered that, when not accounting for blood meal size of Coquillettidia perturbans, the parasitism of Arrenurus danbyensis greater than five mites decreased the egg production of C. perturbans by 3.5 eggs per additional mite [23].
In a natural experiment by Lanciani and Boyt, the female Anopheles crucians that are unengorged with their first blood meal were found to have the highest proportion of pionid-type parasites compared to engorged female mosquitoes [21]. Another laboratory experiment by Lanciani discovered the sexual preference of Arrenurus novimarshallae mite larvae toward female Anopheles crucians pupae hosts compared to male pupae of that species, even when females were reared to smaller sizes with reduced food levels [53]. Female mosquitoes require sufficient energy to conduct flights for their required first blood meal to survive, but if they are heavily parasitized by pionid-type larvae, then they are most likely unable to attain this crucial blood meal, which can severely reduce mosquito densities. Therefore, the effects and rates of parasitism by larval water mites and predation by deutonymph and adult water mites have been found to have a severe effect on population sizes of host and prey species. This, therefore, emphasizes the importance of future studies on loss of biodiversity of water mite parasitism and predation and the effects on the population size of species that they can potentially regulate. Since mosquito larvae have been found to be a possible host and prey for some species of water mites, it is crucial to determine the specific species of water mites that parasitize and prey on mosquito larvae.

Potential Water Mite Adaptations: Speciation and Niche Partitioning
We have also reviewed the water mites that have been shown to prey on mosquito larvae. Although the Arrenurus genus is substantially found to dominate the type of water mites that were found parasitizing mosquito adults (see Figures 4 and 5), we could only find one instance in which an Arrenurus adult water mite was found to prey on mosquito larvae in laboratory experiments (see Table 2) [29]. This demonstrates the hidden complexity within the biotic interactions since Arrenurus madaraszi was the only Arrenurus adult water mite found preying on mosquito larvae, but is the primary genus reported as parasitic (see Table 2 and Figure 5) [29].
Of the 17 genera listed, only six (Parathyas, Piona, Hydryphantes, Arrenurus, Lebertia, and Limnesia) were shown to both prey on and parasitize mosquito larvae and adults (see Table 4). Water mites are known to prey on and parasitize a wide variety of invertebrates. However, not all water mites are parasites. Up to 29 species are thought to not have parasitic larval stages, and studies that compare both parasitic and non-parasitic are needed to clarify possible adaptation benefits of one over the other [54]. A study comparing two species of Arrenurus, A. angustilimbatus (which is a parasite of mosquitoes and mentioned in this review) and A. rufopyriformis (which does not have a parasitic stage in its life cycle), concluded that A. angustilimbatus could be considered to be "ecologically successful" due to its higher heterozygosity and wide geographic range [55]. Studies on other species of water mites have also suggested species' separation as a consequence of parasitism [56]. Additionally, parasitic water mites have been observed to partition on a single host [57]. Up to nine water mite species were observed partitioning Chironomid hosts, with some species demonstrating preferred specificity to the thorax, while others to the abdomen [57]. This type of "niche" partitioning along with the phoresy associated with the parasitism of adult flies could contribute to the highly successful biodiversity and prevalence of water mites. However, due to poor taxonomy of water mites, particularly at the parasitic stage (larval stage), there is still much work needed to be done to fully appreciate the ecological and evolutionary contributions that these life history traits provide. However, with the ability for more accurate genus and species identification of water mites and mosquitoes via genetic analysis, and increased research on the abovementioned life history traits, an expanded understanding of these relationships will be made possible.

Biogeography of Water Mite-Mosquito Biotic Interactions
Because of the high species richness of both mites and mosquitoes and the extensive diversity of species-specific interactions between taxa, further investigation of the abundance of host exploitation and the effects of mite parasitism on mosquitoes seem likely to reveal a range of functional interactions. Even at the species level, these characteristics (attachment rate, mite load, and effects) of mite parasitism on mosquitoes have been previously found to vary depending on the species of both organisms involved in the parasitic interaction. Additionally, the biogeography related to these biotic interactions is also in need of clarification, since our overview of the biogeography of the groups discussed in this review could only be described at the family or subfamily level (see Table 5). Some of the genera have broad distribution, which may imply widespread impact through the diverse biotic interactions with mosquitoes. The cosmopolitan genus Arrenurus has published records of parasitism from widely distant regions, such as Japan and Canada, covering the Nearctic, Neotropical, Palearctic, Oriental, Australasian, and Afrotropical (see Table A3). More work is needed to understand the biodiversity, biogeography, and specificity of these biotic interactions, to document the extensive parasite-host association combinations that are present at a global scale.

Studies on Water Mite Predation of Mosquito Eggs, Larvae, and Pupae Are Needed
Similarly, with respect to predatory impacts of water mites on mosquitoes, further investigations are needed to determine ecological significance and, given the high health impact of mosquito-borne diseases, to determine if mite predation on mosquito larvae can be exploited to reduce these disease burdens. Commenting on the voraciousness of Piona spp. on mosquito larvae, Smith [27] noted that although Piona spp. are able to consume a large number of mosquito larvae, "quantitative studies on the ecology and feeding behavior of these mites are lacking", a statement that is still true almost 40 years later, as we write this review. A similar comment is made by Esteva et al. [58], who created a mathematical model of the roles of parasitism and predation in controlling the population dynamics of water mites and mosquitoes. The modelers observed in their model that predation had a more significant effect than parasitism in controlling the dynamics of mosquito and water mite populations. Indeed, in their model, populations of adult mosquitoes plummet to near zero as the water mite predation rate increases; the range of effective population-reducing predation occurs at a level of <0.9 mosquito larvae consumed per day per mite, which is a modest level compared to laboratory observed rates of six to eight mosquito larvae per mite by Limnesia jamurensis [32]. However, the modelers noted that "systematic studies about the extent of the impact of water mites on mosquito populations that could be used as a basis for a control program are scarce and fragmentary" [58].
Going forward, much new data needs to be collected on the intensity of water mite predation on mosquito larvae and about their impacts on mosquito populations. Our studies, reported elsewhere, applied high throughput sequencing to determine if mosquito DNA can be detected from the molecular gut contents of water mites freshly collected from the field [10]. The molecular gut contents from Lebertia quinquemaculosa, a second species of Lebertia with a novel COI barcode (tentatively named Lebertia davidcooki), and unidentified species of Arrenurus and Limnesia, was amplified with COI primers designed to amplify insects but not arachnids [10]. While DNA of many of the expected prey was present in these specimens (e.g., most sequences in Lebertia were from a multitude of chironomid species; Arrenurus had DNA from the ostracod Podocopida), sequences from mosquitoes were also present [10]. Culex pipiens sequences were observed in 20% of L. quinquemaculosa and 7% of L. davidcooki specimens, and neither of the other species [10]. We plan to apply these techniques to other water mite species, including Piona, which has figured prominently in this review of water mite-mosquito predation, to determine which, if any, species of water mites might utilize mosquito larvae as a predominant part of their diets. In addition, we have initiated mesocosm studies, reported here, to study water mite impacts on naturally recruited mosquito larvae (see Section 3.1).
These studies looked at naturally recruited mosquito larvae in aquatic mesocosms (see Figure 2) to which we have experimentally added water mites at various intervals. Our gut DNA studies in Vasquez [10] identified a potentially novel water mite mosquito predator, Lebertia quinquemaculosa, which we added to the mesocosms reported here. Mimicking "natural" artifacts, such as ponds and streams, vernal ponds, puddles, and other damp areas such as plant phytolemata [59], and also man-made water-retaining structures, such as cisterns, rain gutters, and buckets, we deployed bucket mesocosms in urban parks (see Figures 1 and 2), all of which provide an extensive range of mosquito-breeding habitats. Examples of such features in an urban area have been documented in Detroit [10] and are generally found in urban areas elsewhere [60,61]. We showed, in our results, that the addition of L. quinquemaculosa reduced mosquito larvae population in the mesocosms (see Table 1). We have thus added L quinquemaculosa to the list of water mite mosquito predators (see Table 2).
To our knowledge, these observations of water mite and mosquito larvae in bucket mesocosms constitute the first field test of its kind that investigated water mite impacts on mosquito populations in a naturalistic environment. On a preliminary basis, at least, L. quinquemaculosa seems to have a greater effect on mosquito larvae recruitment or survival than do two other genera of mites (Arrenurus and Hydrachna) which we added to our mesocosms (see Table 1 and Figure 3), a result supported by our molecular diet research on Lebertia and consistent with previous diet preference research on the other species [6,7,10]. As noted above, a systematic investigation of diverse water mite species, including those suggested to have predatory associations with mosquito larvae or eggs, would be warranted, especially to provide data for mathematical models of water mite-mosquito interactions and ultimately to determine whether water mite predation could be exploited to control mosquitoes. These mesocosm studies could also be enhanced with cameras and other observational methods that would clarify the mechanisms by which water mites impacted the recruitment, growth, and/or survival of the mosquito larvae in mesocosms.

Final Considerations
This knowledge of the functional biodiversity of water mites that feed on and parasitize mosquitoes could be of great importance in understanding predator-prey dynamics [20] and developing new methods for controlling mosquitoes. Several different types of diseases, such as West Nile virus, eastern equine encephalitis, dengue, malaria, Zika, yellow fever, and chikungunya, are caused by mosquitoes. The human morbidity due to mosquitoes is estimated at 725,000 worldwide, making it potentially the deadliest animal on earth (https://www.cdc.gov/globalhealth/stories/world-deadliestanimal.html). While our current research is in temperate regions where eastern equine encephalitis and West Nile virus are especially of concern, most mosquito-borne diseases of human pathological importance are primarily found in tropical regions, where, ironically, water mite biodiversity is least understood [4]. DNA barcoding could potentially assist in improving the knowledge of water mite diversity, since water mite adult and larvae DNA barcodes could be matched, thereby greatly facilitating research on mite-mosquito interactions [34].