Phytoseiidae of La Réunion Island (Acari: Mesostigmata): three new species and two males described, new synonymies, and new records

The Indian Ocean includes several Islands and Archipelagos and a survey of the fauna of Phytoseiidae was realized recently in these poorly known territories, despite La Réunion Island is located in an area being a hotspot of biodiversity. The Phytoseiidae fauna of La Réunion Island has been poorly investigated so far with only 33-recorded species including 24 Amblyseiinae, 5 Phytoseiinae and 4 Typhlodrominae, among which 8 species that had been described as new. All data concerning La Réunion have been published in nine papers until now. New results of surveys done recently (2015-2018) are presented in this 10th paper and add 19 newly recorded species among which 3 are new to Science and 21 already known species but with additional data. Among these 19 species, at least nine are already well known as biological control agents (BCA). Demonstration of the natural occurrence of already known efficient BCA in these territories is consequently of great agricultural and commercial interests, in addition of the contribution to fundamental knowledge of biodiversity of these poorly known territories.


Introduction
Several species in the family Phytoseiidae are important natural enemies of phytophagous mite and small insects in natural habitats, outdoor and protected crops all around the world (McMurtry and Croft 1997;McMurtry et al. 2013). However, despite the huge numbers of faunistic surveys carried out for more than 60 years, the fauna of some countries and particular ecosystems remain little explored. Consequently, it is important to survey phytoseiid faunas in poorly known areas in order to document the biodiversity of these areas, especially in biodiversity hotspots, as well as to discover new potential biological control agents (BCA). This is especially important given the context of new international and state regulations concerning import-export of BCA (Kreiter et al. 2020a, b).
The family Phytoseiidae is widespread all over the world and consists of 2,521 valid species dispatched in three sub-families and 94 genera (Demite et al. 2019).
Several terrestrial areas of the Indian Ocean constitute one of the world's hotspots of biodiversity. Myers (1988) defined the hotspot of biodiversity concept in order to identify the areas of biodiversity in most urgent need of conservation/protection. These world hotspots are characterized by high levels of endemism and have lost at least 70 % of their original natural vegetation (Myers et al. 2000). The characterization of the phytoseiid mite diversity in these areas is thus contributing to this general topic of conservation. Located in the Indian Ocean at 700-800 km from the Eastern coast of Madagascar, La Réunion is the main island of the Mascareignes Archipelago (with the two other main islands being Mauritius and Rodrigues). Ueckermann and Loots (1985) published the first paper concerning Phytoseiidae of La Réunion Island 35 years ago. It concerns a description of a new species to Science, Phytoscutus reunionensis (Ueckermann and Loots) found on Prunus persica (L.) in La Plaine des Cafres in 1983 by Dr Serge Quilici (Ueckermann and Loots 1985). In a study of the Tetranychidae and their predators, Guttierrez and Etienne (1986) mentioned two species of Phytoseiidae for La Réunion, Euseius ovaloides (Blommers) and Phytoseiulus persimilis Athias-Henriot, the last probably escaped in the neighbouring areas of greenhouses after releases. Quilici et al. (1988) mentioned then two species on litchi (Litchi chinensis Sonn.) in Bassin-Martin CIRAD research station, Amblyseius largoensis (Muma) and E. ovaloides, bringing to four the number of known species. In a further study focussing on mites of various crops of the Island, Quilici et al. (1997) mentioned six additional species and Quilici et al. (2000) added 14 species recorded for the first time in La Réunion. Just after, Kreiter et al. (2002) described seven new species from the Island, bringing to eight the number of new species described and to 31 the total number of species recorded from this island. Only one was collected in another place after its description, Typhlodromus (Anthoseius) moraesi Kreiter and Ueckermann, very far from La Réunion, in Guadeloupe and then also in Martinique, both in the Caribbean (Kreiter et al. 2013(Kreiter et al. , 2018c. Moraes et al. (2012) added another species for La Réunion, Neoseiulus recifensis Gondim Jr. and Moraes, originally described from Brasil. Surprisingly, this species was found in samples from a survey for selection of a BCA against Raoiella indica Hirst in Brazil, along with two other common species, A. largoensis and T. ( A.) moraesi. Finally, Kreiter et al. (2016a, b) mentioned the unexpected occurrence of Amblyseius swirskii Athias-Henriot in La Réunion, never collected and mentioned in previous studies, collected suddenly in high population in 2015 and 2016, bringing to 33 the number of the known species present in La Réunion Island.
We report in this 10 th paper on La Réunion phytoseiids the results of additional surveys conducted from 2015 to 2018.

Material and methods
The survey took place in La Réunion Island in 2015, 2016, 2017 and 2018. Plant-inhabiting mites were collected from various cultivated and wild plants in various locations of La Réunion. Mites were directly collected on leaves with a fine brush or by using the leaf "dipping-shakingwashing-filtering (dswf)" method of Boller (1984) or by beating the plants (mainly shrubs or trees) and collecting the mites in a black plastic rectangular saucer 45 x 30 cm (Ref. STR 45, BHR, 71370 Saint-Germain-du-Plain, France). The method selected was dependent on the plant investigated: large leaves of shrubs and trees were sampled using the direct collection method or with dswf; small leaves of shrubs and trees with the dswf or by beating and herbaceous plants with dswf.
An experiment was conducted at the Bassin-Plat CIRAD Research Station, Saint-Pierre, in La Réunion Island (altitude above sea level = aasl: 153 m, 55°29'18" E; 21°19'25" S). The experimental site was a 0.3 ha citrus orchard (Citrus sinensis x C. reticulata cv. Tangor grafted on Citrange Carrizo) with 149 meters spaced trees planted in March 2012, after a 2-year spontaneous fallow. Tree rows were planted six meters apart. The wet season spans from November to April and the dry season from June to October. The local average annual precipitation for the period 2014 to 2016 was 1025 mm. From March 2012 to February 2014 (before the experiment started), weeds around the tree base and within 50-cm radii around them were controlled with the herbicide glyphosate (360 g.L-1 and 4 L.ha-1). Weeds in the 5-m-wide area between rows were controlled with a hammer mill.
The experiment began in March 2014. We used a complete bloc design with six replicates to test the effects of different weed management methods on the composition of the ground cover plant community. Four weed management treatments were compared: tillage (T), mowing (M), hammer mill (HM), and herbicide (H). These four treatments were distributed haphazardly within each replicate (four plots of 13 m x 5 m). Mowing and herbicide-spraying treatments were carried out in the same inter-rows so that they would not be disturbed by the tractors used for tillage or hammer mill. Hammer mill is the most commonly used engine for weed management in citrus orchards on the Réunion Island. Mowing was done with an adapted hedge-trimmer, which cuts up weeds at 10 cm above ground. A hammer mill [SML 155 SEPPI®,Caldaro (Bolzano), Italy] was used to crush weeds at the soil level. The herbicide treatment using glyphosate (360 g.L-1 at 4 L.ha-1) eliminated all weeds. A disk harrow (Grégoire and Besson®, Montigné-Montfaucon, France) was implemented once or twice in order to destroy the maximum of weeds. Weed management activities were activated when the ground cover was estimated by a farmer as being too high (70 to 80 cm height for at least one treatment). The timing for treatment was the same for all four treatments, and their timing was spaced out by a period of 93 + 27 days (mean ± SD) depending on the season.
Plots with different weed management (weed communities are variable in space and time and it is impossible to give a precise list of weed species for each date of sampling) were sampled in order to collect arthropods, including phytoseiid mites. Mites collected were then transferred with a brush into small plastic vials containing 70 % ethanol. Mites were then all mounted on slides using Hoyer's medium and identified using a phase or interferential contrast microscope (DMLB, Leica Microsystèmes SAS, Nanterre, France). Characters of specimens were measured using a graduate eyepiece (Leica, see above). We used McMurtry's (1994, 2007) concepts of the taxonomy of Phytoseiidae and the world catalogue database of Demite et al. (2019) for faunistical and biogeographical aspects. Only females were measured unless males were available. Immature will be measured and described in another paper.
In the (re)description of species, the setal nomenclature system adopted was that of Lindquist and Evans (1965) and Lindquist (1994), as adapted by Rowell et al. (1978) for the dorsum and by Chant and Yoshida-Shaul (1991) for the venter. The idiosomal setal pattern follows Chant and Yoshida-Shaul (1992). The notation for gland pores (solenostomes) or lyrifissures (poroids) is based on Athias-Henriot (1975).
Measurements of the main morphological characters were made as follows: dorsal shield length from the anterior to posterior shield margins along the midline; width between lateral margins at the level of setae s4; length of genital shield from the anterior margin of hyaline surface to the posterior margin of the shield; width of genital shield as the distance at the level of setae st5 and between posterior corners of the shield; ventrianal shield length as the distance between anterior and posterior margins; ventrianal shield width between insertions of ZV2 and at level of anus, of paranal setae between margins of the shield; cheliceral movable digit length was measured from basal articulation to tip of the digit; the fixed digit from the dorsal lyrifissure to the tip. Numbers of teeth on the fixed and movable cheliceral digits do not include the respective apical hook. Setae not referred to in the Results section should be considered as absent.
All measurements are given in micrometers (µm) and presented as the mean in bold followed by the range in parenthesis. New measurements for holotypes and paratypes of already described species are presented in tables in bold and underlined. Specimens of all species are deposited in the mite collections of Montpellier SupAgro conserved in UMR CBGP INRA/IRD/CIRAD/SupAgro. Specimens collected in fields in La Réunion within this survey were all identified. Very few single males collected alone within this study were not taken into account.
The following type or additional material have been borrowed and studied: • The holotype of Neoseiulus houstoni (Schicha), from the reference collection of the Biosecurity Collections (NSW [= New South Wales] Department of Primary Industries), Orange NSW, Australia; • The holotype, one paratype and additional material of Neoseiulus recifensis Gondim Jr and Moraes, housed in the mites reference collection of the Department Entomology and Acarology, Escuela Superior de Agricultura Luiz des Queiroz (ESALQ), University of Sao Paulo (USP), Piracicaba, Brasil.
We have also examined type specimens of Amblyseius longipilus Kreiter and Ueckermann (the holotype and 10 paratype females), eckermannseius nesiotus Ueckermann and Kreiter (one paratype female), Phytoseius haroldi Ueckermann and Kreiter (three paratype females), and Neoseiulus barreti Kreiter (one paratype female) of the mite collections of Montpellier SupAgro conserved in UMR CBGP, in order to complete descriptions and compared to specimens collected during this study.

Results and discussion
A total of 44 species were found from the beginning of 2015 to the end of 2018 in our surveys.
• Four species were already well-known in the literature, very common in La Réunion and already recorded; A. largoensis was commonly found in the near past in different localities  and several hundred specimens of this species as well as of the next species (see below), A. swirskii, were collected; A. swirskii (see Kreiter et al. 2016a) of which ten females and five males were measured.; E. ovaloides (mentioned in Quilici et al. 1997Quilici et al. , 2000, not measured in the present paper but this species has very characteristic features and is probably the most common species in La Réunion Island. Five hundred specimens were collected in the present study and everywhere in the Island; T. (Anth.) moraesi (see Kreiter et al. 2002), for which only eight females and one male collected in La Réunion were measured. This species is very characteristic and very common on weeds and low plants where several hundred specimens were collected during this study, mainly in experimental plots of Bassin-Plant CIRAD Research Station. Several specimens of this species was previously measured in the Caribbean islands (Kreiter et al. 2013(Kreiter et al. , 2018c and measurements of specimens of La Réunion agree well with measurements of these Caribbean specimens. These four species are very common and widespread in the island. Measurements of individuals of these four species are very largely overlap with those of original descriptions and of measurements published in other studies.
• For one species, A. longipilus, 14 females have been measured in the original description of Kreiter et al. (2002) which is considered enough for a good estimate of the specific variability (Tixier 2012). Some character measures were however lacking in this original description and we provide here additional/complementary data on the species.
• For one species already known from the Island, N. recifensis, we were suspecting a synonymy with two other similar species of Neoseiulus (N. barreti and N. houstoni). These three species are compared thereafter and the male of one of this species is described for the first time.
This species belongs to the barkeri species group of the genus Neoseiulus, as the spermathecal atrium is large and forked at junction with major duct. It belongs to the barkeri species subgroup as the calyx is not markedly constricted at junction with the atrium, the atrium is deeply forked at the junction with major duct without vacuolated area, and the major duct, atrium and calyx are of approximately the same width (Chant and McMurtry 2003a).
This species has been mentioned by Quilici et al., (2000) in La Réunion, with locations listed but without morphological measurements presented. Measurements of specimens collected during this study are provided in the table 2.
Comparisons with N. barkeri measurements of female and male (Table 2) specimens of various origins in Beaulieu and Beard (2018) shows shorter dimensions of all characters of La Réunion specimens (all ranges of La Réunion specimens are in the lower parts of the ranges mentioned by these authors). These authors already mention in their paper the shorter dimensions of dorsal setae of African female and male specimens (lower part of observed ranges) compared to their own measurements (Beaulieu and Beard 2018).
This species was only known from Australia for a long time (Demite et al. 2019). Quilici et al. (1997Quilici et al. ( , 2000 collected this species on Hibiscus sp., associated with populations of the eriophyid Aceria hibisci (Nalepa), which is common in La Réunion. However, the biology of this predator remains unknown.
Remarks: measurements of characters of La Réunion Island female specimens largely overlap with those of female specimens from Australia (Table 3). Measurements of La Réunion specimens are in general just a few percent's shorter, except for some setae or character dimensions which are quite shorter in La Réunion specimens, for example macrosetae of leg IV and size of ventral shields.   Tixier et al. 2014).
Like the previous species, N. californicus belongs also to the cucumeris species group of Neoseiulus (Chant and McMurtry 2003a).
This widespread species (Moraes et al. 2004;Demite et al. 2019) is considered by McMurtry and Croft (1997) to be a specialized predator, Type 2. Nevertheless, it has characteristics of both specialist and generalist predatory mites (Castagnoli and Simoni 2003). It prefers to feed on spider mites (Gomez et al. 2009), but can also consume other mite species like tarsonemid mites [Phytonemus pallidus (Banks)] (Easterbrook et al. 2001), small insects such as thrips (Rodriguez-Reina et al. 1992) and even pollen when the primary prey is unavailable (Rhodes and Liburd 2006). It can migrate from grasses to fruit trees or grapevines and vice versa (Auger et al. 1999). It is a specialist predator of T. urticae on annual plants and woody species, and of Panonychus ulmi (Koch) and various Tetranychus spp. (and perhaps eriophyid mites) on trees and less frequently on grapevines (Auger et al. 1999). N. californicus is well-known as a BCA sold in many countries around the world for the management of spider mites in greenhouses but also in outdoor crops such as fruit crops in Europe. This is the first mention of that species for La Réunion Island.

Neoseiulus houstoni Schicha
These three species belong also to the cucumeris species group of Neoseiulus like previous species. However, whereas Chant and McMurtry (2003a) classified N. recifensis in the cucumeris species subgroup, N. houstoni was placed in the paraki species subgroup (also in the cucumeris species group), despite the two species having identical spermathecae (Chant and McMurtry 2003a). Neoseiulus barreti was described later than 2003, is not mentioned in Chant and McMurtry (2007) but the spermatheca is also identical to that of the two former species.
Neoseiulus houstoni was the first species collected and described in 1987 on Vigna unguiculata (L.) Walp. in Queensland, Australia (Schicha 1987 Two females of N. barreti were collected, only one time, in Brazil in 2004 on Solanum paniculatum L. in Itapaje, Ceara, Brazil and described later (Furtado et al. 2005).

Not applicable
Remarks: character measurements of the 24 females collected in La Réunion (Table 5) agree very well with those obtained from females of N. barreti or N. recifensis from Brazil and females of N. houstoni from Australia described previously (Gondim Jr. and Moraes 2001, Schicha 1987, Furtado et al. 2005. We consider so far that our specimens can be anyone of the three species and that examination of the specimens collected in this study can lead to anyone of the three species. Consequently, the morphometrics strongly suggest synonymy.
There are however some discrepancies between our measurements and observations and previous descriptions of the three species. In the three descriptions: • Dorsal shield is reticulated in the description of N. barreti in the anterior lateral margins and starting to the back of s4 and on all the posterior part of the dorsal shield except the center; • VA shield has straight slightly convex margins in Schicha 1987 and not in the two other species as a concave part exists in this margin just after ZV2 position; • Z4 and Z5 are progressively tapered in drawings of N. barreti and Z5 seems longer (Furtado et al. 2005) vs for other species they are blunt/rounded apically (or more regularly parallel-sided); • Macrosetae SgeIV, StiIV and StIV are mentioned as setaceous for N. recifensis (Moraes and Gondim Jr 2001) but seem slightly knobbed on illustrations of the species. SgeIV and StIV are slightly knobbed for N. barreti and N. houstoni and StiIV is not mentioned for N. houstoni; • two teeth on the fixed digit and no tooth on the movable digit are mentioned for N. barreti and N. recifensis but three teeth for fixed digit and one recurved tooth for the movable digit are mentioned for N. houstoni; • Spermathecae of N. barreti and N. recifensis are mentioned as trumpet-shaped but as bellshaped for N. houstoni; • Setae JV5 seem longer in N. barreti description and shorter in those of N. houstoni and N. recifensis; • 6 poroids around genital/ventrianal shield are drawn for N. recifensis but only 4 for N. barreti and 0 for N. houstoni; • Occurrence of JV3 is mentioned in the text of description of N. barreti but not illustrated (Furtado et al. 2005). These setae are not indicated for description of N. houstoni and N. recifensis.
• Chaetotactic formulae are not precised in the description of N. houstoni but mentioned only for genu II for N. barreti and for genua II and III for N. recifensis.
Our examination of the type material for N. barreti (one paratype ♀), N. houstoni (the single specimen found, the ♀ holotype) and N. recifensis (the ♀ holotype, one ♀ paratype and seven additional ♀) shows: • dorsal shields of the three species present exactly the same reticulation as drawn in Furtado et al. (2005) for the description of N. barreti; • VA shield present only a slight concavity after ZV2 position in N. houstoni; • Z4 and Z5 are progressively tapered regularly parallel-sided in the three species and Z5 is of the same length (see table 8); • Macrosetae SgeIV, StiIV and StIV are: rounded apically for SgeIV, pointed apically for StiIV and slightly knobbed for StIV for the three species; • Three teeth on fixed digit (two strong anterior and one small tooth posterior to pilus dentilis) and one small recurved tooth in the anterior part of the movable digit for the three species, just as drawn by Schicha (1987); • Spermathecae of N. barreti and N. recifensis are drawn exactly in the same way in original descriptions (Schicha 1987;Moraes and Gondim Jr. 2001;Furtado et al. 2005) and are identical in the three species after our examination (and "bell-shaped" seems more appropriate for the shape description); • Setae JV5 are of similar length for the three species (see table 5); • 6 poroids are present around genital/ventrianal shield for the three species; • The mention of JV3 was an error in the text of the description of N. barreti. This seta is not present in the venter of that species and also not present in the two others; • Chaetotactic formulae are identical for the three species: Genu II: 2-2/0 -2/0-1 (seven setae); Genu III: 1-2/1 -2/0-1 (seven setae).
Considering all these information, we can conclude that the three species are synonyms. Consequently, our specimens are identified as the first species described among the three species, N. houstoni. The valid species name is thus Neoseiulus houstoni (Schicha). Previous specimens collected in La Réunion Island and identified as N. recifensis are renamed N. houstoni. Consequently, this is the first report of that species in La Réunion Island.
The male (for the three species) being unknown, it is herein described for the first time, based on La Réunion specimens.
Description of the adult male of Neoseiulus houstoni (Schicha)

n = 8 (Figs 1a-c)
Diagnosis -The following combination of characters indicated below in the description of the male of this species is quite similar to a lot of species of Neoseiulus belonging to the cucumeris and paraki species subgroups of the cucumeris species group. Not many characters allow to distinguish it from all males of other species if no females are collected in the same time: the peritreme reaching the level between j1 and j3, a limited reticulation compared to several species of these two subgroups, some dorsal setae length, especially s4, S2, Z4 and JV5 approximately of the same length (30 -35) and longest setae after Z5, a spermatodactyl with a terminal part recurved as an open U (hook-like), no additional macrosetae on other legs than leg IV compared to several species of these two subgroups that have macrosetae on leg III, sometimes II, only three pairs of preanal setae instead of 4 pairs in males of several species of these subgroups, one pair of crateriform gv3 very close to setae JV2.
Type material -Eight paratype males in total. Three paratype males deposited in Montpellier SupAgro -INRA Acarology collection, Montpellier and five paratype males deposited in Bassin-Plat CIRAD Research station collection in La Réunion.
Remarks -This combination of characters of the male of this species are not unique and do not allow to distinguish it from all males of other species of Neoseiulus belonging to the cucumeris and paraki species sub-groups without collected females of the species. The male peritreme is shorter than the female peritreme. Other characters are very similar.
This species is distributed in many countries of the world, mainly in tropical areas (Moraes et al. 2000;Mailloux et al. 2010;Kreiter et al. 2013Kreiter et al. , 2018Demite et al. 2019). It was found rarely in surveys made in Guadeloupe, Martinique and La Réunion except in studies on companion plants in citrus orchards (Mailloux et al. 2010;Kreiter et al. 2013Kreiter et al. , 2018cLe Bellec et al., unpub. data). This species seems actually to be more common on weeds with populations of tetranychid mites. Neoseiulus longispinosus, a type II phytoseiid predatory mite, as is N. californicus (McMurtry et al. 2013), has received increasing attention in Asia for the control of different spider mites (of Eutetranychus, Oligonychus, and Tetranychus) since 2010 (Nusartlert et al. 2011). The feeding, development, predation, cannibalism, intra-guild predation and behaviour have thus been extensively studied by several authors (see for example Luong et al. 2017) for pest control purposes. Neoseiulus longispinosus is well-known as a BCA sell in several countries in the world for the management of spider mites. The recent results of Huyen et al.(2017) show that at least in controlled laboratory conditions, N. longispinosus is a potential biological control agent against the citrus red spider mite P. citri. This is the first record of this species for La Réunion Island.  (Table 6) overlap with those obtained for populations of various countries. Measurements are slightly greater than those obtained on specimens of F.C.I. except for setae j4, J2, z5, StIV. On Comoros specimens, setae are longer except sternal shield length (st1-st3), inguinal sigilla (metapodal plates) and of macrosetae of basitarsus IV.
Like N. baraki (see above), N. lula belongs to the paspalivorus species group (Chant and McMurtry 2003a).
N. paspalivorus was found only on coconut and on fruits, in association with A. guerreronis (Moraes et al. 2004b). This species is a promising candidate for the biological control of the coconut eriophyid (Lawson-Balagbo et al. 2008).
Remarks: measurements of characters of the eight female and 3 male specimens ( Like N. barkeri and N. longispinosus, N. scapilatus belongs to the barkeri species group (see above) (Chant and McMurtry 2003a). Quilici et al. (2000) have collected before this species in La Réunion that is distributed in several countries of sub-Saharan Africa. Exact indications of locations were provided but without any measurements of specimens collected. Measurements of specimens collected during this study are provided in table 9. The biology of this species remains unknown.
Remarks: measurements of morphological characters of N. scapilatus female specimens from La Réunion (Table 9) are very close from measurements for specimens from neighbouring countries, except for specimens from South Africa that are larger (van der Merwe 1965). La Réunion specimens have slightly shorter macrosetae.

South-Africa Madagascar
Like N. barkeri, N. longispinosus and N. scapilatus, N. teke belongs to the barkeri species group and like N. longispinosus and N. scapilatus, it belongs to the womersleyi species subgroup (see above) (Chant and McMurtry 2003a).

Not applicable
Not applicable Not applicable Table 9 Character measurements of adult females and one adult male of Neoseiulus scapilatus collected in this study with those in previous studies (localities followed by the number of specimens measured). field conditions (Nwilene and Nachman 1996). Quilici et al.(2000) have collected this species before in La Réunion. Exact indications of locations were provided in the paper but without any measurements of specimens collected. Measurements of specimens collected during this study are provided in table 10.

La Réunion
Specimens examined: 12 ♀♀ in total + 6 ♂♂ + 2 im., 10 ♀♀ + 2 ♂♂ measured. Saint-Pierre -Bassin-Plat CIRAD Research Station (aasl 153 m, Long 55°29'18" E, Lat 21°19'25" Table 10 Character measurements of adult females and males of Neoseiulus teke collected in this study with those in previous studies (localities followed by the number of specimens measured).   (Table 10) are very close from measurements for specimens from neighbouring countries, especially from specimens from various countries in Africa, except for the holotype ) and specimens from South Africa which are larger (van der Merwe 1965). Kampimodromini Kolodochka 1998: 59;Chant & McMurtry, 2003b: 189;2006b: 137; In our specimens of this species, setae S5 are absent. So accordingly with Chant and McMurtry (2003b) all specimens belong to the orientalis species group.

Tribe Kampimodromini Kolodochka
Accordingly with these previous authors, and with , we consider that P. horrifer and P. orientalis are different valid species. Our specimens with longer setae s4, Z4, Z5, and lacking a distinctly short, thick, spatulate macroseta on genu I. This species is widely distributed in Sub-Saharan Africa and Madagascar. The biology of P. horrifer remains totally unknown. This is the first mention of this species from La Réunion Island. Specimens examined: 13 ♀♀ + 1 ♂ + 1 im. in total, 12 ♀♀ + 1 ♂ measured. Le Tampon -Ligne des 400 (aasl 463 m, Long 55°30'36" E, Lat 21°17'24" S), 1 ♂ + 2 im. on  (Table 11) agree well with measurements found in the literature, being however slightly shorter for almost all characters compared with female and male specimens of South Africa (van der Merwe 1968).
Our specimens with shorter s4, Z4 and Z5 setae, having a distinctly short, thick, spatulate macroseta on genu I belong to P. orientalis.
This species is widely distributed in tropical and subtropical areas in South America, Africa and Asia. This species belongs to a genus included in the large polyphagous generalist group named type III phytoseiid mites (McMurtry and Croft 1997;McMurtry et al. 2013). Navasero and Navasero (2016) have studied the life history of P. orientalis on the broad mite (P. latus) as prey. The authors reported high predation rates on the eggs of P. latus, suggesting good potential for the control of this pest. Quilici et al. (2000) have collected before this species in La Réunion but provided no measurements. We herein provide measurements of specimens collected in La Réunion (Table 12).
Specimens examined: 20 ♀♀ + 1 ♂ + 1 im. in total, 5 ♀♀ + 1 ♂ measured. Saint 48 ( 12 (9 -13) 11 (10 -12) 8 -10 14 11 10 -z4 11 (9 -13) 8 (6 -11) 8 -10 9 6 10 -z5 80 ( Phytoseiulus persimilis is a Mediterranean/subtropical predatory mite, a type I species, i.e. a specialist predator of the urticae species group of the genus Tetranychus (McMurtry and Croft 1997;McMurtry et al. 2013). Considerable research has been conducted on this predator-prey interaction (see review by Kostiainen and Hoy 1996), and numerous biological control programs have used P. persimilis against T. urticae on a wide range of ornamental and vegetable crops. Phytoseiulus persimilis was the first greenhouse biological control agents available commercially and it is one of the most successful BCA in the world. It can also be used in temperate climates on open-field crops such as strawberries. Optimum conditions are 20-27°C and relative humidity of 60-90 %. Cooler or warmer temperatures may have a negative effect on reproduction, development and efficiency of this predatory mite. This species is present in Mauritius (Kreiter et al. 2018a) and La Réunion probably because of its commercial introduction and uses in vegetable and ornamental greenhouses, dispersion of some specimens released and establishment in the environment. This species is actually reared and sold in La Réunion and commercialised in Mascareignes since a long time (Quilici, personal communication). Phytoseiulus persimilis was already known from La Réunion (Quilici et al. 1997(Quilici et al. , 2000. Exact indications of locations were provided in these papers but without any Table 12 Character measurements of adult females and one adult male of Paraphytoseius orientalis collected in this study with those in previous studies (localities followed by the number of specimens measured).
This species is mentioned only from the Indian Ocean area, Madagascar (Blommers 1974), La Réunion (Quilici et al. 2000) and recently Mauritius (Kreiter et al. 2018a). Species of this genus Scapulaseius are supposed to be of type III (McMurtry and Croft 1997;McMurtry et al. 2013), i.e. polyphagous generalist predators. However, the biology of S. reptans remains unknown. S. reptans was already mentioned from La Réunion by Quilici et al.(2000). Exact indications of locations were provided in this paper but without any measurements of specimens collected. They are provided for specimens collected during this study and listed in table 15.
Ferragut and Baumann (2019) discussed a possible synonymy between S. reptans and S. asiaticus (Evans) recently. Our specimens of La Réunion Island markedly differ from specimens of S. asiaticus collected recently in Vietnam (Kreiter et al. in prep.) and we therefore disagree with this possible synonymy with arguments that will be developed in a future paper.
Transeius maelliae Kreiter n. sp. is quite similar to the other new species described further, T. mickaeli Kreiter n. sp. with which it was at first confused early on during the identification process. T. maelliae n. sp. is different from T. mickaeli n. sp. by: having only five solenostomes instead of seven, a longer seta z4, almost double, the shape of the spermatheca also bell-shaped but with an undistinct atrium, longer setae JV5, less teeth on both digits of chelicerae and very slightly reticulated ventral shields (see table 27).
In the species subgroup bellottii, the species closest to T. maelliae n. sp. is Transeius jujae El-Banhawy and Knapp. T. maelliae n. sp. resembles to T. jujae, having similar lengths for setae j3, s4, Z4, Z5, and JV5, and the dimensions of the spermatheca. The new species can be distinguished however by the longer length of setae z2, above all by the length of setae z4 which is 2.5 as long as that of T. jujae, and by the shape of the spermatheca with an undifferentiated atrium and walls of calyx slightly converging apically in the new species opposed to a distinct atrium and walls of calyx strictly parallel in T. jujae. In addition, a macroseta is existing in the genu of leg I of T. maelliae n. sp. and lacking in specimens of T. jujae, the ventrianal shield is wider and the dorsal shied is totally smooth in that species.
Peritreme - (Fig. 2a). Reaching the level of j1. Venter - (Fig. 2b). All ventral shields smooth. Sternal shield with three pairs of setae and two pairs of lyrifissures; one pair of sternal setae on a small metasternal shield with one poroid; posterior margin of the sternal shield straight to slightly convex. Distances between st1-st1 66 Chelicera -Chelicerae visible but dorsoventrally oriented; therefore, they are not drawn. Fixed digit 36 (35 -38) long with no discernible tooth on the four females; movable digit 39 (38 -40) long with putatively 2 teeth.
Etymology -The name "maelliae" refers to the stepdaughter of Serge Kreiter, Maëllia Gaultier, to whom the new species is dedicated.
Remarks - Table 16 shows comparison of T. maelliae n. sp. and the closest species in the whole genus Transeius, T. jujae (see diagnosis above) and the next new species described below.
All specimens of T. maelliae n. sp. were found between 1300 and 1500 m aasl and in humid tropical forests, along with two other species, the next new species described below and Amblyseius neoankaratrae Ueckermann and Loots.
Transeius mickaeli n. sp. is different from T. maelliae n. sp. described above, by several characters indicated above in the description of T. maelliae (see diagnosis of T. maelliae).
Venter - (Fig. 3b). All shields very slightly reticulated. Sternal shield with three pairs of setae and two pairs of lyrifissures; one pair of sternal setae on elongate metasternal shields with a pair of pores; posterior margin straight to very slightly convex. Distances between st1-st1 62 Spermatheca - (Fig. 3c). Spermatheca pocular (Denmark et al. 1999), with an unelongate calyx 15 (14 -15) long and 8 (5 -10) wide, a differentiated atrium at the basis of the calyx. Visible short ductus minor and a long ductus major.
Type material -The holotype female, five paratype females, 2 paratype males and 4 paratype immatures in 4 slides deposited in CBGP, in Montpellier SupAgro Acarology collection, France.
Etymology -The name "mickaeli" refers to the stepson of Serge Kreiter, Mickaël Gaultier, to whom this new species is dedicated. Remarks - Table 16 shows comparison of T. mickaeli n. sp. and the closest species, T. maelliae n. sp. and T. quichua (McMurtry and Moraes).
All specimens of T. mickaeli n. sp. were found between 1300 and 1500 m aasl and in humid tropical forests, sometimes in mixed populations with T. maelliae n. sp. and/or with A. neoankaratrae. Just like the two previous new species, T. soniae belongs to the bellottii species group. It belongs to the namurensis species subgroup as the spermatheca has the calyx swollen basally, then narrowing and flaring distally. This species subgroup contains only six species (Chant and McMurtry 2004a, Tixier et al. 2016).

Transeius soniae Zannou, Moraes & Oliveira
This This is the first mention of this species in another country than Kenya and the first mention for La Réunion Island.
Remarks: measurements of specimens of La Réunion (Table 17) fit well with those already published in the literature, especially with those of the original description of Zannou et al.(2007). Some setae such as j4 and j5 are longer in La Réunion specimens than those in Kenya specimens. Setae s4, S2, and Z4 (between 14 and > to 22 %) and in a lesser extent Z1 are however shorter than those collected in Kenya ; El-Banhawy and Knapp 2011).
Diagnosis -Amblyseius djenaeli Kreiter n. sp. belongs to the subfamily Amblyseiinae (absence of dorsolateral setae z3 and s6 and the caudoventral setae JV3), to the tribe Amblyseiini (setae j3, s4, Z4 and Z5 longer than other setae, ratio s4 / Z1 > 3.1, many teeth on the fixed cheliceral digits and macrosetae on legs I, II and/or III[?] in addition of macrosetae on leg IV), to the subtribe Amblyseiina (sternal shield as long as wide, ventrianal shield longer than wide, seta J2 present, genital shield almost as wide as ventrianal shield, ventral shields generally smooth, macrosetae on all legs, setae j5, J2, S2, S4, S5 and Z1 present), to the genus Amblyseius (ratio s4 / S2 > 3.0, chelicerae of normal size with fixed digit of the same size as movable digit, seta JV2 present, without incision in lateral margin of dorsal shield at the level of seta s4, ventrianal shield not reduced to a simple anal shield, Ge III and Ti III each generally with a macroseta) (Chant and McMurtry 2007).
Setae J2 and Z1 are present, dorsocentral setae and setae z2, z4, Z1, S2, S4, and S5 are minute, setae s4, Z4 and Z5 are prominent, elongate and whip-like, female ventrianal shield usually pentagonal, as wide at level of anus than at level of setae ZV2 or wider at this later level, which allows to classify this new species in the species group obtusus (Chant and McMurtry 2004a).
Like A. tamatavensis, A. djenaeli n. sp. belongs to the species subgroup aerialis with spermatheca tubular. This subgroup contains 46 species (Chant and McMurtry 2004a). Many of those species are very different from the new species (Table 18).
The type material of the more similar species, A. solani, was requested in Cuba for comparison with A. djenaeli n. sp. without any success.
Spermatheca - (Fig. 5d). Spermatheca tubular (Denmark et al. 1999), with an elongate calyx with parallel margin 24 (23 -25) long and 3 wide, an atrium included in the basis of the calyx. Visible small ductus minor and large membranous ductus major.
Legs (Fig. 5e)  Etymology -The name "djenaeli" refers to the first name of the stepson of the senior author, Djénaël Gaultier. The species is named in his honour.
Amblyseius herbicolus belongs to the largoensis species group as setae J2 and Z1 are present, setae s4 are minute and the ventrianal shield of the female is vase-shaped. It belongs to the largoensis species subgroup as setae Z4 are long, spermatheca has the calyx elongate and the female ventrianal shield is entire (Chant and McMurtry 2004a).
Amblyseius herbicolus is widespread in all tropical and subtropical regions of the world. It is the second most abundant phytoseiid mites on Coffea arabica L. in Brazil, associated with Brevipalpus phoenicis (Geijskes), vector of the coffee ring spot virus and it was found to be an efficient predator (Reis et al. 2007). A. herbicolus is also found associated with the broad mite, P. latus in crops such as chili pepper (C. annuum) in Brazil and has also a good potential for controlling the pest. Rodriguez-Cruz et al.(2013) have studied biological, reproductive and life table parameters of A. herbicolus on three different diets: broad mites, castor bean pollen (R. communis) and sun hemp pollen (Crotalaria juncea L.). The predator was able to develop and reproduce on all these three diets. However, its intrinsic growth rate was higher on broad mites and castor bean pollen. Feeding on alternative food such as pollen can facilitate the predator's mass rearing and maintain its population on crops when prey is absent or scarce. Many polyphagous generalist phytoseiid mites are important natural enemies because they can feed on plant provided pollen and various prey species, and thus persist in crops even in the absence of target pests (McMurtry et al. 2013). Hence, populations of these predators can be established in a crop by providing alternative food, thus increasing biological control. Alternative food affects P. latus control on chilli pepper plants by predatory mites (Duarte et al. 2015). A. herbicolus had high oviposition and population growth rates when fed with cattail pollen (Typha latifolia L.), chilli pepper pollen and bee-collected pollen, and a low rate on the alternative prey T. urticae. Supplementing pepper plants with pollen resulted in better control of broad mite populations (Duarte et al. 2015). Release of A. herbicolus on young plants with weekly addition of honeybee pollen or cattail pollen until plants produce flowers seems a viable strategy to sustain populations of this predator (Duarte et al. 2015).
Males collected in this study appear to be compatible with the description of adult males of A. herbicolus. Unfortunately, they were collected alone, without females or any association link with a nearby (female) population of A. herbicolus. So they cannot be reliably identified as males of A. herbicolus. We have collected 114 females from different localities, habitats and plants of the Island, with and no sign of males despite females sometimes collected in high densities. The thelytoky of this species is thus strongly suspected.
The biology of this species remains totally unknown. A. longipilus has been described previously by Kreiter et al.(2002) from La Réunion but have never been recorded from other countries or record again in La Réunion since its description. So this is the first new record since the original description.
Remarks: the single description available in the literature is the original description of Kreiter et al.(2002). Measurements obtained in this study (Table 20) fit well with those of the original description, except for JV5 that is a far greater in the new specimens collected in this study. JV5 setae were measured 16 (13 -18) in the original description and this very small size is quite surprising for setae belonging to an Amblyseius species.
We have measured again these setae in 11 specimens, 1 holotype and 10 paratype females and we found 69 (63 -75) (11), very close to the value for the three specimens found in this study.
The first values published were a mistake of measurements and they must be replaced by value of the table 20. This brown-reddish species was collected in numbers in high altitude > 1000 m aasl. This species is peculiar in the genus Amblyseius by having ratios s4 / Z1 and s4 / S2 <3.1, by the shortness of setae Z4 and by the absence of macroseta on tarsus IV. Like A. herbicolus, it belongs to the largoensis species group as setae J2 and Z1 are present, setae s4 are minute and the ventrianal shield of the female is vase-shaped. It belongs to the species subgroup ankaratrae as setae Z4 are short. Only two species are in the subgroup ankaratrae and another one in the close species subgroup nahatius with only one species. Amblyseius nahatius is very different from the two first with its divided ventrianal shield, its shorter setae, and very different insemination apparatus.

Amblyseius neoankaratrae Ueckermann & Loots
This species seems to live in high altitude habitats, such as T. maelliae n. sp. and T. mickaeli n. sp. with which it makes apparently mixed colonies. The biology of that species is totally unknown.
The male of that species was unknown until now and is described thereafter. This is the first mention of this species in another country than South Africa and Mauritius and the first mention from La Réunion Island.
We have no indications of the altitude of collection for the holotype from South Africa but it seems that this species is only met in the highest part of regions in which it is present.
Remarks -The male of A. neoankaratrae has relatively unique combination of characters that can allow to distinguish easily it from males of other species of Amblyseius, even in species subgroups ankaratrae and nahatius. It has very short setae Z4 permitting to distinguish it from any other males of all other species group of Amblyseius. Concerning the species subgroup ankaratrae and nahatius, with two and one species described in these subgroups, respectively, the only male described is for Amblyseius ankaratrae Blommers. The male of A. neoankaratrae can be distinguish from the male of A. ankaratrae by longer Z4 (19 vs 9) and s4 (30 vs 13), shorter Z5 (122 vs 170) and macrosetae of leg IV (50 and 55 vs 70 and 64 for genu and tibia, Figure 6 Male of Amblyseius neoankaratrae Ueckermann and Loots: a -Dorsal shield and peritreme, b -Ventral shield, c -Spermatodactyl respectively), the presence of a macroseta on the basitarsus IV, less teeth on the fixed digit (6 vs 7) and a longer toe in the spermatodactyl (10 vs 7).
Amblyseius tamatavensis belongs to the obtusus species group as setae J2 and Z1 are present, setae z4 are minute and the female ventrianal shield is not vase-shaped or divided. It belongs to the aerialis species subgroup (46 species) as the calyx of the spermatheca is tubular (Chant and McMurtry 2004a).
It seems to fit the functional type III-b (generalist predators living on glabrous leaves) group defined by McMurtry et al.(2013). Cavalcante et al. (2017) reported this species as a promising natural enemy of B. tabaci. Experimental releases of this predator on caged plants in a screenhouse caused the reduction of the density of B. tabaci on pepper plants by up to 60-80 % (Massaro and Moraes 2019). It can be easily produced in large numbers (Massaro et al. 2018) when fed with astigmatine mites, which could allow the mass production for augmentative biological control. This species is reported in tropical areas from over 20 countries around the world (Africa, Asia, America and Oceania). It was already well known from La Réunion since previous studies (Quilici et al. 2000). All details of collections were provided in this paper but without any measurements of specimens collected given. Measurements of specimens collected during this study are provided in table 22.

Proprioseiopsis versutus ( Zack) (Synonymy according to Denmark & Evans 2011).
Proprioseiopsis mexicanus belongs to the belizensis species group as genu I have no Table 22 Character measurements of adult females and males of Amblyseius tamatavensis collected in this study with those in previous studies (localities followed by the number of specimens measured).

Not applicable
Not applicable Not applicable macrosetae. As the spermathecal of that species has a short calyx, cup-shaped, it belongs to the asetus species subgroup (Chant and McMurtry 2005a). This species is known from all islands of French West Indies (Kreiter and Moraes 1997;Moraes et al. 2000, Kreiter et al. 2006Mailloux et al. 2010;Kreiter et al. 2018c) but it was found only in very large number during a previous study on companion plant in Guadeloupe (Mailloux et al. 2010) and in an actual study in La Réunion (Le Bellec, unpub. data). This species seems to be very abundant on weeds in the lower vegetation. Phytoseiid mites of the genus Proprioseiopsis have been found mainly in ground surface, humus, litter, soil, moss or on grass (Muma and Denmark 1970;McMurtry et al. 2015).
Proprioseiopsis mexicanus population increase when fed T. urticae eggs (Megevand et al. 1993) and this species seems to be a good predator of thrips (Kreiter, unpub. data). It is one of the prevailing phytoseiid species on citrus orchards in Alabama (Fadamiro et al. 2009). Denmark and Evans (2011) mentioned that the species can be reared on T. urticae and Oligonychus pratensis (Banks) and is associated with Bryobia praetiosa Koch, Bryobia sp. and P. ulmi. It was also found in association with Tetranychus evansi Baker and Pritchard (Furtado et al. 2014) but mentioned as a poor predator of that species. The biology of this species is however almost unknown.
Proprioseiopsis mexicanus was already recorded by Quilici et al.(2000) and all details of collections were provided but measurements of specimens collected and identified were not published. Measurements of specimens collected during this study are provided in table 23.
Specimens examined: 194 ♀♀ + + 2 ♂♂ + 9 im. in total, 13 ♀♀ + 1 ♂ measured (1 ♂ (Table  23) fit well with all those indicated in Kreiter et al.(2018c) for various countries. All setae and dimensions of specimens of this study have however slightly longer in females and males compared to specimens from other countries.
Like P. mexicanus, P. ovatus belongs to the belizensis species (see above). As the spermatheca of that species is saccular, it belongs to the belizensis species subgroup (Chant and McMurtry 2005a).
This species is known from Guadeloupe, Marie-Galante and Martinique (Kreiter and Moraes 1997;Moraes et al. 2000;Mailloux et al. 2010;Kreiter et al. 2018c). This species was found in very large number only during a previous study on companion plant in Guadeloupe (Mailloux et al. 2010) and in a recent study in La Réunion (Le Bellec, unpub. data). In other habitats, this species seems to be rare. This species like P. mexicanus seems to be abundant on weeds in the lower vegetation. Denmark and Evans (2011) indicated that this species is associated with O. pratensis and Brevipalpus sp. It was also found in association with T. evansi (Furtado et al. 2014) but mentioned as poor predator of that species. Despite this information, the biology of this species remains unknown. This is the first mention of this species from La Réunion.  Tampon  Remarks: measurements of female specimens of La Réunion (Table 24) fit well those obtained for populations of various countries. Despite the great number of females sampled, no male was recorded.
Typhlodromalus spinosus was collected in eastern, western but mainly southern Africa and in La Réunion (Demite et al. 2019). The rapid multiplication of this species on the western flower thrips (WFT), F. occidentalis, was confirmed and clear evidence that T. spinosus predates on WFT under laboratory and field conditions but not on T. urticae was established (Mwangi et al. 2015). This species seems abundant in low vegetation as it was found in high populations in a study of companion plants in citrus orchard (Le Bellec et al. unpub. data). This species have never been record in Guadeloupe or Martinique in similar studies but it is interesting to notice that in those islands, another Typhlodromalus was collected, T. peregrinus (Muma) (Mailloux et al. 2010;Kreiter et al. 2013Kreiter et al. , 2018c. T. spinosus was already recorded by Quilici et al.(2000) and all details of collections were provided but measurements of specimens collected and identified were not published. Measurements of specimens collected during this study are provided in  Chant & McMurtry, 2005b: 203;2007: 115. This species was described as Typhlodromalus nesiotus Ueckermann and Kreiter in Kreiter et al.(2002) but as all setae are short to minute, the setae Z4 are not as long as distance between their base and that of setae S4 and the dorsal shield is smooth except some anterolateral striations, it belongs to the genus Ueckermannseius (Chant and McMurtry 2005b). This species has never been recorded from other countries or record again in La Réunion since its description. The biology of this species remains totally unknown. This is the first new record since the original description in 2002. Specimens examined: two ♀♀ in total, both measured. Petite Île -Piton Bloc, Yébo Luguy farm (aasl 973 m, Long 55°34'64" E, Lat 21°18'64" S), 1 ♀ on R. raphanistrum, 9/12/2015; Forêt de Sans Souci -Ilet Alcide, (aasl 1452 m, Long 55°22'07" E, Lat 21°01'17" S), 1 ♀ on Hypericum lanceolatum Lam., 18/11/2018. Remarks: the two females collected and measured during this study (Table 26) slightly differ from females measured by Kreiter et al.(2002). Dorsal shield dimensions are greater by 15-26%, length and width of the sternal shield are greater by 25 and 20% respectively, and the length of JV5 is greater by 60%.

Ueckermannseius parahavu Moraes, Zannou & Oliveira
Ueckermannseius parahavu : 36, Chant & Mcurtry 2007  Remarks: measurements of morphological characters of the single female found in La Réunion (Table 28) are 6-50% smaller than those obtained on specimen from Kenya, especially j1, j6, J2, s4, S2, z2, z4, Z5 and length of chelicerae. Measurements were however taken on a single specimen from each country, and so this data must be compared with caution.
For the males (Table 29), setae Z5 is shorter and ventrianal shield is longer in specimens from La Réunion Island. This species belongs to the plumifer species group as setae J2 and R1 are present (Chant and McMurtry 1994). It is widely distributed in sub-Saharan Africa, Madagascar (Demite et al. 2019), in La Réunion (Quilici et al., 2000;Demite et al. 2019) and recently from one island of the Comoros Archipelago, Grande Comore (Kreiter et al. 2018b).
Remarks: specimens of this species have been recorded and measured from Burundi. Measurements of our specimens from La Réunion (Table 31) fit well with those of Burundi with sometimes slightly greater values, for example for setae s4 and Z4.
Values for both type of specimens are higher than those obtained for specimens from Asia (Table 31).

Phytoseius haroldi Ueckermann & Kreiter
Phytoseius haroldi Ueckermann & Kreiter, 2002: 339;Chant & McMurtry 2007: 129. This species belongs to the horridus species group as setae J2 and R1 are absent (Chant and McMurtry 1994). This species was described Kreiter et al.(2002) but have never been recorded from other countries or recorded again in La Réunion since its description. This species was abundant on lower vegetation in a study of companion plants in citrus orchard. It seems that this species prefers low plants but despite this observation that has to be confirmed, the biology of this species remains totally unknown. Measurements of some additional females (9) and of some males are provided in table 32. This is the first new record In La Réunion since the original description and the second record in Indian Ocean after recent records of Ferragut and Baumann (2019).
Remarks: measurement values of morphological characters of specimens from La Réunion (Table 33) and specimens from neighbouring countries are very close, especially for specimens from Africa. Some values for specimens from Asia are slightly lower.
Remarks: measurement values of morphological characters (Table 34) were compared to those obtained for specimens of India, the only country in which Phytoseius punicae was recorded and specimens measured. Values of measurements are very close except for seta j3, z3 and st1-st3 length of sternal shield that are smaller compared to specimens of India. Macrosetae StIV is however longer in the single specimen of La Réunion Island.  Moraes et al. 1986: 229;2004a: 258;Chant & McMurtry 2007: 131. This species belongs to the horridus species group as setae J2 and R1 are absent (Chant and McMurtry 1994).

La Réunion
Remarks: all measurement values (Table 39) agree well with those already published on this species with only very slight variations. Measurement values of female specimens of La Réunion are very similar with values for specimens of Kenya and South Africa. Despite the large number of females collected, sometimes from a single location, we did not find any males in our study, suggesting that this predatory mite can reproduce through thelytokous parthenogenesis, as observed by Kishimoto (2015).

Conclusion
Before this 10 th paper, the fauna of Phytoseiidae of La Réunion Island was limited to 33 recorded species including 24 Amblyseiinae, 5 Phytoseiinae and 4 Typhlodrominae, among which 8 species that had been described as new.
This paper reports on results of surveys done recently (2015-2018) and add 19 newly recorded species among which 3 are new to Science and 21 already known species but with additional data. The number of species for La Réunion Island now reached to 52 among which 11 have been described only from this Island (P. haroldi was found in Mauritius recently).
Among these 19 species newly recorded, at least nine are already well known as biological control agents (BCA). This must be underlined as new regulations on importation of macroorganisms are proposed in a lot of countries and specifically for countries like France that have very far over-sea territories. Therefore, it is impossible to import and of course to sell and use exotic species if they are not indigenous in La Réunion Island that is considered as a territory. An importation permit must be requested, but it is expensive and chances to obtain are generally very low (Kreiter et al. 2016). The knowledge of the biodiversity, especially of efficient biological control agents from overseas territories, not only for fundamental biodiversity knowledge or conversation purposes but also for agricultural and economical ones, is so of a considerable importance.  (2014); Japan: Ehara & Kishimoto (2007); South Africa (identified as T. jackmickleyi but synonymized by Denmark & Muma 1968): van der Merwe (1968); Taiwan: Tseng (1983); Holotype South Africa and New Caledonia: Schicha (1981a); -: not provided. , Acarologia 60 (1)