New records of phytoseiid mites (Acari: Mesostigmata) from Madeira Island

Madeira is the largest of the four islands constituting Madeira Archipelago in the North Atlantic Ocean. It is located at 400 km from the Northern Canary Islands, at 500 km from Morocco and between 900 and 1000 km from South Portugal and Spain. So far, nineteen species of the mite family Phytoseiidae had been reported from this island. We report in this paper the results of a survey conducted in May 2019 in Madeira Island, in which 15 species have been found, six being new for the Island fauna.


Introduction
Species of the family Phytoseiidae are all predators of phytophagous mites and small insects like thrips and whiteflies, both on crops and wild vegetation. Several species are biological control agents of pest organisms in both open and protected crops all around the world (McMurtry and Croft 1997; McMurtry et al. 2013; Knapp et al. 2018. This family is widespread around the world, present on all continents except Antarctica, and consists of about 2,500 valid species, dispatched in 94 genera and three subfamilies (Demite et al. 2021).
Biodiversity surveys in poorly investigated areas is still an urgent need and might result in the discovery of additional species potentially useful for biological control as well as in getting more information on the biodiversity of these areas in the context of global climatic changes (Kreiter et al. 2018a, b, c, 2020a, b, c, d, 2021a, b, c; Ferragut and Baumann 2019, 2021; Döker et al. 2020; Kreiter and AboShnaf 2020a, b; Fang et al. 2020; Demite et al. 2021; Toldi et al. 2021. In these perspectives, the most interesting areas are probably those with a high level of biodiversity. Macaronesia is renowned for its biodiversity, with extraordinary high levels of species diversity and endemism in both the terrestrial and marine realms and constitutes a biodiversity hotspot (Kondraskov et al. 2015). Those areas, being called hotspots, were defined by Myers (1988) in order to identify the most immediately important areas for biodiversity conservation. The common characteristics of these hotspots are that they hold high endemism levels and have lost at least 70% of their original natural vegetation (Myers et al. 2000).
Located in the Atlantic Ocean at around 500 km away from the western coast of Morocco, about 400 km from the Northern Islands of the Canary Archipelago in the socalled Macaronesia region, Madeira Island is one of the four main islands constituting Madeira Archipelago, with Porto Santo and the two main Desertas Islands, Deserta Grande and Bugio, the largest of the four ones, with more than 740 km² on a total of 801. This Archipelago was a historical crossing point just like Canary Archipelago, with ships stopping at departure towards Africa or South America or ships returning to Europe full of Tropical plants. Both native and imported species are thus expected to be present.
Therefore, the number of the recorded species is of 19 for Madeira Island, prior to this study.
The objective of this paper is to report the phytoseiid species found in a survey conducted by the senior author in May 2019 in Madeira Island.

Material and methods
The survey took place in Madeira Island from 13 th to 23 rd of May, 2019. Plant inhabiting mites were collected from cultivated and wild plants in all parts of the island.
Mites were directly collected from leaves with a fine brush with or without a pocket lens (large leaves and herbaceous plants) or by beating the plants (mainly shrubs and trees with very small or spiny leaves) and collecting the mites in a black plastic rectangular saucer 45 x 30 cm (Ref. STR 45, BHR, 71370 SaintGermainduPlain, France). Collected mites were then transferred into small plastic vials containing 1.5 ml of 70% ethanol by using a fine brush.
A total of 116 Phytoseiidae in 29 samples were collected and the percentage of some species have been calculated by dividing the number for those species by the total number of phytoseiids or the number in which the species is present by the total number of samples.
The mites were then slidemounted in Hoyer's medium (Walter and Krantz 2009), the slides were dried at 4550 o C for at least two weeks and then all examined and identified using a phase and interferential contrast microscope (DMLB, Leica Microsystèmes SAS, Nanterre, France). Characters of specimens were measured using a graded eyepiece (Leica, see above). McMurtry's (1994, 2007) concepts of the taxonomy of the family Phytoseiidae for identification and the world catalogue database of Demite et al. (2014Demite et al. ( , 2021 for distribution and information on descriptions and redescriptions were used. The setal nomenclature system adopted was that of Lindquist & Evans (1965) and Lindquist (1994) as adapted by Rowell et al. (1978) and Chant & YoshidaShaul (1989) for the dorsal surface and by Chant & YoshidaShaul (1991) for the ventral surface. Pore (= solenostome) and poroid (= lyrifissure) notations are that of AthiasHenriot (1975). Macrosetal notation (Sge = genual macroseta; Sti = tibial macroseta; St = tarsal macroseta) is that of Muma and Denmark (1970). Types of spermatheca or insemination apparatus are those proposed by Denmark and Evans (2011). Numbers of teeth on the fixed and movable cheliceral digits do not include the respective apical tooth. Setae not referred to in results section should be considered as absent.
All measurements are given in micrometres (µm) and presented with the mean followed by the range in parenthesis (data concerning our specimens are in bold). Classification of plants follows the APG IV classification of 2016 (ex. Byng et al. 2018).
Specimens of each species are deposited in the mite collections of Montpellier SupA gro conserved in UMR CBGP INRAE/IRD/CIRAD/Institut Agro (SupAgro)/University of Montpellier.
The following abbreviations are used in this paper for institutions: CBGP = Centre de Biologie pour la Gestion des Populations; CIRAD = Centre International de Recherche Agronomique pour le Développement; IA = Institut Agro; INRAE = Institut National de Recherche pour l'Agriculture, l'Alimentation et l'Environnement; IRD = Institut de Recherche pour le Développement; MSA = Montpellier SupAgro, France; UMR = Unité Mixte de Recherche.

Results and Discussion
A total of fifteen species had been found during this study presented hereafter. Six species are presented with new measurements compared to only few references already available in the literature. Neoseiulus madeirensis belongs to the cucumeris species group of Neoseiulus, as the dorsocentral setae are not short relatively to dorsolateral setae. It belongs to the cucumeris species subgroup as the spermatheca does not have a stalk between calyx and atrium, the atrium is undifferentiated or nodular and joined directly to calyx (Chant and McMurtry 2003a).

Subfamily Amblyseiinae Muma
This species was already recorded in Madeira Island as it was collected and described by Papadoulis and Kapaxidi (2011) Papadoulis and Kapaxidi (2011). We have added some measurements of some important characters not documented before as it is only the second record of that species from Madeira Island. Our specimen was collected at Ribeira Brava, at only 20 km from the botanical garden of Funchal from where the species has been described.
Neoseiulus teke belongs to the barkeri species group and the womersleyi species subgroup (Chant and McMurtry 2003a). This species is reported from subSaharan Africa often associated with Mononychellus tanajoa (Bondar), the cassava green mite (CGM) (Mutisya et al. 2017). It has been studied for its potential as BCA against the CGM. Nwilene and Nachman (1996) studied its reproduction characteristics on M. tanajoa. It was more efficient than I. degenerans, but seems not efficient enough in field conditions (Nwilene and Nachman 1996). This is the first report of this species from Madeira Island. The Madeira population might have originated Table 1 Character measurements of an adult female of Neoseiulus madeirensis collected in this study with those obtained from previous studies (localities followed by the number of specimens measured between brackets).

Characters
Madeira  Moraes et al. 1986: 99, 2004: 149, Chant & McMurtry 2003a: 23, 2007 Similar to the morphological characters already mentioned for N. madeirensis, this species is also belonging to the cucumeris species group of Neoseiulus and to the cucumeris species subgroup (Chant and McMurtry 2003a).
Very few studies exist on its biology. Knisley and Swift (1971) and Kazak et al. (2002) showed its ability to develop feeding on Tetranychus urticae Koch. Sengonca and Dresher (2001) studied the ability of this species to feed and develop on Thrips tabaci Lindeman and concluded that this food alters its biological parameters in comparison to T. urticae. It seems that N. umbraticus able to develop and reproduce also on Panonychus ulmi (Koch), Calvolia lordi (Nesbitt), Aculus schlechtendali (Nalepa), adults of Quadraspidiotus perniciosus (Comstock), and on apple and cherry pollens (Knisley and Swift 1971). Adults of Agistemus fleschneri Summers, and winter eggs of P. ulmi were not fed (Knisley and Swift 1971). This is the first report of this species from Madeira Island. World distribution: Armenia, Azerbaijan, Azores, Belarus, Caucasus Region, Denmark, England, France, Georgia, Germany, Hungary, Iran, Italy, Jamaica, Latvia, Mexico, Moldovia, Montenegro, Morocco, Norway, Poland, Russia, Slovakia, Spain, Switzerland, Turkey, Ukraine, USA.

Tribe Amblyseiini Muma
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 of that species (Reis et al. 2007). Amblyseius herbicolus is also found associated with the broad mite, Polyphagotarsonemus latus (Banks), in crops such as chili pepper (Capsicum annuum L.) in Brazil and has also be proposed as a good potential for controlling this pest. RodriguezCruz et al. (2013) had studied biological, reproductive and life table parameters of A. herbicolus on three different diets: the broad mite, castor bean pollen (Ricinus communis L.) and sun hemp pollen (Crotalaria juncea L.). The predator was able to develop and reproduce on all these three diets. However, its intrinsic rate of natural increase was the highest on broad mites and castor bean pollen. Feeding on alternative food such as pollen can facilitate the predator's mass rearing and maintains 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). Amblyseius herbicolus had high oviposition and population growth rates when fed with cattail pollen (Typha latifolia L.), chilli pepper pollen and beecollected pollen, and a low rate on the alternative prey (Tetranychus urticae Koch). 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). This is the second report of that species from Madeira Island after the recent record of Ferragut and Baumann (2020).  Remarks: morphological and morphometric characters and all measurements fit well with those provided in Kreiter and AboShnaf (2020a, b) and Kreiter et al. (2021a, b) for specimens of the Indian Ocean. Amblyseius herbicolus was previously recorded in many countries of the world and especially in French West Indies (Moraes et al. 2000. It is also reported recently from Vietnam (Kreiter et al. 2020c), Rodrigues and Mauritius Islands (Kreiter and AboShnaf 2020a, b) but only from females, like in Ferragut and Baumann (2020) and this study.
We agree with the opinion of Ferragut and Baumann (2020), as the report of A. largoensis by Carmona (1973) could represent a misidentification. The same conclusion was drawn by Döker et al. (2020) for specimens of both species in Turkey. However, Carmona's specimens were not available for examination by the former authors. Just like the former authors, we found that A. herbicolus was one of the most abundant phytoseiid species (35.3% of all the phytoseiids collected for them and 21% for us) and frequent (19% of samples for them and 31% for us) in material collected in the Madeira archipelago. Moreover, many confusions were made in the literature between A. largoensis and A. herbicolus, these two species being separate by only three characters: the shape of the calyx of spermatheca (McMurtry and Moraes 1984; Döker et al. 2020, the size of atrium compared to the base of calyx and the number of teeth on the movable digit (Döker et al. 2020 Moraes et al. 1986: 52, 2004b: 82, Chant & McMurtry 2005: 216, 2007 The 200 species of the genus Euseius are considered as Type IV species that are pollen feeding generalists predators (McMurtry andCroft 1997; McMurtry et al. 2013). Euseius scutalis can be reared on pollen and was recorded as a predator of Panonychus citri (McGregor) in citrus orchards (McMurtry 1977; Kasap andSekeroglu 2004); it is also reported as a biological control agent of Bemisia tabaci (Gennadius) (Nomikou et al. 2003). Euseius scutalis was observed in high numbers on olive trees in late spring (Chatti et al. 2017). The biology of E. scutalis, however, remains poorly known. This species was recently recorded from Madeira Island by Ferragut and Baumann (2020). Recorded from Maghreb and South of Spain (Kreiter et al. 2004, Ferragut andBaumann 2020), it was probably introduced in the Madeira Island by commercial exchanges with Morocco.
Remarks: this species is very common in Maghreb and South of Spain al. 2004, Ferragut andBaumann 2020). It is also widespread in the Tunisian orchards , Sahraoui et al. 2012.

Genus Iphiseius Berlese
The biological characteristics of this Mediterranean species have been well documented because of its use in controlling thrips on various cultivated plants in greenhouses. Iphiseius degenerans is a commercially available biological control agent of thrips and spider mites in greenhouse crops. It is able to feed on a variety of foods, but thrips' larvae and sweet pepper pollen are unfavourable food for immature development. This could compromise the establishment of this biological control agent when used against thrips in sweet pepper crops. According to the classification by McMurtry et al. (2013), I. degenerans is a typeIV polliniphagous predator.
It is one of the most common native phytoseiid mite species on cassava in south Africa (Zannou et al. 2005) and feeds on M. tanajoa (Nwilene and Nachman 1996), a widely distributed neotropical mite pest of cassava in Africa, insect larvae and pollen of many plants (Vantornhout et al. 2005). Another study concluded that I. degenerans can be considered a suitable biological control candidate based on its preference for Eutetranychus orientalis (Klein) in the Mediterranean region (Fantinou et al. 2012). Iphiseius degenerans preys on Oligonychus perseae Tuttle, Baker and Abbatiello. Although I. degenerans contribution to O. perseae biocontrol can be limited, it needs to be assessed, also taking into account the importance of alternative food source (e.g. Castor oil pollen) for predator population growth (Zappala et al. 2015). This species was already known from Madeira Island, first mentioned by Carmona (1962) and then by Ferragut and Baumann (2020).
World distribution: numerous countries in Northern and Southern Africa, in Mediterranean area (Cyprus, Greece, Italy, Portugal), in Near East or Middle East (Egypt, Israel, Lebanon, Saudi Arabia, Syria, Turkey, Yemen), in Europe (Georgia), in South America (Brazil) and in North America (USA in California, Florida, Georgia, New Hampshire). Also Grande Comore Island (Kreiter et al. 2018b and Kreiter et al. submitted).
Specimens examined: 4 specimens (2 ♀♀, 1 ♂ and 1 imm.) collected during this study. Porto da Cruz, Rum distillery (15 m aasl, 32°48 ′ 18 ″ N, 16°49 ′ 46 ″ W), 2 ♀♀, 1 ♂ and 1 imm. on Acalypha wilkesiana Müller Argoviensis (Euphorbiaceae), 18/V/2019. Remarks: measurements of the 2 ♀♀ and 1 ♂ fit well with measurements of specimens reported in the literature from close countries and with those concerning the specimens previously collected in Grande Comore in Moroni (Kreiter et al. 2018b). This species belongs to the tiliarum species group of the genus Neoseiulella as setae JV3 are present and chelicerae have only few teeth. Neoseiulella canariensis was only known from the Macaronesia Region, from the Canary Islands PeñaEstevez 2003, 2007) from where this species was described on various plants belonging to 15 botanical families and then from Madeira Island (Papadoulis and Kapaxidi 2011). Its biology is totally unknown.
Remarks: morphological and morphometric characters and all measurements of our specimens (Table 2) fit well with measurements of Ferragut and PeñaEstevez (2003) completed by Kanouh et al. (2012), except for the 20% larger dimensions of the genital shield. We have added in the table 2 measurements of some important characters not documented before as it is only the second record of that species from Madeira Island. This species belongs to the bergi species group (Chant and McMurtry 1994). The biology of that species is totally unknown. This is the first report of this species outside the African continent.

Genus Typhlodromus Scheuten
World distribution: Kenya, South Africa. Remarks: morphological and morphometric characters and all measurements of our speci mens (Table 3) fit well with measurements of the original description and other measurements of the literature, especially with those for specimens of South Africa (van der Merwe 1968, Ueckermann et al. 2008. The male of this species is apparently already known and mentioned in the paper of ElBanhawy et al. (2011) but it was not illustrated. Therefore, we do not describe and only measurements of the male specimen are given here (Table 4).

Characters
Grande Comore Island (1) (this study) Typhlodromus (T.) exhilaratus has been wrongly considered as a synonym of T. tiliae Oudemans by Denmark (1992). This species has been mainly reported in the Mediterranean basin. It is quite common in some vineyards in France and Italy (Tixier et al., 2006; Castagnoli et al., 2002. Liguori and Guidi (1990) have shown its ability to feed on the main mite pest in South European vineyards, Eotetranychus carpini (Oudemans). Typhlodromus (T.) exhilaratus is known from Morocco, it was reported by Tixier et al. (2003) from vineyards in the region of Meknes but it is also observed on various plants and locations (Tixier et al. 2016). This is the first record of that species in Madeira Island.

Characters
Specimens examined: 20 specimens (19 ♀♀ and 1 ♂) collected during this study.Canical Remarks: morphological and morphometric characters and all measurements of our specimens fit well with measurements of the redescription of Chant and YoshidaShaul (1987). This species is the third more abundant (17%) after E. stipulatus (22%) and A. herbicolus (21 %) but it is not very frequent (only 10% of the total samples against 31% for E. stipulatus and 31% for A. herbicolus).
Remarks: the measurements of the adult females collected (Table 5) agree with those provided by Denmark (1992) for specimens from Morocco as reported by Tixier et al. (2016). Typhlodromus (T.) setubali and T. (T.) moroccoensis are morphologically close and Tixier et al. (2016) indicated that further analyses would be required to determine if T. (T.) moroccoensis is valid or synonym. But comparisons provided in tables 5 and 6, show some differences especially the number of teeth on the fixed digit and spermatheca shape.

Typhlodromus (Typhlodromus) phialatus Athias-Henriot
Typhlodromus phialatus AthiasHenriot 1960b: 100. Typhlodromus (Typhlodromus) phialatus, Moraes et al. 2004: 366, Chant & McMurtry 2007 This species is mainly known from the Western Palearctic region and is very common in the Mediterranean basin. It has been reported from many plants and some orchards such as vineyards, apple, almond, pear but essentially on citrus (Ferragut et al. 1983; Papaioannou Souliotis et al. 1994; Espinha et al. 1995; Kreiter et al. 2000; Sahraoui et al. 2012. Some studies have been carried out on the biology of this species (Meszaros et al. 2007) and some publications reported that it could limit the development of mite pests in citrus orchards. This species was reported from Morocco by McMurtry and Bounfour (1989) on citrus, weeds and Argania spinosa (L.) Skeels and is present in Canary Island. It was first mentioned by Papadoulis and Kapaxidi (2011) from Madeira Island. The report here confirms the occurrence of this species in the island.
Remarks: the measurements of the adult females collected agree with those provided by Ferragut et al. (2010) for specimens from Spain and by Tixier et al. (2019) for specimens from Sources of measurements -Morocco: Denmark (1992) in Tixier et al. (2016) as proposed by Tixier (2012); -: not provided. France.

Typhlodromus (Typhlodromus) pyri Scheuten
Typhlodromus pyri Scheuten 1857: 104, Moraes et al. 1986: 246. Typhlodromus (Typhlodromus) pyri, Chant 1959. Typhlodromus (Typhlodromus) pyri, Moraes et al. 2004: 367, Chant & McMurtry 2007 This species is cosmopolitan but it is one of the most common and dominant species in vineyards and orchards in the western part of Europe. It has been introduced in various countries such as Australia, New Zealand and the USA for biological control purposes. It has been reported on a wide range of plants, essentially on cultivated and uncultivated shrubs and trees This species is an efficient predator of red and yellow spider mites, and eriophyid mites mainly in orchards and vineyards as well as the grape thrips Drepanothrips reuteri (Uzel) in France (Serrano et al. 2004 Chant & YoshidaShaul 1987). This species is mainly known from the South of the Mediterranean basin, and it is not very common. It has been reported in Portugal and Spain by Dosse (1961) and in Moroccan crops by Tixier et al. (2003Tixier et al. ( , 2016 and Kreiter specifically from Oulmès region (unpub. data). Its biology has been recently studied (Ouassat and Allam 2020 and Allam et al. unpub. data) with specimens collected on apple in Oulmès region of Morocco (Ouassat 2017, Ouassat andAllam 2019). It seems an efficient predator of P. ulmi. It was already observed in Morocco by McMurtry & Bounfour (1989) on Olea europea L., Cupressus sp. and Cynodon dactylon Richard. This is the first record of that species in Madeira Island.
Remarks: morphological and morphometric characters and all measurements of our specimens (Table 6) fit well with measurements of the redescription of Chant and Yoshida Shaul (1987) for specimens from Portugal and measurement of specimens from Morocco (Tixier et al. 2016   Three species previously reported by Carmona (1973) were not found in subsequent studies by Papadoulis and Kapaxidi (2011) and Ferragut and Baumann (2020) as well as in this study. Those species are, Amblyseius largoensis, Euseius hibisci and T. (A.) rhenanus. Similar to the conclusion by Ferragut and Baumann (2020), we think that the report of A. largoensis by Carmona (1973) represents probably a misidentification, though those specimens were not available for study. Just like the former authors, A. herbicolus was the most abundant (18% of all the phytoseiids collected) and frequent (31% of samples) in our material from the Madeira archipelago. As reported in the literature (Döker et al. 2020, Ferragut andBaumann 2020), confusions between A. largoensis and A. herbicolus might be the reason. Three characters allow to distinguish the two species as highlighted by Döker et al. (2020) (2020) highlighted the same hypothesis for specimens from Madeira. The species A. largoensis was not listed in Carmona (1962 and, Papadoulis and Kapaxidi (2011), Ferragut and Baumann (2020) and in the present paper. We thus assumed that specimens of A. herbicolus might also get misidentified as belonging to the species A. largoensis by Carmona (1973) and that the occurrence of the species A. largoensis have to be confirmed. Concerning E. hibisci, Carmona (1973) reported E. hibisci in Madeira Islands, a species usually reported from the American continent. Just like Papadoulis and Kapaxidi (2011), we think that as E. hibisci closely resembles to E. stipulatus, it is possible that the abovementioned record was not E. hibisci but actually E. stipulatus, which is the main Euseius species in Mediterranean region.
And finally, T. (A.) rhenanus which is very close to several species of Typhlodromus (Anthoseius), was often misidentified, and a misidentification might have been also happened in Madeira reports.
On three surveys conducted after the two of Carmona (1962Carmona ( , 1973, none of these three species were recovered.
The Phytoseiidae fauna of Madeira is essentially composed of West Palearctic species with some African species.
Among the 15 species recorded in this study, at least seven species [N. teke, A. herbicolus, E. stipulatus, I. degenerans, T. (T.) exhilaratus, T. (T.) setubali and T. (T.) pyri] are already known as biological control agents (BCAs), some having been experimented and some others sold in several places of the world.
In addition to the intrinsic value of phytoseiid mite biodiversity in such environments, demonstration of the natural occurrence of efficient BCAs in an isolated island such as Madeira Island is of great agricultural, commercial and strategic interests for the country.