Type I lifestyle—Specialized mite predators

Since the first attempt to classify the phytoseiids according to their feeding habits (McMurtry & Croft 1997), the specificity of certain phytoseiid assemblages to mite prey groups other than tetranychids was pointed out by different authors. Thus, Type I lifestyle is now considered to include phytoseiids that are specialized predators of different mite groups. This led to the need to divide this lifestyle type into three subgroups.

Subtype I-a—Specialized predators of Tetranychus (Tetranychidae)

This subtype contains phytoseiids that have adapted to attacking spider mites producing the so-called “complicated web” (CW-u life type of Saito 1985). Until now, the examples refer to phytoseiids associated with Tetranychus species. However, it is possible that these predators can also be effective as control agents of spider mites of other genera, also known to produce the CW-u web type. Saito (1985) mentioned some species of Eotetranychus as also producing that web type.

It corresponds to the species included in the Type I lifestyle of McMurtry & Croft (1997), containing Phytoseiulus species. These predators could have coevolved with Tetranychus species, which have highly clumped distributions (see McMurtry & Croft 1997).

There is a vast literature indicating the association of Phytoseiulus with Tetranychus species under field conditions. Several new studies under laboratory conditions confirm the ability of Phytoseiulus to develop and reproduce when feeding on Tetranychus species (e.g., Pratt et al. 2003; Kazak 2008; Abad-Moyano et al. 2009). Phytoseiulus persimilis (Athias-Henriot), Phytoseiulus fragariae Denmark & Schicha and Phytoseiulus macropilis (Banks) are the most comprehensively studied species. The first species has been extensively used commercially for the control of pest Tetranychus mites, especially Tetranychus urticae Koch, a major pest of many crops in temperate and tropical countries, but its performance seems to be limited on solanaceous plants, apparently because of the action of trichomes present on the aerial parts of those plants (Krips et al. 1999; Skirvin & Fenlon 2001; Sato et al. 2011).

A recent laboratory study has suggested that P. macropilis can be an effective predator of T.urticae on tomato (Sato et al. 2011). This predator may make use of the webbing produced by T.urticae to avoid contact with deterring glandular trichomes of tomato leaves (Sato et al. 2011).

Significant efforts have been dedicated in the last decade to the determination of prospective predators of Tetranychus evansi (Baker & Pritchard), an actual or potential pest of solanaceous plants in African, American, Asian and European countries that produces profuse webbing (Navajas et al. 2013). Despite that effort, only a single species of predator, Phytoseiulus longipes (Evans), was found as potentially useful for the control of that pest, showing preference for T. evansi over T.urticae as prey (Furtado et al. 2006, 2007; Ferrero et al. 2007; Silva et al. 2010).

Subtype I-b—Specialized predators of web-nest producing mites (Tetranychidae)

This subtype contains phytoseiids that have adapted to attacking Schizotetranychus, Stigmaeopsisand some Oligonychus spider mites producing the so-called web-nests (WN-u life type of Saito 1985). These phytoseiids also seem to show co-evolution with their prey (Saito 1990). Typhlodromus (Anthoseius) bambusae (Ehara) has been considered an effective biological control agent of Schizotetranychus celarius (Banks) in China (Zhang et al. 1999). This predator apparently evolved with that prey, having the ability to penetrate its web-nests.

Subtype I-c—Specialized predators of tydeoids (Tydeoidea)

This group was suggested by C. Duso (pers. comm. with JAM, 2009). It comprises representatives of Paraseiulus and Typhlodromina. It may also contain some species of Proprioseiopsis (Momen 2011).

Paraseiulus talbii (Athias-Henriot) is the most extensively studied member of this group.Camporese & Duso (1995) showed that it was closely associated with tydeoid mites in Italy, and that in laboratory studies it had a higher reproductive potential on Tydeus caudatus (Dugès) than on other food items. Of several phytoseiid species inhabiting European vineyards, only P. talbii has been mentioned to feed on tydeoids (Camporese & Duso 1995). Its impact on tydeoid population densities was determined to be moderate, but apparently insufficient to maintain these mites at low population levels in vineyards (Camporese & Duso 1995).

Typhlodromina eharai Muma & Denmark is associated with Tydeus californicus (Banks) on avocado trees in California, USA, on which it has been observed to prey (J.A. McMurtry pers. obs.). Proprioseiopsis cabonus (Schicha & Elshafie) was reported by Momen (2011) in association with the tydeoids Neoapolorryia aegyptiaca (El-Bagoury & Momen), Lorryia aegyptiaca (Rasmy & El-Bagoury) and T. caudatus in soils of Egypt. In a laboratory study, Momen (2011) showed the ability of P. cabonus to feed, develop and reproduce efficiently on N. aegyptiaca and L. aegyptiaca, but not on pollen of Ricinus communis or Phoenix dactylifera, reported to be suitable for laboratory rearing (Abou-Setta et al. 1997) of Proprioseiopsis cannaensis (Muma) and Proprioseiopsis rotundus(Muma). Life table parameters of a Brazilian population of Proprioseiopsis cannaensis were calculated by Bellini et al. (2010) on four different food items, including species of Eriophyidae, Tenuipalpidae and Tetranychidae, as well as Typha angustifolia pollen. In that paper, the authors reported difficulties in maintaining a colony of this predator on a mixture of Typha angustifoliapollen and all stages of the mite Tyrophagus putrescentiae (Schrank) (Acaridae). The results suggested that neither of those items were adequate as food sources to the predator. Laboratory studies are needed to assess the possible preference of P. cannaensis for tydeoids as food sources. The suitability of tydeoids as prey for some phytoseiids has been known for a long time. Knop & Hoy (1983) reported Homeopronematus anconai (Baker) as an important prey to maintain the population of Galendromus occidentalis (Nesbitt) under field conditions, when the population of Tetranychus pacificus (McGregor), a major pest of grape, was low. However, G. occidentalis should not be included in this group, given that it is not a specific predator of H. anconai.

Type II lifestyle—Selective predators of tetranychid mites

This category includes selective predators of tetranychid mites, most often associated with dense web producing species, such as Oligonychus and Tetranychus species. It includes species of Neoseiulus, Galendromus and apparently, the rickeri group of Typhlodromus (Anthoseius). These are associated with Tetranychus species (CW-u life type) or species of other genera producing other types of webbing (WN-u web and WN-c life types; see McMurtry & Croft 1997). Unlike Type I species, members of Type II have a preference for a broad range of tetranychid species, but they also feed and reproduce on mites of other groups, including Eriophyidae, Tarsonemidae and Tydeoidea, and also on pollen.

It has been questioned (Croft et al. 1998a) whether Neoseiulus californicus (McGregor), one of the main phytoseiids placed by McMurtry & Croft (1997) in this group, should not be better classified as a member of Type III group. However, it is kept here in the Type II group because it is nearly always associated with tetranychids producing heavy webbing; Type III phytoseiids do not do well on mite prey of this type, often getting stuck in their webbing. Neoseiulus californicus shows adaptations for living in spider mite colonies with heavy webbing; it has the ability to cut strands of webbing with the chelicerae (Shimoda et al. 2009) and to use the front legs to tear holes in web-nests of Oligonychus perseae (Tuttle, Baker & Abatiello) (Montserrat et al. 2008).

Because of the widely known efficiency of N. californicus as a control agent of tetranychid mites, many new publications have dealt with different aspects of its biology, ecology and practical use to control tetranychid mites (i.e, Jolly 2000; Zalom 2002; Nancy et al. 2005; Pringle & Heunis 2006; Walzer et al. 2007; Abad-Moyano et al. 2009; Elmoghzy et al. 2011; Toldi et al. 2013). It has also been evaluated as a predator of T. evansi, with poor performance (Escudero & Ferragut 2005). Neoseiulus californicus has been used for the biological control of O. perseae in the USA (Hoddle et al. 1999, Hoddle et al. 2000), and was recommended for the control of this pest in Spain (Montserrat et al. 2008). This predator has also been mentioned as potentially effective for the control of Phytonemus pallidus (Banks) on strawberry (Easterbrook et al. 2001) and Polyphagotarsonemus latus (Banks) on pepper (Jovicich et al. 2008), both Tarsonemidae, and for the control of thrips on pepper (Van Baal et al. 2007, Rahman et al. 2009).

Several studies on the closely related N. fallacis were also conducted in recent years. This phytoseiid was reported to feed on a wide range of mite pests of ornamentals (Pratt & Croft 2000). Minimum inoculations of this species were reported to provide successful control of T. urticae on strawberries in the USA (Croft & Coop 1998). In field experiments, N. fallacis was reported to be effective in controlling T. urticae on Malus rootstocks or Acer shade trees, as well as Oligonychus ilicis (McGregor) on Rhododendron (Pratt & Croft 2000). Releases of N. fallacis provided effective control of T. urticae on apple seedlings in Oregon (Croft et al. 2004). It was also reported as effective for management of Schizotetranychus longus (Saito) on bamboo in the USA (Pratt & Croft 1999). Neoseiulus fallacis has been reported to be able to control P. pallidus on strawberry (Croft et al.1998b), and to develop and reproduce adequately when offered Aculops lycopersici as prey (Brodeur et al. 1997).

As observed for N. californicus in the American and European continents, Neoseiulus longispinosus (Evans) has received extensive attention in Asia, according to Nusartlert et al. (2010), for the control of spider mites. These authors reported its development on species of the tetranychids Eutetranychus, Oligonychus and Tetranychus. This range of prey includes species that web profusely and species that produce little webbing. They did not observe the development of this predator on Oligonychus mangiferus (Rhaman & Sapra), a species that produces little webbing. Yet, the potential of this predator to control Oligonychus coffeae (Nietner), a serious pest of tea plantations that also produces little webbing, was reported by Vattakandy et al. (2013). Neoseiulus longispinosus was also observed to develop and reproduce on Eotetranychus cendanai (Rimando), so it might be used to control this spider mite pest in greenhouses (Tongtab et al. 2001).Works by Zhang et al. (1999)showed the potential of this predator as a biological control agent of Stigmaeopsis nanjingensis (Ma & Yuan) on bamboo in China. Neoseiulus longispinosus has also been observed to feed on the tenuipalpid Raoiella indica (Hirst), a serious pest of coconut, especially in the Caribbean area (Carrillo et al. 2012).

The preference of Neoseiulus womersleyi (Schicha) for Amphitetranychus viennensis (Zacher) over T. urticae was reported by Furuichi et al. (2005). The authors attributed this difference to the more favorable web produced by the former prey. The potential of N. womersleyi to control the spider mite Tetranychus macfarlanei (Baker & Pritchard) was recently evaluated under laboratory conditions by Ali et al. (2011). Reduced predation of Tetranychus kanzawai (Kishida) by N. womersleyi was attributed by Oku (2008) to the presence of the yellow excreted pellets deposited by the spider mite prey on its own web.

Galendromus occidentalis was found to be effective in the control of O. perseae, whereas the opposite was observed for Galendromus annectens (DeLeon) (Hoddle et al. 1999). Galendromus helveolus (Chant) was shown to be one of the most promising predators for that purpose, as it readily entered the nests of O. perseae to feed on the prey (Takano-Lee & Hoddle 2002).

Some Neoseiulus and Galendromus species have been found in association with web-net producing spider mites. However, they are not specific predators of those mites, and for that reason they are placed in Type II lifestyle, rather than in subtype I-b. Neoseiulus fallacis has shown some potential in controlling S. longus in the United States (Pratt & Croft 1999), being able to successfully invade the nests of this species, at least up to ten days after the beginning of nest building process by the mite. Zhang et al. (1999) showed the ability of N. longispinosus to consume S. nanjingensis in short duration laboratory experiments, but the authors pointed out that this predator is rarely found on bamboo, the host plant on which S. nanjingensis should be controlled. Neoseiulus cucumeris(Oudemans) has also been shown to feed and develop when offered S. nanjingensis as prey under laboratory trials (Zhang et al. 2000). However, they are only able to enter broken web-nests. Neoseiulus californicus and G. helveolus have been reported to provide effective control of O.perseae (Tuttle, Baker & Abatiello), also a web-nest producing mite (Kerguelen & Hoddle 1999; Takano-Lee & Hoddle 2002).

Type III lifestyle—Generalist predators

Species in this category feed and reproduce on a wide range of prey (at least under laboratory conditions), including mites of families belonging to Astigmata such as Acaridae and Pyroglyphidae, and Prostigmata, particularly Eriophyidae, Tarsonemidae, Tetranychidae, Tenuipalpidae and Tydeidae, as well as thrips, whiteflies, mealybugs, nematodes, etc. Many species in Type III have been shown to be able to feed and reproduce well on pollen. They can utilize plant exudates and honeydew as survival food in the absence of prey, or as complementary food, which increase their reproductive capacity when prey is present (McMurtry & Croft 1997).

In addition, several phytoseiids of this group have the ability to feed on fungi. Association or actual development and/or reproduction have been reported for the following phytoseiids on different plant pathogenic fungi: Typhlodromus (Typhlodromus) pyri (Scheuten) by Chant (1959), Eichhorn & Hoos (1990), Zemek & Prenerova (1997) and Pozzebon & Duso (2008); Ricoseius loxocheles De Leon by Flechtmann (1976) and Oliveira (2012); Amblydromalus manihoti (Moraes)(mentioned as Typhlodromalus limonicus s.l.) by Bakker & Klein (1992); and Amblyseius andersoni(Chant) by Pozzebon & Duso (2008). Some phytoseiids have been found to consume the microcrustacean species Artemia franciscana (Kelogg) (Artemiidae) as factitious food (Vangansbeke et al. 2013). Amblydromalus limonicus (Garman & McGregor) and Amblyseius swirskii Athias-Henriot have been reported to feed and maintain themselves on those organisms when the latter are placed on plant leaves (Audenaert et al. 2013).

The subsequent grouping of these mites in subtypes is based on their preferred microhabitats.

Subtype III-a—Generalist predators living on pubescent leaves

Species of Paraphytoseius, Phytoseius as well as some Kampimodromus, Typhlodromalus and Typhlodromus species are commonly found on pubescent leaves (leaves with trichomes). The typically small and laterally compressed idiosoma of these species apparently aids them in moving between leaf trichomes (Duso 1992; Walter 1992; Karban et al. 1995; Kreiter et al. 2002; Kreiter et al. 2003; Tixier et al. 2007). Those phytoseiids characteristically have some of the dorsal shield setae stout and usually serrate. These morphological characteristics enable those mites to colonize microhabitats not occupied by larger phytoseiids, avoiding possible competition and escaping predation by the latter (Seelman et al. 2007), while taking advantage of the presence of prey which also prefer the same microhabitat.

Also, the elongate gnathosoma of Eharius phytoseiids seems to be an adaptation to aid them in reaching the leaf surface from the trichomes of Marrubium vulgare foliage. Eharius chergui (Athias-Henriot) has been observed on this plant in the apparent absence of any prey or pollen grains, possibly piercing leaf cells (D. Mahr pers. comm.).

Trichome shapes and structures are widely variable (Metcalfe & Chalk 1979). Some authors have pointed out their deleterious effect on some phytoseiid species, whereas others have reported the opposite (Sato et al. 2011). While the non-glandular trichomes conceivably affect the phytoseiids and other mites physically, glandular types may also cause chemical interference with them. Walter (1996) conducted a thorough revision about the effect of different types of trichomes on mites. The author referred to several studies showing that plants with pubescent leaves had a higher diversity and density of predatory mites than plants with glabrous leaves, which was especially true for the small phytoseiid species. Different authors mentioned by Walter (1996) reported the following phytoseiids to predominate on pubescent leaves: Euseius victoriensis Womersley, G. occidentalis, Kampimodromus aberrans Oudemans, Phytoseius species, Typhlodromus (Anthoseius) caudiglansSchuster and T. (T.) pyri. As explained elsewhere in this text, the first and the second species are not classified as Type III lifestyle. Kampimodromus aberrans has been widely known to occur mainly on plants with pubescent leaves (Kreiter et al. 2002; Kabicek 2008). Kreiter et al. (2003) observed that leaf characteristics (trichomes, domatia) as well as pollen densities and plant species influenced greatly the occurrence and abundance of this predator. Additional supporting information has been published about the importance of K. aberrans as a biological control agent of spider mites on some grape varieties in Europe, especially in France and Italy (Duso et al. 2009; Kreiter et al. 2000;Ozman-Sullivan 2006). It has also been reported from hazelnut orchards in Turkey in association with the eriophyoid mites Phytoptus avellanae (Nalepa) and Cecidophyopsis vermiformis (Nalepa)(Ozman-Sullivan 2006). In a laboratory study, these authors reported the ability of K. aberrans to develop and reproduce on P. avellanae. This predator has also been reported to prey, develop and reproduce on Cecidophyopsis vitis (Nalepa) (Schausberger 1992). Higher density of T. (T.) pyri was reported by Duso et al. (2003) on an apple variety with highly pubescent leaves as compared to varieties with less pubescent leaves. The preference of Typhlodromalus aripo De Leon for cassava varieties with pubescent tips over varieties with glabrous tips was reported by Onzo et al. (2012).

Many studies have been published about the association of Phytoseius species with tetranychid, eriophyid, tarsonemid and tenuipalpid mites (e.g., El-Laithy & Fouly 1998; Tixier et al. 1998; Mailloux et al. 2010; Vassiliou et al. 2012). Differently from what was reported by McMurtry & Croft (2007) in relation to the potential of species of this genus as biological control agents, a species reported as Phytoseius finitimus (Ribaga) was mentioned by Duso & Vettorazzo (1999) as potentially useful for the control of Panonychus ulmi Koch on grape. In laboratory experiments,Pappas et al. (2013) observed phytoseiids identified as P. finitimus to feed and reproduce on larvae of T. urticae, as well as on crawlers of the greenhouse whitefly, Trialeurodes vaporariorum Westwood. Populations from Iran identified as Phytoseius plumifer (Canestrini & Fanzago) have been mentioned as predators of tetranychid and eriophyid mites on various crops (Hajizadeh et al.2002; Kodayari et al. 2013; Nadimi et al. 2009).

Subtype III-b—Generalist predators living on glabrous leaves

This is probably the most diverse subgroup. It seems to contain most of the species of large genera, e.g. Amblyseius and Neoseiulus, but also species of smaller genera, such as Amblydromalus. This subtype contains phytoseiids that are not particularly small.

Amblyseius swirskii has been extensively used for thrips and whitefly control in greenhouses (e.g., Bolckmans et al. 2005; Messelink et al. 2005, 2006; Kutuk et al. 2011). It can be mass produced at relatively low cost on acarid mites. Provision of pollen to this predator in the field may enhance its efficiency in controlling whiteflies (Nomikou et al. 2010; Kutuk & Yigit 2011). This predator was mentioned by van Maanen et al. (2010) as useful for the control of P. latus.

Amblydromalus limonicus is often found on the glabrous leaves of citrus and other plant species (Moraes et al. 1986), whereas the closely related A. manihoti is a common predator on cassava plants in South America (Yaninek et al. 1998). The latter species is found mostly on leaves of the median section of cassava plants, where leaf pubescence is less than those of the apical section (Bonato et al. 1999). In the last few years, A. limonicus has been commercialized for the control of whiteflies and thrips in protected crops, following the discovery of successful off-plant rearing on factitious prey. Amblydromalus manihoti was reported (as Amblyseius limonicus s.l.) to feed on thrips (Noronha & Cobo 1990) and whiteflies (Noronha & Moraes 1992). Amblydromalus lailae (Schicha) was reported as potentially useful for the biological control of thrips, A. lycopersici and P. latus in Australia (Steiner et al. 2003b).

Amblyseius andersoni was observed to be abundant and more effective as a predator of spider mites on grape varieties with more glabrous leaves (Camporese & Duso 1995; Duso et al. 2003). Studies conducted under laboratory and greenhouse conditions also indicated the potential of Transeius montdorensis Schicha to control A. lycopersici, P. latus and thrips (Steiner et al. 2003a). It can be reared exclusively on Typha pollen. Given the successful adoption of off-plant rearing methods using factitious prey, this predator is now produced commercially in Europe and Australia mainly for thrips and whitefly control in greenhouses.

Studies in China (Wu et al. 2010) have shown Amblyseius eharai Amitai & Swirski and Scapulaseius newsami Evans to be effective natural enemies of Panonychus citri (McGregor). The potential of the former species to control thrips has been reported by Kakimoto et al. (2004).

Subtype III-c—Generalist predators living in confined space on dicotyledonous plants

Examples are phytoseiids (Walter 1996) on galled leaves of willow and poplar trees. Gall-forming eriophyids are often present as possible food sources, but so are spider mites (JAM, pers. obs.). Thus, these phytoseiids are apparently distinguished by the type of microhabitat they favour, represented by the galled leaf surface, rather than by the available prey type. They are elongate but not particularly small, comprising mostly the desertus group of Neoseiulus (Prischmann et al. 2005). Little is known about the potential of species of the desertus group of Neoseiulus as biological control agents of gall-forming eriophyid mites on willow and poplar trees.

Domatia are structures of different constitutions always corresponding to a confined space on leaves of some plants (O´Dowd & Willson 1989; Walter 1996). They have been reported on over 2000 plant species of nearly 300 families (Brouwer & Clifford 1990). As previously mentioned, domatia favor the survival and development of several mites, including phytoseiids. Pemberton & Turner (1989) hypothesized the existence of a widespread facultative mutualism between plants with leaf domatia and beneficial mites. The following phytoseiids have been noted to occur in domatia: Amblyseius herbicolus (Chant), Euseius hibisci (Chant), Galendromus longipilus (Nesbitt), Phytoseius hawaiiensis Prasad, Typhlodromus (Anthoseius) haramotoi Prasad and Typhlodromus (Anthoseius) rhenanoides (Athias-Henriot), by Pemberton & Turner (1989); Euseius elinae (Schicha) and Typhlodromus (Anthoseius) dossei Schicha, by Walter (1992); Iphiseius degenerans (Berlese), by Faraji et al. (2002) ; Iphiseiodes zuluagai Denmark & Muma, by Matos et al. (2004) ; Typhlodromus (Anthoseius) doreenae Schicha, by Rowles & O´Dowd (2009).

Although several other phytoseiids have been noted to occur in domatia, it does not seem appropriate to include them as part of this subtype because they are also often found on leaves (either pubescent or glabrous) of plants that do not have domatia. Thus, it seems that they may benefit from the presence of domatia, but that domatia are not essential for them (Ferreira et al. 2008). For example, I. zuluagai is one of the common predatory mites found in domatia of coffee leaves in Brazil (Matos et al. 2004). However, it has also been commonly reported on the glabrous leaves of several plant species, including citrus (Albuquerque & Moraes 2008) and other plant species that do not have domatia (Moraes et al. 1986). Thus, we do not consider it appropriate to include this species as a member of subtype III-c.

Similarly, T. aripo has been reported to spend the day between the unfolded leaves within the growing tips of cassava plants (Onzo et al. 2005, 2009). Growing tips of cassava could be considered as a confined space (Onzo et al. 2005), but T. aripo has also been reported from leaves of several plant species (Moraes et al. 1986). Even on cassava, this predator is found at night wandering on the upper leaves searching for its preferred prey, Mononychellus tanajoa (Bondar) (Tetranychidae). Thus, here it is considered as belonging to subtype III-a, given its preference for plants with pubescent tips (Onzo et al. 2012). The preference of K. aberrans and Neoseiulella tiliarum(Oudemans) to inhabit pubescent domatia has been reported (Kreiter et al. 2000, 2003; Kabicek 2008). However, these mites are commonly found on the surface of pubescent leaves, and are for this reason classified in subtype III-a.

Subtype III-d—Generalist predators living in confined spaces on monocotyledonous plants

This includes phytoseiids adapted to living in restricted spaces on monocotyledonous plants, between leaf sheaths or between bracts and subjacent fruit surfaces. The best known case refers to phytoseiids of the paspalivorus group of Neoseiulus, including Neoseiulus baraki (Athias-Henriot), Neoseiulus neobaraki (Zannou, Moraes & Oliveira) and Neoseiulus paspalivorus (DeLeon). These species are small, flat, elongate, and short-legged, which allows them to move into the tiny spaces between the tightly appressed bracts and the subjacent fruit surface of coconut, in search of their eriophyid prey, Aceria guerreronis (Keifer) (Moraes et al. 2004).

Considerable effort was dedicated in the last decade to the development of a program for the biological control of A. guerreronis, a major pest of coconut. In this regard, several studies have been conducted in tropical countries of the American, African and Asian continents to search for prospective biological control agents. Neoseiulus baraki, N. paspalivorus and N. neobaraki have been found in association with A. guerreronis (Moraes et al. 2004; Lawson-Balagbo et al. 2008; Negloh et al. 2008; Navia et al. 2013). The first two species were most common in Africa, South America and Asia, while the last was only recorded in Africa. Their ability to feed and reproduce on A. guerreronis has been shown in studies conducted under laboratory conditions (Lawson-Balagbo et al. 2007; Negloh et al., 2008). They have also been reported to develop and reproduce on T.putrescentiae, occasionally found in the same microhabitat (Moraes et al. 2004; Lawson-Balagbo et al. 2007). Neoseiulus paspalivorus was also shown to develop and reproduce on Steneotarsonemus furcatus (De Leon) and coconut pollen (Lawson-Balagbo et al. 2007), whereas N. baraki had poor development and reproduction when fed with Steneotarsonemus concavoscutum Lofego & Gondim Jr. as well as coconut pollen (Negloh et al. 2008; Domingos et al. 2012). Neoseiulus baraki and N. paspalivorus were originally described from grasses, on which they were probably occupying the leaf sheaths, as suggested by the confined microhabitat they occupy on coconuts.

Subtype III-e—Generalist predator from soil/litter habitats

Some species of this group periodically move up onto low-growing plants. These correspond to a diverse assemblage of species that may have secondarily moved from the aerial to the ground habitat (McMurtry 2010). It includes many Neoseiulus and Arrenoseius, some Amblyseius and most Proprioseiopsis, Chelaseius and Graminaseius (e.g., Salamane & Patrova 2002; Amano et al. 2011). The last three genera could have derived from Amblyseius. The Neoseiulus probably evolved in a separate event and they do not seem to be morphologically distinguishable as a group from the arboreal species.

Two species showing this lifestyle, Neoseiulus barkeri (Hughes) and N. cucumeris, have been extensively used commercially for the biological control of pest species, especially thrips (Gerson et al. 2003). Neoseiulus barkeri has also been used for the biological control of P. latus in protected crops (Fan & Petitt 1994). Laboratory and greenhouse experiments conducted by Messelink et al.(2012) have shown this predator to be potentially useful for the control of Steneotarsonemus laticeps(Halbert), a serious pest of amaryllis in Europe. Its ability to control the western flower thrips, Frankliniella occidentalis (Pergande), has also been suggested (Fan & Petitt 1994). Neoseiulus cucumeris is also known to feed on P. latus, P. pallidus, T. urticae and fungi (Gerson et al. 2003; Weintraub et al. 2003). Both N. barkeri and N. cucumeris are commercially produced using acarid mites as prey.

Species of Proprioseiopsis are commonly found in the soil/litter habitats. Despite being so common in different parts of the world (Moraes et al. 2004), very little has been reported about the biology of mites of this group. The few papers dealing with the biology of these mites refer to their ability to consume mites found on the aerial plant parts as prey (Ball 1980; Meshkov 1996; Fouly 1997; Abou-Setta et al. 1997; Momen 1999, 2009; Momen & El-Borolossy 1999; Navasero & Corpuz-Raros 2005).

Type IV lifestyle—Pollen feeding generalist predators

This category contains phytoseiid predators for which pollen constitute an important part of the diet. It includes the genera Euseius, with nearly 212 species (Demite et al. 2012), Iphiseius and Iphiseiodes (Reis & Alves 1997; Villanueva & Childers 2007), with a single and a few known species, respectively. These species generally have high reproductive capacity when feeding on pollen, and population increases often follow blossoming periods of the crop or adjacent plants.

Many of these mites prefer glabrous leaves, as inferred from an analysis of the plants on which they have been found (see Moraes et al. 1986). However, leaves of some of these plants are covered by trichomes over certain regions. Euseius finlandicus (Oudemans) has been reported mainly on glabrous substrates (Tuovinen 1994; Kabicek 2005, 2008). Kabicek (2008) reported the common occurrence of this species on the glabrous regions of the leaves of Corylus avellana, moving to pubescent patches when disturbed, where they are more protected. This predator is known to feed on eriophyoid and tetranychid species (Schausberger 1998; Broufas & Koveos 2000; Awad et al. 2001). Abundance of this predator was correlated with the presence of pollen on hedgerows adjacent to vineyards in northern Italy (Duso et al., 2004). A positive effect of pollen was also reported by Broufas & Koveos (2000) in some fruit tree orchards. Euseius victoriensis has been mentioned as a remarkable predator in Australia, playing an important role in the control of eriophyids (Smith & Papacek 1991), T. urticae as well as of P. latus (James 2001).

Field observations in Saudi Arabia indicated Euseius species, particularly Euseius scutalis(Athias-Henriot), to be important as predators of tetranychid mites, scale insects and whiteflies (Nomikou et al. 2001; Raza et al. 2005). A recent study conducted by Al-Shammery (2010) showed E. scutalis to be a promising control agent of those pests. Euseius hibisci is of some importance to management of the spider mite Oligonychus punicae (Hirst) in avocado orchards in California (McMurtry 1985), as is E. scutalis of O. perseae in avocado orchards in Israel (Maoz et al. 2011). In a study conducted by Abad-Moyano et al. (2010) in Spain, Euseius stipulatus (Athias-Henriot) was found to adversely affect the establishment of N. californicus and P. persimilis, thus negatively affecting control of T. urticae in Spanish Clementine orchards.

Iphiseius degenerans is a common predator on cassava plants in Africa (Yaninek et al. 1998; Zannou et al. 2005). This species was reported to feed and oviposit when diet consisted of pollen, T.urticae, larvae of the thrips F. occidentalis or eggs of Ephestia kuehniella (Zeller) (Vantornhout et al. 2005). Following the determination by van Houten et al. (2005) of its potential as a predator of F. occidentalis, this predator is now commercialized for thrips control.

Iphiseiodes quadripilis (Banks) could not complete its development when fed Phyllocoptruta oleivora (Ashmead) under laboratory conditions, while it could complete its development and reproduce on ice plant or oak pollens, or on Eutetranychus banksi (McGregor) and P. citri(Villanueva & Childers 2007). These results corroborate previous observations indicating availability of pollen on citrus leaves to be correlated with the abundance of I. quadripilis (Villanueva & Childers 2004). Iphiseiodes zuluagai is one of the most common phytoseiids in Brazil and a few other south American countries (Demite et al. 2012). Independently of its preference for pollen, I. zuluagai has also been mentioned as an important predator of Brevipalpus phoenicis(Geijskes) in Brazil, and to be able to develop and reproduce when offered eggs of T. putrescentiaeas factitious food (Albuquerque & Moraes 2008).

Some phytoseiids placed in this group have been shown to be able to pierce leaf cells. This ability will be discussed in the following section of this publication.

Aquatic phytoseiids

A distinct group of phytoseiids consisting of a single species in each of the genera Evansoseius and Macrocaudus (Neoseiulini) may constitute a new type (or subtype) of lifestyle. These mites live on floating plants and have similar primitive morphological phytoseiid features, especially the presence of some setae of uncommon occurrence on the posterior half of the dorsal shield. Otherwise they share common features with other Neoseiulini, a tribe considered by Chant & McMurtry (2003) to consist of primitive Amblyseiinae. Both genera were described from South America, and neither has been reported after the original description. Nothing is known about their biology. Macrocaudusmultisetatus Moraes, McMurtry & Mineiro has been occasionally observed from its type locality and type host plant (unpublished observation). The only obvious arthropods found in association with this species were larvae of an unidentified species of chironomid fly. Attempts to rear that phytoseiid on T. putrescentiae, Typha pollen or the nematode Rhabditella axei (Cobbold) (Rhabditidae) as factitious food sources have not been successful (unpublished observation).

Phytoseiids with plant cell piercing abilities

A “cross type” was suggested by Adar et al. (2012) to contain phytoseiids that apparently complement their nutrition requirements by feeding on leaf cells. Although evidence suggests their ability to pierce leaf cells, the primary reason for this type of behaviour is still unknown. Given that the limiting factor for any organism is usually the availability of water, it seems that leaf cell piercing may be related to water uptake (from within cell cytoplasm), the uptake of nutrients being a consequence of their presence in the imbibed liquid.

This potential group contains mostly Euseius mites, but also species of Kampimodromus, Phytoseius, Typhlodromalus, Typhlodromus (Anthoseius), Typhlodromus (Typhlodromus) and probably some Amblydromalus. It is expected that many of the species included in this group are also pollen feeders. However, research suggested that some pollen feeders were unable to take up liquid from at least some of the plant species on which they were found (Porres et al. 1975; Adar et al.2012).

Uptake of plant liquid by phytoseiid mites was first reported by Chant (1959) from laboratory experiments with E. finlandicus, Typhlodromus (Anthoseius) rhenanus (Oudemans) and T. (T.) pyri. It was later suggested by Porres et al. (1975) and Congdon & Tanigoshi (1983) for E. hibisci, Congdon & Tanigoshi (1983) for Euseius tularensis (Congdon), Kreiter et al. (2002) for K. aberrans, Magalhães & Bakker (2002) for T. aripo and Nomikou et al. (2003) for E. scutalis. Field studies by Grafton-Cardwell & Ouyang (1995, 1996) also provided indications that E. tularensis could consume the content of citrus leaf cells. Adar et al. (2012) conducted a detailed evaluation of the ability of E. scutalis and A. swirskii to pierce leaf cells, determining that the former was able to extract liquid from leaf cells but that the latter was not, although it ingested free water on the surface of the substrate. Nomikou et al. (2003) also did not find evidence of liquid uptake from leaves by A.swirskii. Adar et al. (2012) evaluated the cheliceral morphology of several phytoseiid species, including species reported and not reported as plant piercers, concluding that plant piercing phytoseiids had movable cheliceral digit shorter than fixed cheliceral digit and the latter less curved than the non-plant piercers.

Despite the ability of some phytoseiids to take up liquid from plants, economic damage by those mites has not been reported. Feeding scars have been mentioned by Sengonca et al. (2004) to be caused by T. (T.) pyri on apple leaves and fruits, but even in total absence of other food items, the damage caused by those mites was minor, represented by few small scars. In addition to the reduced ability of each individual to cause damage, phytoseiid populations rarely attain high levels on plants under natural conditions. High phytoseiid population densities usually only occur when food is abundant, and that circumstance is less conducive to piercing of plant cells by phytoseiids (Sengonca et al. 2004).

Some authors have discussed the biological importance of plant piercing by phytoseiids. Grafton-Cardwell & Ouyang (1995) reported that pruning of citrus trees resulted in higher E.tularensis population densities. In a subsequent paper, the same authors (Grafton-Cardwell & Ouyang 1996) observed a positive correlation between the number of eggs laid by E. tularenis and the concentration of fertilizer supplied to the plants. These observations do not preclude the possibility of the effect being indirect, resulting basically from the effect of better plant nutrition on the prey. Magalhães & Bakker (2002) concluded that T. aripo feeds in cells of cassava leaves, suggesting that this habit could increase the survival of the phytoseiid in the absence of its main food source (M. tanajoa) on cassava. Gnanvossou et al. (2005) observed that males of this species were able to complete immature development when confined on a cassava leaf disk without any prey. It has often been observed under laboratory conditions that some phytoseiid species seem unable to develop and reproduce on non-leaf substrates, as reported by McMurtry & Scriven (1965) for E.hisbisci and by Bakker & Klein (1992) for A. manihoti. However, it is not known whether they really depend on plants for their survival, or whether other unaccounted factors could be involved. Omnivory is widespread in the animal kingdom, but many predators (canines, for example) are known to directly ingest plant material only under abnormal circumstances, although plant material is often taken by consuming the guts of their prey.