PHYTOSEIID MITE DIVERSITY (ACARI: MESOSTIGMATA) AND ASSESSMENT OF THEIR SPATIAL DISTRIBUTION IN FRENCH APPLE ORCHARDS

The present study reports the results of surveys carried out over two years in 173 apple orchards in France. Eleven species of Phytoseiidae were observed, among them three were dominant: Amblyseius andersoni, Kampimodromus aberrans and Typhlodromus (Typhlodromus) pyri. Cydnodromus californicus was also found but only in some orchards and nearly always in association with one of three dominant species. This observation confirms the faunal modification initiated more than ten years ago. Amblyseius andersoni was recorded in high densities in nearly all the regions considered. Typhlodromus (T.) pyri was also widespread, even if particularly frequent and abundant in the Rhone-Alpes region. Kampimodromus aberrans was localised in some regions; it was especially frequent and abundant in the Mediterranean region. An identification key containing the eleven reported species is provided. Taylor’s law was applied in order to characterize the Phytoseiidae distribution in apple orchards. The distribution is clearly aggregative, whatever the region and the Phytoseiidae species considered. Relationships between the occupation rate and the mean number of Phytoseiidae per leaf was established and an abacus was constructed to facilitate surveys and the counting during practical assessments of Phytoseiidae fauna in apple orchards.


INTRODUCTION
The mite family Phytoseiidae comprises several predator species which can effectively reduce densities of mite pests in various agrosystems (McMurtry and Croft 1997;Gerson et al. 2003). Most species of this family are generalist predators; they can feed on their prey (especially of the families Tetranychidae and Eriophyidae) but can also develop feeding on pollen, plant exudates, fungi and small insects (McMurtry and Croft 1997;Kreiter et al. 2005).
The family presently comprises more than 2,000 valid species widespread in the world (Moraes et al. 2004;Tixier et al. 2012). In France, more than one hundred species have been reported, especially in crops. However, their occurrence has not been equally studied in all agrosystems. The most wellknown French acarofauna is that of vineyards (i.e. Kreiter et al. 2000Kreiter et al. , 2002Tixier et al. 2002Tixier et al. , 2005Barbar et al. 2006). However, despite the economic importance of apple orchards, little is known on the French Phytoseiidae fauna in this crop. Few studies carried out more than ten years ago (Fauvel and Gendrier 1992;Fauvel et al. 1993;Bourgouin et al. 2002) showed the presence of Typhlodromus (Typhlodromus) pyri Scheuten, Amblyseius andersoni Chant, Cydnodromus californicus (McGregor) and Kampimodromus aberrans (Oudemans). However, these species were differently distributed depending on the regions considered. Furthermore, the augmentation of some species as A. andersoni in south-western France and K. aberrans in southeastern was observed in the last survey in 2002.
In other neighbouring countries as Italy, Spain and Switzerland, these same species are observed in apple orchards (Linder 2001;Miñarro et al. 2002;Iglesia et al. 2005;Duso et al. 2009).
The present work presents the results of surveys carried out in 173 apple orchards in several regions of France. The first aim of this survey was thus to determine the Phytoseiidae species encountered. For each species, a discussion on its biology, its occurrence in apple orchards sampled and the factors that could affect its presence is provided. An identification key to females is proposed to assist Phytoseiidae species diagnosis in French apple orchards.
The second aim was to determine the spatial distribution of Phytoseiidae into orchards. According to this spatial distribution, the third aim was to provide an optimal sampling method to assess their densities for practical and accurate counting.

Sampling and species identification
One hundred and seventy-three apple orchards (planted with 24 different cultivars) in thirteen regions were sampled from May to September in 2011 and 2012 (Table 1, Figure 1). The number of orchards surveyed in each region was similar. Only one or two orchards have been sampled in Alsace, Centre, Haute-Normandie, Franche-Comté, Limousin and Poitou-Charentes. Thus for these regions, no conclusion would be drawn on the Phytoseiidae fauna.
Fifty leaves were collected in each plot. Each leaf was individually observed; Phytoseiidae were counted and collected for identification with a fine hair-brush. Then the females were mounted on slides in Hoyer's medium and identified under 98   (Tixier et al. 2008a, b;Tixier 2012;Akashi et al. 2012). The nomenclature used was that proposed by Chant and McMurtry (2007) in the last revision of the family except for the species Neoseiulus californicus, recently re-associated to the genus Cydnodromus (Tsolakis et al. 2012).
All Phytoseiidae found in apple orchard are generalist predators. Conversely to specific predators, their occurrence is usually not linked to their prey and no correlation is usually observed between densities of Phytoseiidae and preys (Slone and Croft 2001). Thus, species of Tetranychidae were not identified nor counted. Furthermore, their densities were usually very low.

Phytoseiidae distribution in apple orchards
In order to characterize Phytoseiidae distribution at the plot level, Taylor's law has been used (Taylor 1961). This law relates the standard deviation (S 2 ) to the mean (m) according to the following relation: S 2 = am b . When applying a log transformation, the latter relation describes a straight line (logS 2 = log(a) + b x log(m)), where b is the slope. The b value of this relation provides information on distribution: when b = 1, the species is randomly distributed, when b > 1 the distribution is aggregative and when b < 1, the distribution is regular. To establish such a relation and calculate the b value, the mean and the standard deviation of each plot have been calculated (and log transformed) and a simple correlation test has been applied (Statsoft 2008).

Characterisation of the sample size for optimal sampling
In order to define the number of leaves to be sampled in further surveys for characterising the number of Phytoseiidae in apple orchards (N), the following relation (Nachmann 1984) was applied: N = am (b-2) /E 2 where a and b are Taylor's law variables, m the mean, and E 2 the accepted error around the mean. We herein tested two errors : 10 % and 20 % of variation around the mean, i.e. for a mean of 0.5 Phytoseiidae/leaf that means that the samplings can provide estimation intervals of 0.45 -0.55 and 0.4 -0.6 Phytoseiidae/leaf, respectively.

Species of Phytoseiidae in French apple orchards
One thousand, nine hundred and eighty-five specimens have been identified. Eleven species of Phytoseiidae have been observed. The list of species, localities and regions is provided in Table 1. Three species were particularly abundant and frequently observed: A. andersoni, K. aberrans and T. ( T.) pyri (Table 2). Cydnodromus californicus was observed in several regions in medium densities, whereas the other seven species were observed in some plots but usually in low densities and in association with the three dominant species (Tables 2, 3). Apart from the three dominant species, only two (Euseius gallicus Kreiter and Tixier and Euseius stipulatus [Athias-Henriot]) have been found alone (not occurring with other Phytoseiidae species) in one and three plots, respectively (Tables 1 and 2). In most of orchards, the three dominant species were not found in association with each other (Table 3). The highest co-occurrence was found between T. ( T.) pyri and A. andersoni co-found in 11 plots among the 173 sampled (Table 3). Cydnodromus californicus was always found with one of the three dominant species and highest co-occurrence was observed with A. andersoni (in 17 plots among the 20 in which C. californicus was found). Duso et al. (2009) reported the dominance of four species of Phytoseiidae in European orchards: A. andersoni, K. aberrans, T. (T.) pyri as in the present survey. However, they also reported the dominance of another species, Euseius finlandicus (Oudemans), rather rare in the apple orchards considered in our study. Table 4 shows the cultivars on which the three dominant species have been found. No clear relationship between apple cultivar and Phytoseiidae species appears as the three dominant species have been found on the most common cultivars.
The list of the eleven species found is provided below with some information on their occurrence and biology.

plots -
Amblyseius andersoni was the most frequent and abundant species. It was observed in more than half of the orchards sampled, on 27 apple cultivars (among 34 sampled) and in nearly all the French regions sampled (Figure 1a, b). Amblyseius andersoni was dominant in seven of the regions sampled. We can thus question the occurrence of this species in such regions if the number of plots would have been higher. This species is quite common in agrosystems, especially in vineyards and apple orchards in Europe (i.e. Spain, Turkey, Switzerland, Slovenia and Italy) (Moraes et al. 1986). It is reported to feed and develop on tetranychid mites and to ensure efficient biological control of these mites (Duso and Camporese 1991;Genini et al. 1991;Koveos and Broufas 2000;Fischer and Mourrut-Salesse 2005;Houten et al. 2005;Lorenzon et al. 2012). Some studies have also shown its ability to develop resistance to pesticides (i.e. Duso et al. 1992;Pozzebon et al. 2002;James 2002James , 2003. Sundourner 100 one orchard in North West France (Table 1). One specimen has been observed in this study, but this species is quite common in France, especially on uncultivated shrubs and trees but rarely in crops (Moraes et al. 1986(Moraes et al. , 2004. It is reported as a frequent species of apple orchards in Europe by Duso et al. (2009).

Euseius finlandicus has been observed in only
Euseius gallicus has been observed in four orchards located in the South-East of France (Table 1) on three apple cultivars. This species recently described has been reported from shrubs and trees. Nothing is known on its biology and it is morphologically close to E. stipulatus (Okassa et al. 2009;Tixier et al. 2010).
Euseius stipulatus has been found in four orchards located in the South-East France (Table 1). This species is commonly found in the southern Europe. It is a very common species in crops, especially in citrus orchards (Ferragut and Escudero 1997;Sahraoui et al. 2012). Several studies have shown its ability to feed on pollen but also on pests such as Tetranychus urticae (Koch) and Panonychus citri (McGregor) (i.e. Ferragut et al. 1992;Abad-Moyano et al. 2009;Pina et al. 2012). It is usually found on plants with smooth leaves. In the present study, it has been reported on four apple cultivars, but essentially on Golden Delicious, whereas in other orchards with the same cultivar, other Phytoseiidae species have co-occurred.
Kampimodromus aberrans was the second most abundant and frequent species; it was found in 43 orchards and 16 apple cultivars. It was however less widespread than A. andersoni (Figure 1a, b). Indeed, it was only present in four regions and it prevailed in Languedoc-Roussillon only. In Franche-Comté, K. aberrans was the unique species sampled; however only one plot was considered. This species has a Palearctic distribution; it has been observed both in natural vegetation and crops, especially apple orchards and vineyards (i.e. Tixier et al. 1998Tixier et al. , 2000Kreiter et al. 2002;Duso et al. 2009). However, it is more often reported in untreated apple orchards than in commercial plots (Duso et al. 2009). In France, this species is the prevalent species in vineyards of southern France, whereas in the North T. (T.) pyri prevails (Kreiter et al. 2000).
This southern distribution is similar to what has been presently observed in French apple orchards. Climatic conditions, especially dry conditions of Mediterranean climate might favour the presence of K. aberrans. However, this species has been reported from higher latitudes, as Germany, Ukraine and Slovakia in orchards (Schruft 1967, Jedlickova 1991, Kolodochka and Omeri 2007 and presently in North-East France, suggesting that other factors could explain its distribution. Duso et al. (2009) suggested that the occurrence of K. aberrans in Italian apple orchards was linked to pesticide applications and tolerance to pesticides applied. Duso et al. (2009) also showed the importance of apple cultivar leaf characteristics on the occurrence of this latter species. In the present study, we can note that K. aberrans was particularly abundant on cultivars "Reinette", known to have hairy leaves and on the cultivar Chanteclerc. However this latter cultivar has only been sampled in Languedoc-Roussillon, thus it is impossible to determine if the dominance of K. aberrans is due to cultivar or climatic conditions.
Cydnodromus californicus was observed in twenty orchards and eleven cultivars but always at low densities (Table 4). It was mainly observed in Provence-Alpes-Côte d'Azur and Midi-Pyrénées (Table 1). It was the prevailing species in apple orchards in surveys carried out more than ten years ago (Bourgouin et al. 2002). This species tends thus to disappear in apple crops. This evolution is similar to what has been observed in vineyards (Kreiter et al. 2000). Cydnodromus californicus has been often reported as a species usually present in highly treated plots because of its ability to develop resistance to pesticides (i.e. Fauvel and Bourgoin 1993;Castagnoli et al. 2005;Cloyd et al. 2006). The fact that this species has disappeared from French orchards and vineyards could be explained by the development of Integrated Pest Management practices and the decreasing of toxic pesticide applications. Duso et al. (2009) observed that K. aberrans increased its densities in apple orchards when pesticides less toxic to Phytoseiids are used; it would be more competitive than other species. However, additional studies should be carried out to confirm this hypothesis.
Paraseiulus triporus (Chant and Yoshida-Shaul) was found on five cultivars in five orchards located in South of France (Table 1). The densities were always low. This species is rather common in the entire West Palearctic region; it has been reported from apples in Sweden, Italy and The Netherlands (Moraes et al. 2004). Nothing is known on its biology.
Typhlodromus (Anthoseius) rhenanoides Athias-Henriot was found in one apple (cultivar Golden Delicious) orchard in South of France, only (Table  2). It has been observed on apples in Spain and Portugal and on vines in France (Kreiter et al. 2000). However, this species is rarely reported from crops.
Typhlodromus (Typhlodromus) baccettii Lombardini was found in one apple orchard (cultivar Red Winter) in South of France. It has been reported from apples only from Norway (Edland and Evans 1998). Nothing is known on its biology. It is a quite rare species only reported from Europe. It is the second record of this species in France .
Typhlodromus (Typhlodromus) pyri was found in nearly all the regions, 40 orchards and 21 apple cultivars (Figure 1a, b). However, even if present in many regions, it was observed only in some plots. It is interesting to note the dominance of this species in the regions Rhône-Alpes and Basse-Normandie. This species is quite common in apple orchards and vineyards all over the world (Hardman et al. 1991;Moraes et al. 1986Moraes et al. , 2004Roda et al. 2003). Several studies have shown its ability to control mite pests and to resist somewhat to pesticide applications (i.e. Genini et al. 1991;Bonafos et al. 2007). Roda et al. (2003) showed that apple leaf pubescence could affect the densities of T. (T.) pyri because of pollen and fungi spore retention. In the present study, no clear effect of apple cultivar on its occurrence has been observed (Table 4).
Phytoseius horridus Ribaga was found in one apple orchard (11 specimens identified on the cultivar Karmerrien) in North West of France (Table 1). It is the first report of this species in France. It has also been observed on apples in Spain (Miñarro et al. 2002). Nothing is known on the biology of this west Palearctic species.
In order to assist the identification of the females of the Phytoseiidae species reported in apple orchards in France, an identification key is provided below.

Phytoseiidae distribution in apple orchards
A high and significant correlation was observed between log(m) and log(S 2 ) (R 2 = 0.96, P < 0.0001), allowing to use the indices of Taylor's law to characterise Phytoseiidae distribution. The slope value was 1.26 (Standard Error = 0.02), showing an aggregative distribution. Table 5 shows the parameters of the regression for different regions (where the number of plots was sufficient to carry out the analysis). In all the regions considered, the slope value was significantly higher than 1 (except in Aquitaine), showing an aggregated distribution. However, sometimes two species were co-occurring in a same plot. It is thus impossible to determine if the distributions of the two species are differ- ent. In order to assess the distribution of the three dominant species, only plots where one of these latter species represented 80 % of the densities were considered and then the Taylor's law was applied.
The slope values obtained were significantly higher than 1 for the three species, showing a clear aggregated distribution for all of them (Table 5).

Characterisation of the sample size for optimal sampling
As Phytoseiidae distribution is aggregated the number of leaves to be sampled should be important. To estimate the average densities of Phytoseiidae per leaf with an error of 10 % around the mean, 260 leaves per orchard would have to be sampled. For practical work and producers, this is clearly not possible as too time consuming. However, with an error of 20 % around the mean, the average number of apple leaves to be sampled would be 40. Below 40 leaves, the precision around the mean might be too low to estimate sufficiently accurately the densities of these predators. In order to simplify sam-plings, the relation between the average number of Phytoseiidae per leaf and the rate of leaves occupied by at least one mite has been assessed ( Figure  3). The good correlation obtained (R 2 = 0.88) enables to establish an abacus to determine the average densities of Phytoseiidae when only their presence/absence of the leaves is assessed (Table 6).

CONCLUSION
This paper is the first one presenting such a great survey in apple orchards in France. It shows the importance of three species and their relative abundance. Amblyseius andersoni was clearly the dominant species whereas the samplings carried out ten years ago demonstrated the dominance of C. californicus. This survey confirms thus the fauna modification seemingly initiated ten years ago.
The prevalence of A. andersoni does not apply in all the regions. Kampimodromus aberrans is dominant in Languedoc-Roussillon and Franche-Comté whereas T. ( T.) pyri prevails in Rhône-Alpes, Limousin, Poitou-Charentes and Basse-Normandie. However, in these three latter regions as well in Franche-Comté, the number of orchards sampled is too low to consider this distribution representative. However, it is clear that K. aberrans predominates in Mediterranean climates whereas T. ( T.) pyri prevails in Rhône-Alpes. Cultural practices and climatic conditions could certainly explain such different localisations. Kampimodromus aberrans seems to be less affected by low relative hygrometry than the other two species (K. aberrans, RH 50 = 56 %, T. ( T.) pyri RH 50 = 58 %, A. andersoni RH 50 = 62 %) (Duso and Camporese 1991;Genini et al. 1991;Gambaro 1994). This could explain why K. aberrans has been mainly recorded in orchards of Languedoc-Roussillon (the driest region considered) and A. andersoni in more humid regions. However, as K. aberrans has also been found in North of France, as well as in North and central Europe (Moraes et al. 2004) and A. andersoni in Provence where humidity is quite low, other factors probably affect Phytoseiidae occurrence. For instance irrigation type but probably essentially pesticide applications could be involved. Kreiter et al. (2000) and Duso et al. (2009) have indeed shown the importance of pesticide application to explain the distribution of Phytoseiidae in French vineyards and Italian apple orchards, respectively.
Furthermore, K. aberrans is usually observed on plants with hairy leaves (Kreiter et al. 2002), suggesting that apple cultivar could also affect the diversity of Phytoseiidae. Duso et al. (2009) also emphasized the influence of apple cultivar on Phytoseiidae densities, especially for the species K. aberrans. However, in the present case, cultivar does not seem to affect Phytoseiidae mite species occurrence. Other factors, such as cover-crop type, not much studied until now in French apple orchards could affect the Phytoseiidae diversity. Some studies in apple orchards but also in citrus orchards indeed showed exchange between plants in the interrows and trees (Alston 1994;Coli et al. 1994;Nyrop et al.1994;Tuovinen 1994;Stanyard et al. 1997;Fitzgerald and Solomon 2004;Pereira et al. 2006;Aguilar et al. 2008Aguilar et al. , 2011Mailloux et al. 2010). Nothing is known on the occurrence in inter-rows of the main Phytoseiidae species found on apple trees. Kampimodromus aberrans and T. ( T.) pyri are essentially recorded on trees and shrubs but are little known from herbaceous plants. Amblyseius andersoni has been more frequently observed on herbaceous plants. The occurrence of such species in inter-rows should thus be studied especially to develop weeding practices for biological control improvement and natural enemy efficiency. Finally, this study shows the aggregative distribution of Phytoseiidae in apple orchards and provides useful information for improving samplings and more accurately determining Phytoseiidae densities on apple trees. In further surveys, it could be interesting to determine the relation between Tetranychidae and Phytoseiidae mites in order to propose decision rule for managing pesticide application.