Plant-feeding and non-plant feeding phytoseiids: differences in behavior and cheliceral morphology

Transcription

1 DOI /s y Plant-feeding and non-plant feeding phytoseiids: differences in behavior and cheliceral morphology Einat Adar Moshe Inbar Shira Gal Noam Doron Zhi-Qiang Zhang Eric Palevsky Received: 23 February 2012 / Accepted: 31 May 2012 Ó Springer Science+Business Media B.V Abstract In previous studies plant feeding behavior of plant- and non-plant feeding phytoseiids was never examined directly. Moreover, in these studies the cheliceral morphology of phytoseiids was not associated with their ability to feed on plants. In the present study, we monitored the plant-feeding behavior of Euseius scutalis and Amblyseius swirskii. OnlyE. scutalis was observed penetrating the leaf surface with the movable digit and feeding. Second, using a dye and coloring the gut as an indicator for feeding, we found that E. scutalis pierced an artificial membrane and fed whereas A. swirskii did not. Finally, to identify morphological characteristics typical of plant feeders versus non-plant feeders, we used scanning electron microscopy to examine the adaxial (inner) profile of the chelicerae in 13 phytoseiid species. The only parameter that distinguished between plant- and non-plant feeders was the ratio of the dorsal perimeter length of the fixed digit to the ventral perimeter length of the movable digit. Plant-feeders were characterized by ratio values greater than one whereas the values for non plant-feeders were lower than one. We suggest that a shorter and less curved movable digit, expressed by a high ratio, will facilitate the penetration of the leaf surface. Cheliceral traits proposed here as typical of plant feeders, were observed for five genera, indicating that plantfeeding may be more common in the Phytoseiidae than previously reported. We propose that the ability to feed on plants be added as a cross type trait of phytoseiid life-style types. Keywords Phytoseiidae Amblyseius swirskii Euseius scutalis Predatory mites Chelicerae Scanning electron microscopy Electronic supplementary material The online version of this article (doi: /s y) contains supplementary material, which is available to authorized users. E. Adar S. Gal N. Doron E. Palevsky (&) Department of Entomology, Newe-Ya ar Research Center, Agricultural Research Organization (ARO), Bet-Dagan, Israel palevsky@volcani.agri.gov.il E. Adar M. Inbar Department of Evolutionary and Environmental Biology, University of Haifa, Haifa, Israel Z.-Q. Zhang Landcare Research, 231 Morrin Road, Auckland, New Zealand

2 Introduction Predatory mites of the family Phytoseiidae are important acarine biological control agents (ABAs), and therefore have been investigated extensively (reviewed in Gerson et al. 2003). Plant feeding may play the same role as pollen feeding in promoting pest control efficacy (Sabelis and Van Rijn 2005), either by population increase of the phytoseiids (McMurtry and Scriven 1966; Zhao and McMurtry 1990), or by higher sustainability of the predators on the host plant when prey is scarce (McMurtry 1992; Coll and Guershon 2002; Tanigoshi et al. 1993). The advantage of plant tissue over pollen as an alternative food source is that it is available for these predators all year round. Several studies have shown that specialized pollen feeders and generalist phytoseiids feed directly on the plant. It was first experimentally documented using systemic dyes with Typhlodromus rhenanus (Oudemans), Typhlodromus pyri Scheuten and Euseius finlandicus (Oudemans) (Chant 1959). Later, plant feeding by T. pyri was also linked to minute feeding damage of apple leaves and fruit (Sengonca et al. 2004). Porres et al. (1975) evaluated the leaf feeding of three Euseius species, E. hibisci (Chant), E. stipulatus (Athias-Henriot) and E. fructicolus (Gonzales and Schuster), on plants labeled with radioactive phosphoric acid, and found that only E. hibisci fed from avocado leaves, and none showed evidence of feeding from lemon foliage. Similarly, Kreiter et al. (2002) demonstrated the feeding of Kampimodromus aberrans (Oudemans) on plants of the European nettle tree, Celtis australis, marked with rubidium. Using the systemic insecticide aldicarb, Magalhães and Bakker (2002) monitored the survival of Typhlodromalus aripo DeLeon, a suspected plant feeder of cassava, versus the spider mite predators Neoseiulus idaeus Denmark and Muma and Phytoseiulus persimilis Athias-Henriot. As expected, only the survival of T. aripo was negatively affected on plants treated with aldicarb, thus confirming its ability to feed from the leaves, versus the inability of the two species of spider mite predators. Using the same technique, Nomikou et al. (2003) showed that Euseius scutalis (Athias-Henriot) fed from cucumber leaves, whereas the generalist Amblyseius swirskii Athias-Henriot did not. Additionally, treatments such as applying fertilizers and pruning in citrus enhanced fecundity in Euseius tularensis Congdon (Grafton-Cardwell and Ouyang 1995, 1996), possibly due to facilitation of plant feeding on young supple foliage. While the above studies show that generalist and pollen feeding phytoseiid predators can feed from plant tissues, they did not document plant feeding directly. Buryn and Brandl (1992), in an attempt to quantify Karg s (1961) hypothesis, correlated diet to cheliceral morphology (length, width and teeth number on the fixed and movable digits) of gamasid mites (suborder including the superfamily Phytoseioidea). Their results showed that arthropod feeders have relatively large chelicerae, nematode feeders bear more teeth and mites feeding on plant material, have relatively small chelicerae. Phytoseiid feeding and functional morphology studies were conducted by Flechtmann and McMurtry (1992a, b) on spider mite predators, generalists and pollen feeders. The movable digit of spider mite predator species (of the genera Phytoseiulus, Galendromus, Typhlodromus and Neoseiulus) were described as having a long nearly right angled hook. In contrast, the chelicerae of pollen feeding predators (of the genera Euseius and Iphiseius) were broadly triangular and blunt. However, quantitative measurements of the chelicerae in these phytoseiid species were not performed, and plant feeding was not discussed. In the present study we describe plant feeding behavior, and document feeding through an artificial membrane. We then test the hypothesis that plant feeding behavior of phytoseiids, and the ability to pierce and feed through an artificial membrane, is related to the functional morphology of the chelicerae.

3 Materials and methods General methods Names of phytoseiid mites mentioned in the text follow the genera under which each species is presently placed (Chant and McMurtry 2007). Feeding observations and comparative feeding experiments were conducted at Newe Ya ar research center, Israel. Euseius scutalis, collected from avocado leaves, were reared on potted plants of either Capsicum annum (sweet pepper, Red Rock variety 7180), or Solanum nigrum (black nightshade), in a growth chamber maintained at 27 ± 4 C, 34 ± 9 % RH, 16L:8D. Iphiseius degenerans, collected from citrus, were reared on S. nigrum, in potted plants, at ambient room temperature. Quercus ithaburensis (Mount Tabor oak) pollen, stored at -20 C, was applied to leaves of potted plants twice a week as a food source. Mites of the species A. swirskii, originally collected from citrus foliage and reared on Carpoglyphus lactus L. (Acari: Carpoglyphidae), were received from Bio-Bee Biological Systems, Sdeh Eliyahu, Israel. Predators were starved on C. annum leaves in a refrigerator at 4 C, for two days before experimental use. Observations and monitoring of feeding behavior of. E. scutalis, A. swirskii and I. degenerans were performed at magnifications of using a stereo microscope with a planapo 1.69 objective (Leica M205 FA, Leica Microsystems, Heerbrugg, Switzerland) with a full high definition digital video camera (Optronics, Goleta, California, USA). Feeding arenas described below, were fastened to a plastic bottle cap (external dimensions (d 9 h): mm), mounted on a compact ball tripod mount (SLIK SBH- 100, Hidaka, Japan) with modeling clay (Figs. 1, 2a), which facilitated lateral viewing. To prevent predator escape from the arena, a thin coat of hair gel (Natural Formula Fix Curl Fig. 1 Feeding arena (ar) fastened to a plastic bottle cap (bc) mounted on a compact ball tripod mount (tp) with modeling clay (mc)

4 Fig. 2 Schematic illustration of feeding arena (ar) a fastened to a plastic bottle cap (bc) using adhesive tape (at). A thin coat of hair gel (hg) applied around the edges of the arena surface (as). Three types of arena surfaces were used: b Leaf disc arena, abaxial (lower) leaf surface facing up (lab). c Parafilm Ò arena (pf). d Green silk paper (sp) arena covered on both sides with a thin layer of beeswax (bw-sp) and wrapped over the edges of the bottle cap filled with water-soaked cotton wool (cw). The beeswax was carefully scratched on its lower surface (sbw) to allow water to penetrate the paper from below keeping the upper surface intact and dry (ibw) Gel, Careline Pharmagis,Yeruham, Israel) was applied around the edges of the arena surface (Fig. 2a). Feeding behavior observations Females were starved for two days at 4 C, and subsequently transferred to an arena (1 3 mites/arena). Feeding behavior of adult female predators was video recorded on pepper or nightshade leaf disc arenas, fastened with double sided adhesive tape to the top of the plastic bottle cap, abaxial (lower) leaf surface facing up (Fig. 2a, b). Penetration of chelicerae into an artificial membrane was documented using two techniques: (1) Parafilm Ò (Pechiney plastic packaging, Menasha, WI, USA) stretched over the edges of an empty bottle cup (Fig. 2a, c). This method enabled us to clearly observe the penetration and retraction of the chelicerae from the feeding surface. (2) Green silk paper (Office Depot, Afula, Israel) covered on both sides with a thin layer of beeswax (Argov et al. 2002), the lower surface of the beeswax carefully scratched to allow water to penetrate the paper from below, keeping the upper surface intact and dry. The waxed paper was then wrapped over the edges of a bottle cap filled with water-soaked cotton (Fig. 2a, d). The plasticity of the beeswax membrane facilitated the documentation of puncture holes, following the removal of the chelicerae.

5 Feeding experiments The arena used for examination of the feeding ability through an artificial membrane, was made of green silk paper covered in beeswax, as described above (Fig. 2a, d). When a predator pierced the beeswax layer and fed on the colored water, its gut was dyed green. This method enabled us to distinguish between predators that fed through the membrane, from those that did not. Starved mites of each species were placed individually (n = 60 for each species) on the arenas, at room temperature. After 125 ± 5 min (phase one of the experiment), the number of green colored mites was recorded. As none of the A. swirskii (see Results ) was colored green, in the second phase of the experiment, the silk paper of the arenas was punctured (10 15 times) with a fine pin, allowing for the formation of small green pools. After an additional 20 min, the number of green predators was counted, which gave an indication of the proportion of predators (that did not feed in phase one) that fed when free water was available. To test statistically the hypothesis that E. scutalis could penetrate the artificial membrane and drink, while A. swirskii could not, we used a Z test on the proportions of green dyed E. scutalis and A. swirskii in phase 1 of the experiment. For testing the hypothesis that A. swirskii were hungry in phase 1, but were incapable of reaching the water, we used the non-parametric McNemar s test for related samples on the occurrences of dyed A. swirskii, before and after membrane puncturing (phase 1 and 2). Comparative functional morphology of chelicerae For the functional morphology study of chelicerae, using a scanning electron microscope (SEM), females of 13 species of phytoseiids from eight genera, were sent to the New Zealand Arthropod Collection, Landcare Research, Auckland, New Zealand, as preserved specimens in 70 % ethanol from laboratories and biocontrol companies in North America, Europe and Israel (Table 1). Prior to shipping, mites (ca. 60 specimens per species) were killed in hot water to force the protrusion of the chelicerae (Flechtmann and McMurtry 1992a). This was performed by placing live mites in a watch glass on a hot plate (80 C). After 30 s, the mites were transferred to 50 % ethanol at room temperature for 15 min, and then to a sealed vial with 1.5 ml ethanol 70 % for shipping. To avoid the introduction of impurities, distilled water and absolute ethanol were used to prepare solutions. Upon receipt, mites were transferred with a micropipette to micro-porous (78 lm pore diameter) specimen capsules (ProSciTech, Thuringowa, QLD, Australia), and dehydrated in three additional 15 min ethanol rinses (once in 95 % and twice in 100 %), after which mites were dried in a critical point dryer (CPD 030 BAL-TEC, Schalksmühle, Germany), following the CPD protocol specified by BAL-TEC. Dried mites were emptied from the specimen basket into a watch glass, and placed with a minute pin ventral side up on a carbon tab, mounted on a pin type SEM mount. Using a stereo microscope at 1009 (MZ12.5 Leica Microsystems, Heerbrugg, Switzerland), the gnathosoma was severed with a minute pin, and the chelicerae were separated with an eye lash, and laid flat on the carbon tab. SEM mounts were gold coated for 6 min in a NeoCoater MP-19020NCTR (Jeol, Tokyo), and images were taken with a SEM NeoScope JCM-5000 (Jeol, Tokyo), at magnifications of 1,300 and 2,7009. Measurements, using the imaging software NIS-Elements BR 3.1 (Nikon, Japan), were only conducted on chelicerae with the adaxial (inner) profile facing upwards (Fig. 3c), as features of the movable digit (MD) are hidden by the lobe and fixed digit (FD) in the abaxial profile (Fig. 3a, b). We used the apex of the FD and the constriction at the base of

6 Table 1 Phytoseiid species received for the functional morphology study from research institutes and biocontrol companies Type a Plant feeding Genus Species Company/Institute Citation I - Phytoseiulus persimilis Bio-Bee b Magalhães and Bakker (2002) I Phytoseiulus longipes Bio-Bee II III Neoseiulus californicus CRA-ABP c III Neoseiulus cucumeris KNL d III - Amblyseius swirskii Bio-Bee Nomikou et al. (2003) III Amblyseius largoensis TREC e III? Typhlodromus pyri AS CR f Sengonca et al. (2004) III Typhlodromus exhilaratus CRA-ABP III Phytoseius plumifer ARO III? Kampimodromus aberrans CRA-ABP Kreiter et al. (2002) IV? Iphiseius degenerans ARO g Adar (unpublished) IV Euseius ovalis KNL IV? Euseius scutalis ARO Nomikou et al. (2003) Genera are sorted by feeding types described by McMurtry and Croft (1997). Literature sources suggesting species capacity for plant tissue feeding (?) or not (-) are listed a Type I, Tetranychus specialist; Type II, tetranychid specialist; Type III, generalist; Type IV, pollen specialist b Bio-Bee Biological Systems, Kibbutz Sdeh Eliyahu, Israel c CRA-ABP, Research Center for Agrobiology and Pedology, Firenze, Italy d Koppert Biological Systems, The Netherlands e Tropical Research and Education Center, University of Florida, USA f Entomology, Biology Centre AS CR, Ceske Budejovice, Czech Republic g Entomology, Newe-Ya ar Research Center, ARO, Israel the MD, anterior to the arthrodial process, as points of reference for measuring (Fig. 4b), which are evident in the SEM images (Fig. 3c) but apparently not in light microscopy. To identify traits that could serve to characterize facultative plant feeders, the following measurements were recorded for 6 9 specimens per species. Lengths measured (Fig. 4a): (1) Dorsal perimeter of FD from apex to tip (DP FD ). (2) Ventral perimeter of MD from base to tip (VP MD ). (3) Height of MD tip from plant surface (Height MD ). (4) Space or cavity between MD and FD (Cavity MD-FD ) when the chelae are in a normal closed position. In addition, we used the DP FD /VP MD ratio to compare the cheliceral shape among species, thereby precluding length measurement variability. Angles measured, using the middle article ventral base as a reference plane (Fig. 4b): (1) FD apex to FD tip (apex FD tip FD ). (2) FD apex to MD tip (apex FD tip MD ). (3) MD base to MD tip (base MD tip MD ). One way ANOVA was used to analyze all measured parameters, followed by LSMeans differences separated with Tukey HSD. We compared between species that are known as plant and non-plant feeders, between species of different feeding types (Table 1), and between species within genera, assuming that the capacity for plant feeding may be characteristic of some genera. For comparison between males and females, we also measured the chelicerae of T. pyri males and used a T test for comparison between the DP FD /VP MD ratios.

7 Fig. 3 Scanning electron microscope images of chelicerae of phytoseiid females taken at 91,300, white scale bars 10 l. a Iphiseius degenerans abaxial profile. b Neoseiulus cucumeris abaxial profile. c Iphiseius degenerans adaxial profile Results Feeding behavior observations Prior to feeding E. scutalis was observed walking or standing on the leaf, moving its palps and forelegs, crimping and piercing the leaf surface (veins or between veins) (Online Resource 1). Penetration holes were documented on a beeswax membrane (Online Resource

8 Fig. 4 Schematic illustrations of chelicerae of phytoseiid females. a Length measurements (dashed lines): (1) Fixed digit (FD) dorsal perimeter (fdp), from apex (fda) to tip (fdt). (2) Movable digit (MD) ventral perimeter (mdp) from base (mdb) to tip (mdt). (3) Height of MD tip from plant surface (mdh). (4) Space or cavity between MD and FD (ca). b Angles measured, using the ventral base of the middle article (mab) asa reference plane (rp). (1) FD apex (fda) to FD tip (fdt). (2) FD apex to MD tip (mdt). (3) MD base (mdb) to MD tip 2). It appears that the leaf surface is pierced with the movable digit, while the fixed digit presses down on the leaf surface, allowing for the crimping action. When feeding, the piercing of the movable digit is followed by the penetration of the distal tip of the subcapitulum, evidently with the corniculi (Online Resource 3). During feeding, forelegs are held in the air, lightly vibrating (Online Resource 4), shimmering light is discernible in the gnathosoma (Online Resources 3, 5, 6), possibly indicating the uptake of fluid or air, and stirring movements are evident in the digestive tract (Online Resource 6). Feeding sessions can take seconds to several hours. Once ended, mouthparts are removed from the feeding surface, starting with the retraction of the subcapitulum, followed by alternate movements of the movable digits (Online Resource 7). During this process the mite sometimes touches the surface with its palps, in what seems to be supporting the retraction of the chelicerae (Online Resource 8). After feeding, chelicerae are apparently brushed clean by the palps (Online Resource 9). Behavior studies, parallel to those described above, were conducted for starved A. swirskii, but feeding from a leaf or piercing a membrane were never observed. Feeding experiments Proportion of feeding mites recorded as 0.61 green dyed E. scutalis (N = 56), were significantly higher (Z = 9.22, P \ 0.01) than the nil proportion of A. swirskii (N = 52).

9 Following membrane puncturing, the proportion of dyed E. scutalis increased to 0.76 (N = 55), while the proportions of dyed A. swirskii increased substantially from 0 to 0.45 (N = 49, P \ 0.01). Comparative functional morphology of chelicerae Significant differences among species (Fig. 5) were found in all length (Table 2) and angle (Table 3) measurements. Similarly, DP FD /VP MD ratio yielded significant differences among species (F (12,82) = 72.45, P \ 0.01, Fig. 6). However, length measurements (DP FD, VP MD, height MD and cavity FD-MD ) did not differentiate between known plant and non plant feeding species (Table 2, column 1), and between species belonging to different types. DP FD were similar for species within genera, except for the two Neoseiulus species. VP MD only differed between the two Phytoseiulus and the two Neoseiulus species. Height MD differed only between the two Phytoseiulus species, and cavity MD FD differed between the two Phytoseiulus and the two Amblyseius species. In contrast to the absolute length measurements, the DP FD /VP MD ratio differentiated between known plant feeding and non plant feeding species (Fig. 6 solid and dashed lined rectangles respectively), and between species that are spider mite specialists (type I) and species that are pollen specialists (type IV) (Fig. 6). Additionally, species within all genera studied, were similar. However, females of T. pyri had a significantly higher ratio than males (t (12) = 8.49, P \ ). None of the angle measurements differentiated between known plant feeding species and non-plant feeding species. Apex FD tip FD was the only angle that differed between species that are spider mite specialists (type I) and species that are pollen specialists (type IV). Within genera, all angles measurements differed between the two Phytoseiulus species, and base MD tip MD differed also between the two Typhlodromus species. Discussion In previous studies, where food uptake from leaves by phytoseiids has been demonstrated using dyes, radioactive isotopes and systemic insecticides (Chant 1959; Porres et al. 1975; Magalhães and Bakker 2002; Nomikou et al. 2003), it was not clear how the mites actually fed. Do the mites feed by piercing the plant cuticle or could they be utilizing natural opening such as stomata (Ochoa et al. 2011)? In the present study, we have shown on leaves and artificial membranes, that feeding by E. scutalis and I. degenerans entails crimping and piercing the feeding surface. Our video recording of plant feeding is very similar to Flechtmann and McMurtry s (1992b) description of feeding on spider mites, where the movable digit penetrates the feeding surface, followed by the corniculi, with the fixed digit remaining outside and the front legs lightly vibrating in the air. Additionally, our description confirms that the corniculi are not used for penetrating the leaf, as the penetration of the subcapitulum [bearing the opening of the pharynx (Evans 1992; De Lillo et al. 2001)], seemed to follow the penetration of the movable digit. Given these similarities, we hypothesize that the principal difference between plant and spider mite feeding, relates to the shape of the movable digit that is required to pierce the leaf surface, in contrast to the shape required for clutching and cutting into the body of a prey (see cheliceral morphology discussed below). The experiments with beeswax covered silk paper that showed E. scutalis penetrating and feeding through an artificial membrane, and our video recordings of piercing the artificial membranes, suggest that plant feeding utilizes a similar piercing mechanism.

10 Fig. 5 Scanning electron microscope images of chelicerae, adaxial profile, of phytoseiid females of 13 c species taken at 92,700, white scale bars 10 l. Species sorted by ratio of the dorsal perimeter (DP) of the fixed digit (FD)/ventral perimeter (VP) of the movable digit (MD) as presented in Fig. 6 (DP FD /VP MD ) except for the male of Typhlodromus pyri. a Neoseiulus californicus. b N. cucumeris. c Amblyseius largoensis. d A. swirskii. e Phytoseiulus persimilis. f P. longipes. g T. pyri. h T. pyri male. i T. exhilaratus. j Iphiseius degenerans. k P. plumifer. l Kampimodromus aberrans. m Euseius ovalis. n E. scutalis In contrast, A. swirskii did not feed through the membrane, even though it was hungry and fed when the green dyed water was made accessible. These results concur with those of Nomikou et al. (2003), who concluded that E. scutalis fed on plant tissue while A. swirskii did not. As previously indicated, in the functional morphology study conducted by Flechtmann and McMurtry (1992a), spider mite feeders (type I and II) were described as having a nearly right angled hooked movable digit. Accordingly, we assumed that the angle Base MD tip MD would allow to distinguish between spider mite feeders and pollen feeders (type IV), but our results led us to refute this hypothesis. The angle that did separate between spider mite and pollen feeders was Apex FD tip FD, the highest angle characterizing pollen feeders. This seems to fit Flechtmann and McMurtry (1992a) triangular description of the pollen specialist chelicerae. Notably, the Apex FD tip FD angle of the three pollen feeders (Table 3, column 1) did not differ from K. aberrans, possibly indicating that this species may be classified as a type IV phytoseiid. The only parameter that distinguished between plant (i.e. E. scutalis, T. pyri, I. degenerans and K. aberrans) and non plant (i.e. P. persimilis and A. swirskii) feeders was the DP FD /VP MD ratio. Plant feeders were characterized by ratio values greater than one, whereas the values for non-plant feeders were lower than one. Tracing the perimeters of the digits creates a schematic triangle, the Apex FD Base MD being the base, and the tip FD the head. A DP FD /VP MD ratio of one, implies that cheliceral perimeters are equal; similarly, a DP FD /VP MD ratio greater than one, implies that the movable digit is shorter than the fixed digit, and a DP FD /VP MD ratio smaller than one, implies that the movable digit is longer than the fixed digit. Note that the MD of all species studied curved up to, and met with the FD, posterior to the tip FD, thus the longer the VP MD relative to the DP FD,the greater the curve of the MD. Indeed, non-plant feeders studied here had relatively long curved MD (low DP FD /VP MD ratio), which is comparable to the spider mite specialists MD described by Flechtmann and McMurtry (1992a). We suggest that a dagger like movable digit (shorter and less curved), expressed by a high DP FD /VP MD ratio, will facilitate the penetration of the leaf surface (Online Resource 3). The longer, more curved movable digit, of other plant feeding phytoseiids such as T. pyri, (lower DP FD /VP MD ratio than pollen specialists), may make it difficult for the mite to pierce the leaf in the same way, possibly explaining the scarring made by T. pyri, reported by Sengonca et al. (2004). They also found that males do not feed on plants and do not create scars. Interestingly, we found that the DP FD /VP MD ratio of the T. pyri females is larger than the ratio of the males. Another option that we cannot rule out, is that phytoseiids such as T. pyri use a different method to obtain plant fluids altogether. Whether Euseius or Iphiseius species cause leaf damage is still an open question. The pricking damage caused by E. scutalis, following the penetration on bee wax coated paper (Online Resource 2), does indicate their ability to penetrate a membrane, but it does not prove that the mites cause damage to the leaf surface. More ultra-structural research is needed to determine the feeding mechanism.

11

12 Fig. 5 continued DP FD /VP MD ratios were always similar for species within genera, whereas this was not always the case for length and angle measurements. For instance species with large and small chelicerae within the same genera (for example P. persimilis and P. longipes; N. californicus and N. cucumeris), have different absolute values for the VP md. Variation in angle measurements could result from measurement error, caused by the flexing of the middle article (non-sclerotized tissue), during cheliceral placement on the stub (see Materials and methods for angle measurement protocol). Based on the similarities of the DP FD /VP MD ratio within genera, and the literature (Chant 1959; Sengonca et al. 2004; Nomikou et al. 2003; Porres et al. 1975) we hypothesize that T. exhilaratus and E. ovalis are also plant feeders (Fig. 6). Furthermore, as DP FD /

13 Table 2 Mean (±SE) length measurements (microns) of females of 13 phytoseiid species Type a Plant feeding Genus Species DPFD VPMD HeightMD CavityFD-MD I - Phytoseiulus persimilis 23.3 ± 0.2 fg 25.0 ± 0.3 cde 10.9 ± 0.5 b 0.7 ± 0.2 f I Phytoseiulus longipes 21.1 ± 0.3 g 21.9 ± 0.3 f 8.7 ± 0.2 cd 3.6 ± 0.1 abc II III Neoseiulus californicus 23.8 ± 0.2 f 26.9 ± 0.3 c 14.8 ± 0.4 a 2.9 ± 0.1 bcd III Neoseiulus cucumeris 31.1 ± 0.2 a 34.6 ± 0.5 a 15.3 ± 0.4 a 3.6 ± 0.1 ab III - Amblyseius swirskii 28.3 ± 1.2 bc 30.5 ± 1.3 b 14.8 ± 0.4 a 0.9 ± 0.6 ef III Amblyseius largoensis 28.9 ± 0.3 b 32.0 ± 0.3 b 15.1 ± 0.4 a 2.2 ± 0.3 d III? Typhlodromus pyri 24.5 ± 0.5 def 23.8 ± 0.5 def 9.7 ± 0.3 bc 2.0 ± 0.2 de III Typhlodromus exhilaratus 26.4 ± 0.3 cd 25.1 ± 0.2 cde 9.4 ± 0.3 bcd 2.2 ± 0.1 cd III Phytoseius plumifer 23.8 ± 0.4 ef 22.1 ± 0.2 f 8.3 ± 0.1 cd 3.9 ± 0.2 ab III? Kampimodromus aberrans 26.2 ± 0.3 cde 23.0 ± 0.3 ef 8.9 ± 0.3 cd 3.2 ± 0.2 abcd IV? Iphiseius degenerans 27.9 ± 0.3 bc 26.1 ± 0.1 cd 10.9 ± 0.2 b 3.5 ± 0.1 abc IV Euseius ovalis 27.0 ± 0.2 bc 23.2 ± 0.3 ef 8.4 ± 0.2 cd 4.4 ± 0.4 a IV? Euseius scutalis 26.8 ± 0.2 c 22.9 ± 0.5 ef 8.0 ± 0.3 d 4.3 ± 0.2 a F (12,82) P (1) Dorsal perimeter of fixed digit (FD) from apex to tip (DPFD). (2) Ventral perimeter of movable digit (MD) from base to tip (VPMD). (3) Height of MD tip from plant surface (Height MD ). (4) Space or cavity between MD and FD (Cavity FD MD ). Different lower case letters indicate a significant difference between species (a = 0.05) Type I, Tetranychus specialist; Type II, tetranychid specialist; Type III, generalist; Type IV, pollen specialist (McMurtry and Croft 1997) a

14 Table 3 Mean (±SE) angles measurements (degrees) of females of 13 phytoseiid species, using the second digit base as a reference plane Type a Plant feeding Genus Species Apex FD tip FD Apex FD tip MD Base MD tip MD I - Phytoseiulus persimilis 8.8 ± 0.9 e 7.7 ± 0.7 e 34.7 ± 0.8 ab I Phytoseiulus longipes 18.9 ± 0.6 bc 20.9 ± 0.7 abc 28.9 ± 0.9 cd II III Neoseiulus californicus 10.3 ± 1.9 de 8.0 ± 1.8 e 34.8 ± 2.0 ab III Neoseiulus cucumeris 14.5 ± 0.3 cd 9.2 ± 0.6 de 31.1 ± 0.8 abc III - Amblyseius swirskii 18.6 ± 1.3 bc 11.6 ± 2.4 de 35.3 ± 1.5 a III Amblyseius largoensis 19.8 ± 0.7 b 7.9 ± 1.0 e 34.8 ± 1.0 a III? Typhlodromus pyri 16.6 ± 0.4 bc 13.7 ± 0.6 d 31.5 ± 0.5 abc III Typhlodromus exhilaratus 15.5 ± 2.0 bc 15.0 ± 1.3 cd 24.4 ± 1.6 d III Phytoseius plumifer 19.8 ± 0.6 b 19.8 ± 1.0 bc 26.9 ± 0.6 cd III? Kampimodromus aberrans 26.6 ± 0.7 a 24.6 ± 1.0 ab 29.4 ± 1.0 bcd IV? Iphiseius degenerans 27.8 ± 0.6 a 24.3 ± 0.5 ab 31.8 ± 0.8 abc IV Euseius ovalis 28.9 ± 0.8 a 27.1 ± 1.7 a 26.2 ± 0.8 cd IV? Euseius scutalis 26.4 ± 0.7 a 25.6 ± 0.8 ab 24.3 ± 1.0 d F (12,82) P Fixed digit (FD) apex to FD tip (Apex FD tip FD ). 2. FD apex to movable digit (MD) tip (Apex FD tip MD ). 3. MD base to MD tip (Base MD tip MD ). Different lower case letters indicate a significant difference between species (a = 0.05) a Type I, Tetranychus specialist; Type II, tetranychid specialist; Type III, generalist; Type IV, pollen specialist (McMurtry and Croft 1997) Fig. 6 Ratio of the dorsal perimeter (DP) of the fixed digit (FD)/ventral perimeter (VP) of the movable digit (MD) (DP FD /VP MD ) for phytoseiid females of 13 species. Species suspected to feed on plant tissue have a ratio[1. Species documented in the literature as plant feeding and non-plant feeding are framed with solid lined and dashed rectangles, respectively

15 VP MD ratio of P. plumifer is between the ratio values of I. degenerans and K. aberrans, both known as plant feeders (Adar unpublished) (Kreiter et al. 2002), we expect that it also has the capacity to plant feed. Non plant feeding has been suggested experimentally using systemic insecticide (Temik 10G Ò, active ingredient: aldicarb) treated potted plants for the species P. persimilis and N. idaeus on cassava leaf discs (Magalhães and Bakker 2002), and for P. persimilis and A. swirskii on detached cucumber leaves (Nomikou et al. 2003). Clearly, non-plant feeding is dependent on the experimental setup and cannot be extrapolated, meaning, a non-plant feeder in one experimental setup might become a plant-feeder when placed in a different setup. A good example of the effect of the plant host on the ability of phytoseiids to feed from the leaf, is the study conducted by Porres et al. (1975), who evaluated the ability of three Euseius species (E. hibisci, E. stipulatus, E. fructicolus) to feed on avocado and lemon leaves. All three Euseius species did not feed from lemon, but E. hibisci did feed from avocado leaves, its natural host. Surprisingly, E. stipulatus originating from citrus in Spain did not feed on lemon, possibly indicating that even the variety can have an effect on the capacity to feed. Euseius species were found to be morphologically similar in the present study and by Flechtmann and McMurtry (1992a), thus, apparently additional factors are involved in determining the predators ability to plant feed. Buryn and Brandl (1992) correlated cheliceral morphometrics with diet in a wide variety of gamasid mites. Their analysis included species specializing on worm-like prey, micro-arthropods, as well as polyphagous predators, but did not include omnivorous predators since only a few species were found. While they were able to make broad generalizations regarding the correlation between diet and cheliceral morphology, they concluded that diet, feeding behavior and morphology studies must be combined to accurately predict the preferred palette of a predator from morphological data. Our research, focused on Phytoseiidae, adopted this approach by combining video recordings of leaf feeding behavior with cheliceral measurements. To determine whether plant feeding is more common than previously envisioned, we suggest studying the cheliceral morphology and plant feeding behavior of additional species of the five genera: Euseius, Iphiseius, Kampimodromus, Phytoseius and Typhlodromus. Furthermore, we propose that cheliceral measurements should be taken from additional predators, that are known in the literature as plant feeders (such as Typhlodromalus aripo (Magalhães and Bakker 2002). Correlating plant feeding behavior with morphological traits for additional plant feeders will confirm whether these characteristics could be used to predict whether the species can plant feed. The division of the Phytoseiidae to life styles types, established by McMurtry and Croft (1997), became an important tool for understanding and utilizing ABAs from this family. They originally divided it to four types: I. Specialized predators of spider mites of the genus Tetranychus; II. Selective predators of spider mites; III. Generalist predators; and IV. Specialized pollen feeders. Adopting SEM imaging (becoming more affordable with the appearance of bench-top SEMs) of the cheliceral morphology for new species descriptions and for re-descriptions of existing species, will add important information relevant to their life style categorization. Recently, McMurtry and Moraes (2012) proposed to split type III and to add new life style types. Since plant feeding is crucial for the sustainability of predators, the latter, an especially important characteristic for biocontrol agents in stable systems, we suggest that the ability to plant feed be added as a cross type trait in this revision. Acknowledgments We thank Birgit Rhode, Leone Clunie, Rosa Henderson, Grace Hall & Chris Winks of Landcare Research New Zealand (LRNZ) for their technical support. We would like to express our gratitude

16 to Zengqi Zhao of LRNZ and Qing-Hai Fan of Biosecurity, Ministry of Agriculture and Forestry, New Zealand for their insights in development, and interpretation of the functional morphology study. For the inspiration for the scanning electron microscope study we are thankful to Norm Fashing of Department of Biology, College of William and Mary, VA, USA. For the mite species received for SEM imaging, we express our gratitude to Koppert NL; Sauro Simoni et al., CRA-ABP, Florence Italy; Bio-Bee, Israel; Rostislav Zemek et al., Biology Centre AS CR, Czech Republic; Jorge Pena, UFL, USA. We are grateful to Karin and Goni Shavit for their assistance in the video and figures editing. We express our appreciation to Uri Gerson of the Hebrew University of Jerusalem, Israel, and 3 anonymous reviewers for their helpful comments and Urs Wyss of University of Kiel, Germany for sharing his thoughts on video recording the feeding behavior of predatory mites. This work was partly supported by the European Commission, FP7, EUREKA project PHYTOFEED. This manuscript is a contribution of the Institute of Plant Protection, Volcani Center, ARO, Israel. References Argov Y, Amitai S, Beattie GAC, Gerson U (2002) Rearing, release and establishment of imported predatory mites to control citrus rust mite in Israel. Biocontrol 47: Buryn R, Brandl R (1992) Are the morphometrics of chelicerae correlated with diet in mesostigmatid mites (Acari)? Exp Appl Acarol 14:67 82 Chant DA (1959) Phytoseiid mites (Acarina: Phytoseiidae). Part I. Bionomics of seven species in southern England. Part II. A taxonomic review of the family Phytoseiidae, with descriptions of 38 new species. Can Entomol 91:1 166 Chant DA, McMurtry JA (2007) Illustrated keys and diagnoses for the genera and subgenera of the Phytoseiidae of the world (Acari: Mesostigmata). Indira Publishing House, West Bloomfield Coll M, Guershon M (2002) Omnivory in terrestrial arthropodes: mixing plant and prey diets. Annu Rev Entomol 47: De Lillo E, Di Palma A, Nuzzaci G (2001) Morphological adaptations of mite chelicerae to different trophic activities (Acari). Entomologica 35: Evans GO (1992) Principles of acarology. C.A.B International, Wallingford Flechtmann CHW, McMurtry JA (1992a) Studies on cheliceral and deutosternal morphology of some Phytoseiidae (Acari: Mesostigmata) by scanning electron microscopy. Int J Acarol 18: Flechtmann CHW, McMurtry JA (1992b) Studies on how phytoseiid mites feed on spider mites and pollen. Int J Acarol 18: Gerson U, Smiley RL, Ochoa R (2003) Mites (acari) for pest control. Blackwell Science, Oxford Grafton-Cardwell EE, Ouyang Y (1995) Augmentation of Euseius tularensis (Acari: Phytoseiidae) in citrus. Environ Entomol 24: Grafton-Cardwell EE, Ouyang Y (1996) Influence of citrus leaf nutrition on survivorship, sex ratio, and reproduction of Euseius tularensis (Acari: Phytoseiidae). Environ Entomol 25: Karg W (1961) Ökologische Untersuchungen von edaphischen Gamasiden (Acarina, Parasitiformes) (Part I and II). Pedobiologia 1:53 98 Kreiter S, Tixier MS, Croft BA, Auger P, Barret D (2002) Plants and leaf characteristics influencing the predaceous mite Kampimodromus aberrans (Acari: Phytoseiidae) in habitats surrounding vineyards. Environ Entomol 31: Magalhães S, Bakker FM (2002) Plant feeding by a predatory mite inhabiting cassava. Exp Appl Acarol 27:27 37 McMurtry JA (1992) Dynamics and potential impact of generalist phytoseiids in agroecosystems and possibilities for establishment of exotic species. Exp Appl Acarol 14: McMurtry JA, Croft BA (1997) Life-styles of phytoseiid mites and their roles in biological control. Annu Rev Entomol 42: McMurtry JA, Moraes GJ (2012) Revision of the life style system of phytoseiid mites. IOBC/wprs Bulletin (in press) McMurtry JA, Scriven GT (1966) Studies on predator-prey interactions between Amblyseius hibisci and Oligonychus punicae (Acarina: Phytoseiidae, Tetranychidae) under greenhouse conditions. Ann Entomol Soc Am 59: Nomikou M, Janssen A, Sabelis MW (2003) Phytoseiid predator of whitefly feeds on plant tissue. Exp Appl Acarol 31:27 36 Ochoa R, Beard JJ, Bauchan GR, Kane EC, Dowling APG, Erbe EF (2011) Herbivore exploits chink in armor of host. Am Entomol 57:26 29

17 Porres MA, McMurtry JA, March RB (1975) Investigations of leaf sap feeding by three species of Phytoseiid mites by labelling with radioactive phosphoric acid (H332PO4). Ann Entomol Soc Am 68: Sabelis MW, Van Rijn PCJ (2005) When does alternative food promote biological pest control? In: Hoddle M (ed) Second international symposium on biological control of arthropods, Davos, Switzerland, pp Sengonca C, Khan IA, Blaeser P (2004) The predatory mite Typhlodromus pyri (Acari: Phytoseiidae) causes feeding scars on leaves and fruits of apple. Exp Appl Acarol 33:45 53 Tanigoshi LK, Megevand B, Yaninek JS (1993) Non-prey food for subsistence of Amblyseius idaeus (Acari: Phytoseiidae) on cassava in Africa. Exp Appl Acarol 17:91 96 Zhao Z, McMurtry JA (1990) Development and reproduction of three Euseius (Acari: Phytoseiidae) species in the presence and absence of supplementary foods. Exp Appl Acarol 8: