Life-style classification of some Phytoseiidae (Acari: Mesostigmata) species based on gnathosoma morphometrics 

Abstract

Life style of phytoseiid species (Acari: Mesostigmata) is an important feature in evaluating their potential as biological control agents. Previous studies suggested that some morphometrics of gnathosoma are related to the feeding habits of phytoseiids. In this study, we showed it is appropriate to estimate feeding habits and/or life styles of phytoseiids based on variations in gnathosoma morphology. Scanning electron microscope pictures of 10 phytoseiid species belonging to four major life styles were taken. For each species, 23 morphometrics of gnathosoma and the length of dorsal shield were measured. Principle Component Analysis (PCA) showed that the 10 species can be divided into three groups based on the first two principle components of these parameters. The 1 st group includes one specialized predators of Tetranychus species (Phytoseiulus persimilis) and two selective predators of tetranychid mites (Neoseiulus pseudolongispinosus, Neoseiulus californicus), the 2 nd group includes six generalist predators (Amblyseius orientalis, Amblyseius swirskii, Amblyseius tsugawai, Neoseiulus barkeri, Neoseiulus bicaudus, Neoseiulus cucumeris), and the 3 rd group includes one specialized pollen feeder (Euseius utilis). This grouping is consistent with the life style division of McMurtry & Croft (1997), except that Type I and II were not separated. However, there is a trend that both Type II species were closer to each other. Overall, species in Group II have larger chelicerae and hypostome than those of Group I and III, while species in Group III have larger lobes and angle of fixed digit than those of the other two groups.

Feeding design in free-living mesostigmatid chelicerae (Acari: Anactinotrichida)

“…Despite the clue about aspect ratios in Table 2 of Buryn and Brandl (1992), that their second principal component (PC2) differentially weights the (size-adjusted) width of chelal measures against their (size-adjusted) lengths, they did not use a mechanical model. This study deploys one and, so far as practical, re-analyses their data, testing it against phytophagous mite-consuming phytoseiids (Adar et al 2012;Liu et al 2017). Differences in bill musculature is correlated with differences in adductive power in finch feeding (Bowman 1961)-so explicitly including an estimated crushing force allows adding conclusions around soft-body mite/collembola prey feeding specialisms in mesostigmatids that Buryn and Brandl (1992) could not make easily (and also to make omnivory easy to explain).…”

“…In retrospect, this is rather surprising given the earlier success of relating chelal form to function in cryptostigmatids (Schuster 1956;Kaneko 1988;Perdomo et al 2012) and astigmatids (Akimov and Gaichenko 1976). Moreover, later work (Adar et al 2012;Liu et al 2017) focusing on a narrow set of phytoseiid plant-pest predators has suggested that some morphometrics of mainly the chelicerae of their gnathosoma are related to their feeding habits.…”

“…Although Walter and Ikonen (1989) concludes that feeding behaviours in grassland soil mesostigmatids are rarely predictable above the generic level, given that most other animals' bodies suit their life-style and home (Manton 1958a), and that certainly vertebrate animal jaws and teeth show adaptations for food or morsel processing, it is worth trying to extend and augment Buryn and Brandl (1992)'s approach to see if stronger further insights can be gained. Furthermore to validate these new insights using the data from Adar et al (2012) and Liu et al (2017) for phytoseiids whose cheliceral morphologies are historically hard to correlate with feeding types a priori (McMurtry 1982). An extra objective is also to seek to better understand the trophic design for phytoseiid pollenophagy-a common habit in that group, where they digest the exine externally and ingest resultant fluids (Walter and Proctor 2013).…”

“…The test data set comprised those without feeding type designation from Buryn and Brandl (1992) plus the eight additional species in this study. Validation datasets were constructed from Adar et al (2012) and Liu et al (2017) as in Table 1 Lower, and from Hyatt (1980) as described in the text.…”

“…Predictions were then also made for each of the extra eight species (A. halleri, V. nemorensis, G. confusa being used for explicit validation) as well as all previously studied mites of unknown status given their data, and each mite allocated the most probable feeding type class. The validation datasets (Adar et al 2012;Liu et al 2017) were plotted on the resultant displays. Data for a validation cohort of parasitines was extracted from illustrations in Hyatt (1980) and means calculated as in text.…”