Predation and oviposition rates of Gaeolaelaps aculeifer and Parasitus bituberosus (Acari: Laelapidae and Parasitidae) on pre-pupae/pupae of Thrips tabaci (Thysanoptera: Thripidae)

Thrips tabaci Lindeman is one of the main insect pests of onion (Allium cepa L.) in Colombia and several other countries. Strategies for its control are currently based on the use of chemical products. In a recent survey, Gaeolaelaps aculeifer (Canestrini) and Parasitus bituberosus Karg, two soil-dwelling predatory mite species (Acari: Mesostigmata), were found associated with this crop at Boyacá department, Colombia. Given that T. tabaci spends its pre-pupal and pupal stages in the soil, this study was conducted to evaluate the predation and oviposition of G. aculeifer and P. bituberosus on these developmental stages under laboratory conditions. The predators consumed up to 6.8 ± 0.52 and 6.9 ± 0.45 T. tabaci pre-pupae/ pupae, respectively, when offered 6 –10 prey a day. The maximum oviposition rates were 4.4 ± 0.25 eggs/female/day for G. aculeifer and 6.9 ± 0.26 eggs/female/day for P. bituberosus, with a mean egg viability higher than 91% for both predators. These results suggest that the evaluated predators may affect the population of T. tabaci under field conditions, and that the conduction of complementary studies on those predators is warranted, envisioning their practical use for T. tabaci control in Colombia.


Introduction
Thrips tabaci Lindeman is one of the main onion (Allium cepa L.) pests in Colombia as well as in several other countries (Rueda et al. 2007; Riley et al. 2011; Pal et al. 2019, due to its invasive capacity, high mobility and polyphagous behavior (Smith et al. 2011). The damage is caused by the feeding of nymphs and adults, which result in the appearance of silvery spots on the leaves that turn into white blotches, followed by the development of silvery patches and leaf curling (Waiganjo et al. 2008; Munoz et al. 2014. These injuries result in reduced photosynthesis (Jensen et al. 2003) and decreases in bulb size (Waiganjo et al. 2008). In addition, this species, as well as Frankliniella fusca (Hinds) (tobacco thrips), a less common species, have been reported as vectors of the Iris Yellow Spot Virus (IYSV) in onions, garlic, chives, leeks and several ornamentals (Srinivasan et al. 2012; Bag et al. 2015.
Worldwide, T. tabaci is controlled mainly by sprays of synthetic insecticides (Foster et al. 2010; Wu et al. 2014. However, given the pupal developmental stages are normally found in the litter or the topsoil layer (1.5-2.0 cm) (Tommasini and Maini 1995), these can be hidden and not affected by the applied insecticide (Cannon et al. 2007). This has led growers to intensify insecticide use, negatively affecting the environment, human health and promoting the selection of resistant thrips populations (Herron et al. 2008; Nazemi et al. 2016. In an attempt to break that cycle, alternative control strategies have been evaluated, as the use of biological control. Soil Mesostigmata are a diverse mite group, where numerous species are predators of small arthropods and nematodes (Lindquist et al. 2009; Carrillo et al. 2015. Mesostigmatid mites have been shown to prey on pre-pupae and pupae thrips (Wiethoff et al. 2004; Wu et al. 2014; Rueda-Ramírez et al. 2018, 2019, which would interrupt the cycle of these organisms, thereby preventing their return to the plants. Recent studies reported species of soil-dwelling predaceous mesostigmatids of the families Laelapidae and Parasitidae in onion growing areas of the Colombian Department of Boyacá (Castro-López 2018). Hence, it is presumable that some of these species play a natural role in thrips control in that region or could be utilized for that purpose.
One of the mesostigmatid species collected, Gaeolaelaps aculeifer (Canestrini) (Laelapidae), has been mass produced and commercialized in Africa, Asia, Europe, North America and Oceania for applied biological control of pest species, including thrips (Knapp et al. 2018; van Lenteren et al. 2018. It has been reported that each female of a Colombian population of this predator (collected in the Department of Cundinamarca) is able to consume daily 2.6 prepupae/pupae of the thrips Frankliniella occidentalis (Pergande) (Rueda-Ramírez et al. 2018). Previously, Berndt et al. (2004b) reported predation of G. aculeifer on F. occidentalis on substrate soil, and Navarro-Campos et al. (2012) on the thrips Pezothrips kellyanus (Bagnall) in citrus.
Parasitus bituberosus Karg (Parasitidae), found in Europe, Africa, Asia (Karg, 1972), and recently in Colombia , is also a potential biological control agent. This species is reported to prey on fly larvae, (Al-Amidi and Downes 1990; Al-Amidi et al. 1991; Szafranek et al. 2013 nematodes (Szafranek et al. 2013; Rueda-Ramírez et al. 2019) and pygmephorid mites (Szafranek et al. 2013). In Colombia it was recently reported to prey daily 4.4 pre-pupae and pupae of F. occidentalis , suggesting that it could be evaluated on other thrips species.
As the predation of G. aculeifer and P. bituberosus on T. tabaci has not been evaluated, the objective of this study was to determine their predation and oviposition rates on T. tabaci pre-pupae/pupae under laboratory conditions.

Material and Methods
This study was conducted between August of 2017 and February of 2018, at the Entomology Laboratory of the Biological Crop Management Group (GMBC), Universidad Pedagógica y Tecnológica de Colombia, Tunja, Boyacá, Colombia.
These mites were used to establish stock colonies in rearing units modified from Abbatiello (1965) and Freire and Moraes (2007). Each unit consisted of a plastic container (10 cm diameter x 7 cm high), whose bottom was covered with a layer of a plaster made of a mixture of nine parts gypsum and one part activated charcoal. Gaeolaelaps aculeifer was fed with a mixture of all developmental stages of the mite Aleuroglyphus ovatus (Troupeau) (Sarcoptiformes, Astigmatina, Acaridae) reared on crushed commercial dog food (Purina®; nutritional content: 9% fat, 12% moisture, 8% ash and 25% protein). Parasitus bituberosus was fed with Rhabditella axei nematodes reared on decomposing bean pods (Phaseolus vulgaris L.) and a mixture of all developmental stages of A. ovatus. The units were maintained in a growth chamber, in the dark, at 19 ± 3°C, 60 ± 10% RH. The units were closed with a piece of plastic film and maintained permanently moist by daily addition of distilled water.Thrips tabaci pre-pupae and pupae were obtained from a colony started with adult specimens collected from onion plants. The identification of the thrips species was confirmed by a Thysanoptera specialist, Dr. Everth Ebrat Ravelo. The colonies were maintained as described by López et al. (2007), in transparent plastic containers (11 cm diameter x 9 cm high), with bottom covered with a thin layer of cotton overlaid by a sheet of paper towel maintained humid daily with distilled water. Germinated faba bean (Vicia faba L.) seeds were placed onto the towel to serve as a substrate for insect feeding and oviposition. Each container was sealed with a transparent plastic film, to prevent thrips from escaping, and maintained in a growth chamber in the dark, at 19 ± 3 ºC and 60 ± 10 % RH.

Experimental procedure
For each predatory mites, three treatments were evaluated: a daily density of six, eight and ten of T. tabaci pre-pupae or pupae as prey. The experimental unit consisted of a plastic Petri dishes (4 cm diameter, 1.3 cm high), whose bottom was covered with plaster as previously described for the mite rearing units. Thirty gravid females of 3-8-day-old each mite species per treatment were individually transferred from the stock colony to each experimental unit, creating 30 repetitions for each treatment and species.
Experimental units with P. bituberosus were evaluated daily for eight days, while units with G. aculeifer for ten days. The differences in the evaluation time of these two species were due to differences in their longevity (Rueda-Ramírez et al. 2018, 2019. Units were evaluated counting the number of consumed prey and laid eggs. Consumed and non-consumed prey were replaced by new ones after each evaluation. Daily eggs laid were transferred into new containers and maintained until larval emergence, to assess egg viability. Experiments were conducted at controlled laboratory conditions identical to those previous described for mite colony maintenance.

Statistical analysis
Mean daily predation and oviposition rates, and mean egg viability (proportion of eggs hatched/female) were calculated. The predation rate was analyzed with a two-way (2 x 2) analysis of variance (ANOVA) in a factorial design, with three prey densities (densities of six, eight and ten T. tabaci pre-pupae/pupae provided daily per unit) and predators (G. aculeifer and P. bituberosus) as factors, as normality and homoscedasticity assumptions were met. Oviposition and egg viability were analyzed with a generalized linear model (GLM) with treatment and predator as factors and a quasi-binomial distribution. Post hoc Tukey's test was used for testing differences between means considering the best fit model. Statistical analyses were performed using the R program (Packages ExpDes.pt, lme4, multcomp and ggplot2, version 3.6.2, The R foundation for Statistical Computing, 2019-12-12).

Predation rate
Significant differences in predation rates were observed between prey densities (F = 311.6, d.f. = 2, P < 0.0005) but not between predator species at any prey density (F = 1.3, d.f. = 1, P = 0.25). Number of consumed preys was highest when ten prey were offered daily to G. aculeifer and to P. bituberosus.
For both species, predation rate increased with increasing prey density. The number of consumed individuals of T. tabaci pre-pupae / pupae by G. aculeifer was 6.8 ± 0.52, 6.4 ± 0.39 and 5.1 ± 0.33 and by P. bituberosus was 6.9 ± 0.45, 6.2 ± 0.35 and 5.1 ± 0.28 when ten, eight and six prey were offered daily, respectively. However, statistically predation rate of both predators did not tend to increase above a density of eight pre-pupae/pupae of T. tabaci, with a smaller difference between predation at eight and ten than between six and eight ( Figure 1).
Considering the predatory mite species and the prey density, the highest daily oviposition was observed for P. bituberosus (6.9 ± 0.26 eggs/female/day) when the offered prey density was 10 pre-pupae/pupae, and the lowest for G. aculeifer (3.6 ± 0.36 eggs/female/day) when the offered prey density was 6 pre-pupae/pupae of T. tabaci. The highest daily oviposition for G. aculeifer (4.4 ± 0.25 eggs/female/day) was also observed when the offered prey density was 10 pre-pupae/pupae of T. tabaci. A significant positive correlation was observed between daily oviposition rate and the number of T. tabaci pre-pupae/pupae preyed by both predators (r = 0.8, P < 0.001 and r = 0.53, P < 0.001 for P. bituberosus and G. aculeifer, respectively).

Mean egg viability
Significant differences between the viability of eggs produced by females offered different prey densities were recorded (Chi 2 = 22.02, d.f. = 2, P = 1.65 x 10 -5 ), but no significant differences  were observed between predator species (Figure 3). Viability was lowest (86 ± 5.8 %) for G. aculeifer offered six prey day compared to other combinations of predator and prey densities, which did not differ significantly among themselves (> 91%).
In summary, for both G. aculeifer and P. bituberosus, predation in the current study on T. tabaci was approximately 1.6 times higher than F. occidentalis in the cited studies. These differences in predation rate may be related to characteristics of T. tabaci. Shaikh et al. (2015) reported the length of pre-pupae and pupae of T. tabaci to be respectively 0.91 ± 0.10 mm and 0.96 ± 0.12 mm, whereas Cárdenas and Corredor (1989) reported for F. occidentalis length of respectively about 1.1 mm and 1.3 mm. The smaller size of T. tabaci could account for the greater predation rate than the former. However, in addition to size, other factors could be involved, resulting from the metabolism of organic acids as glutamic, malic, citric, in onion plants (Rodríguez-Galdón et al. 2008). These hypotheses need to be further explored.
The oviposition rates of G. aculeifer were higher when fed on T. tabaci pre-pupae/pupae G. aculeifer P. bituberosus than the rates reported by Rueda-Ramírez et al. (2018), Navarro-Campos et al. (2016) and Berndt et al. (2004a) when this species was fed with F. occidentalis (about 2.9 ± 0.1,2.2 and 2.5 ± 0.87 eggs/female/day, respectively). Another laelapid species as C. jaboticabalensis, only laid 0.2 eggs/female/day (Moreira et al. 2015) and S. miles with 0.8 ± 0.53 eggs/female/day when fed F. occidentalis. In the case of P. bituberosus, the oviposition rate was lower than that reported by Rueda-Ramírez et al. (2019) when F. occidentalis pre-pupae/pupae was offered as prey (8.9 ± 0.8 eggs/female/day). The oviposition capacity depends on several factors, among which the nutritional content of the food source and the use of resources to extend longevity in presence of nutrient-poor or stressful conditions for predatory species (Gotoh and Tsuchiya 2008). Both factors need to be analyzed. McMurtry (1982) indicated that the growth capacity of a population is only one feature determining the performance of a biological control agent; several other factors may influence the efficacy of predators such as intraguild predation, competition, response to abiotic environmental factors, functional and numerical response, and others (Skirvin andFenlon 2001; Gontijo et al. 2012), thus further experiments should be performed. The difficulty in obtaining large numbers of T. tabaci to conduct this study hampered the possibility to evaluate a larger array of prey densities or a larger number of replicates. Under greenhouse conditions, a reduction of 78% and 72% were observed in the population of T. tabaci in the presence of G aculeifer and P. bituberosus, respectively in onion plants (Castro-López and Martínez-Osorio 2021). However, complementary field studies should be conducted to explore other possibilities, including the association of T. tabaci pre-pupae and pupae with other food sources such as nematodes (Rueda-Ramírez et al. 2019; Azevedo et al. 2019, 2020, naturally found in agricultural ecosystems (Rueda-Ramírez et al. 2018, 2019. The use of entomopathogenic fungi in combination with predatory mites has shown good potential, as shown in the study conducted by Saito and Brownbridge (2016), in which mortality of thrips was higher than 90%. These complementary studies should lead to the practical use of G. aculeifer and or P. bituberosus in Colombia where thrips species are still controlled primarily with insecticides.
In conclusion, while previous studies reported on the predation of G. aculeifer and P. bituberosus of edaphic phases of thrips species (Berndt et al. 2004a; Navarro-Campos et al. 2012; Rueda-Ramírez et al. 2018, 2019, this is the first study demonstrating the predation capacity on T. tabaci pre-pupae/pupae.