The gall mite, Aceria pallida Keifer (Acari: Eriophyoidae) is an economically important pest of wolfberry Lycium barbarum L. and can cause significant losses to plant production. Two species of phytoseiid predatory mites, Amblyseius andersoni Chant and Neoseiulus neoreticuloides Liang & Hu were found on L. barbarum in Bayan Nur city, Inner Mongolia, China. We assessed the potential of these two phytoseiid species as biological control agents against A. pallida, using functional response experiments with seven prey densities (5, 10, 20, 40, 60, 80 and 100 adults of A. pallida) on a wolfberry leaf under 25°C ± 1°C, 60% ± 5% RH and a 16:8 h (L:D) photoperiod. Overall, the predation of both species increased with increase in prey density. The consumption of A. andersoni female was significantly greater than that of conspecific males and N. neoreticuloides female at high prey densities. Both phytoseiid species displayed a type II functional response to A. pallida. Female A. andersoni had a higher attack rate (5.961) and a shorter handling time (0.014 d) than male A. andersoni (1.619; 0.019 d) and female N. neoreticuloides (0.719; 0.023 d). The maximum attack rate (T/Th) was estimated to be 71.43 and 52.63 for female and male A. andersoni, respectively, while it was 43.48 for female N. neoreticuloides. Both female and male A. andersoni consistently consumed significantly more prey than N. neoreticuloides across all densities of A. pallida adults.
Introduction
Wolfberry, Lycium barbarum L. is used as a medicinal plant and food in China. The gall mite, Aceria pallida Keifer (Eriophyoidae) is an economically important phytophagous mite associated with wolfberry (Rong & Wang 1983; Wu et al. 2017; Liu et al. 2019b), in the major wolfberry production areas of China (Kuang 1983; Rong & Wang 1983; Zhang et al. 2000; Wu et al. 2017). Like most eriophyoid mites (Keifer et al. 1982; Westphal & Manson 1996), A. pallida caused the gall formation on the plant and other abnormalities resulting in loss of plant growth (Wu et al. 2017). Aceria pallida is mainly controlled using acaricide applications (Xu et al. 2014). However, A. pallida completes most of its development in galls, and this protects it from acaricide treatments (Xu & Duan 2005). Biological control, using predatory mites that mainly feed on tetranychid mites and eriophyoid mites (McMurtry & Croft 1997), can be a viable alternative to chemical control and can be used to control A. pallida.
Two species of predatory mites, Amblyseius andersoni Chant and Neoseiulus neoreticuloides Liang & Hu were found on wild Chinese wolfberry, L. barbarum, in Inner Mongolia (Liu et al. 2019a). Amblyseius andersoni is a type III generalist predator that feeds on a variety of prey and non-prey foods (Zhang & Sanderson 1993; Koveos and Broufas 2000; Duso et al. 2011; Lorenzon et al. 2012; Li et al. 2019; McMurtry et al. 2013), and it is used in the biological control of spider mites (Tetranychidae) in orchards (Markó et al. 2012; Szabó & Pénzes 2013). Neoseiulus neoreticuloides was initially described on elms in China (Liang & Hu 1988), but no further studies have been conducted on this species. However, several species of the genus Neoseiulus are known for their potential as biological control agents of eriophyoid mites (McMurtry & Croft 1997). One example is Neoseiulus baraki Athias-Henriot that controls Aceria guerreronis Keifer (Lima et al. 2015).
Thus, it is worthy of interest and useful to be considered in the general scenario of control of A. pallida, one of the most serious pests on wolfberry, by exploring effective predatory mites that were found to co-occur with this eriophyid species. This study evaluated the predatory potential of two phytoseiid species—A. andersoni and N. neoreticuloides—on the gall mite A. pallida, basing functional response experiments.
Functional response (FS) experiments are used to evaluate the effectiveness of a predator for controlling a particular prey species (Xiao & Fadamiro 2010). It describes the relationship between consumption rate of an individual predator and different prey density (Farazmand et al. 2012), and can be classified into three types (Holling 1959): Type I FS (the number of prey consumed increases linearly to a maximum, then remains constant as prey density increases), Type II FS (the prey consumption increases as prey density increases but at a decelerating rate towards an asymptote), and Type III FS (the number of prey eaten approaches an asymptote as a sigmoid function).
Materials and methods
Predator source and rearing
Colonies of A. andersoni and N. neoreticuloides were initiated from specimens collected in September 2017 on a Chinese wolfberry plant at Bayan Nur city (40°28′N–108°11′E), Inner Mongolia, China. Both predator species were maintained with all stages of the fruit mite, Carpoglyphus lactis L. on refined sugar and yeast. The rearing unit consisted of a piece of plastic film spread on a water-saturated foam, which sat in a 30×20×10 cm plastic box half-filled with water to prevent mites from escaping. All rearing units were maintained under laboratory conditions (25°C ± 1°C, 60% ± 5% RH and a 16:8 h (L:D) photoperiod) since they were collected in September 2017.
Prey culture
The colony of A. pallida was obtained from Chinese wolfberry at the same location of predatory mite collection. Wolfberry branches with mite galls were collected and stored at 4°. The mites within the galls were directly used in the following experiments. Aceria pallida are very small (the adult is 200–240 µm) (Kuang 1983), and the juvenile stage is difficult to distinguish under stereoscope. Thus, only adults of A. pallida, which are distinguishable by yellow-orange colour, were used in the experiments. These were collected from galls with diameters of 2–4 mm as described in Wu et al. (2017).
Experimental arenas
The experimental arena used was modified from Munger cells (Nguyen et al. 2013). From top to bottom, the cells were successively made up of a transparent acrylic board (top board; 20×35 mm, 2-mm thick), another transparent acrylic board (middle board; 20×35 mm, 2-mm thick) with a 10-mm diameter hole in the centre, a wolfberry leaf (20× 35 mm), a piece of wet filter paper (20× 35 mm) and another transparent acrylic board (bottom board; 20×35 mm, 2-mm thick) with a 10-mm diameter hole in the centre. The top board, middle board, wolfberry leaf and bottom board were tightly clamped together with two metal clips to form an enclosed cell. The filter paper enabled the leaf to remain fresh during the experiment. The cells were placed on a PVC tray covered with wet gauze.
Functional response
The predatory mites were shifted from the stock colony to a similar rearing unit and fed with all stages A. pallida for three successive generations (30 days) before being used in the experiments. Three-day-old males and gravid females of A. andersoni, and three-day-old gravid females of N. neoreticuloides were randomly selected from the cultures and individually transferred into the experimental cells. The reason for the omission of male N. neoreticuloides was that it had lower consumption for A. pallida adults and the data couldn't fit equation models to discriminate the type of functional response (unpublished data). The predatory mites were starved for 24 h before being used in the experiments. Female mites who had oviposited during starvation were placed singly in the new cell. Adults of A. pallida were offered as prey to male and female adults. Each cell was randomly subjected to one of the seven densities of A. palida (5, 10, 20, 40, 60, 80 and 100 adults), which were transferred from the galls into the cells with a fine camel-hair brush. Then females/males of A. andersoni or N. neoreticuloides were allowed to feed for 24 h, and the number of prey eaten was recorded. Each density treatment was replicated 10 times.
Statistical analysis
To discriminate between type II and type III functional responses, a polynomial logistic regression was performed between the proportion of prey consumed (Ne/N0) and initial prey density (N0) (Juliano 2001; Timms et al. 2008), using SigmaPlot 12.5 software (SigmaPlot 2013).
where Ne = number of prey consumed, N0 = initial number of prey, a = intercept, b = linear, c = quadratic and d = cubic coefficient. From the equations obtained for the proportions of prey consumed, the linear coefficients (b) and the quadratic coefficient (c) were observed, thus allowing determination of the type of functional response. If b < 0, the functional response is type II, if b > 0 and c < 0, the functional response is type III. The handling time and attack rate were estimated using non-linear least squares regressions from PROC NLIN of SAS (SAS Institute 2002). Since prey items were not replaced during the experimental period, the random predator equation of Rogers was appropriate to describe the type II functional response parameters (2) (Rogers 1972; Mendes et al. 2018):
where α = attack rate, Th = handling time and T = experimental time. The values of α and Th were compared between two predators using the 95% confidence intervals. Two-way ANOVA followed by Tukey's multiple comparisons test was applied to compare the consumption of predators among prey densities within and between predator species.
Results
Predation of A. andersoni and N. neoreticuloides
The predation of the two predators increased significantly with increasing prey density (female A. andersoni: P <0.0001; male A. andersoni: P <0.0001; female N. neoreticuloides: P <0.0001). Female A. andersoni consumed significantly more prey than male A. andersoni and female N. neoreticuloides (F12, 189 = 257.8, P <0.0001). The interaction between predator sex for A. andersoni and gall mite density was also significant for the number of prey eaten (F6,126 = 232.6, P <0.0001). The maximum number of prey killed was 57.6, 33.5 and 23.7 for female and male A. andersoni, and female N. neoreticuloides, respectively (Table 1).
TABLE 1.
Consumption (mean ± SE) of A. pallida adults by A. andersoni and N. neoreticuloides at different prey densities
Functional response of A. andersoni and N. neoreticuloides to adult A. pallida
Both phytoseiid species displayed a Type II functional response to densities of A. pallida adults (b<0) (Figure 1; Table 2). The attack rate (α) of female A. andersoni was 3-fold higher and the handling time (Th) was 0.75-fold lower than that of males (Table 3). Female A. andersoni had an attack rate 8-fold higher and a handling time shorter than female N. neoreticuloides (Table 3). The maximum attack rate (T/Th) was estimated to be 71.43 for female A. andersoni, 52.63 for male A. andersoni and 43.48 for female N. neoreticuloides.
TABLE 2.
Estimates of coefficients in a binomial logistic regression of the proportion of A. pallida adults consumed by A. andersoni and N. neoreticuloides as a function of initial prey density.
TABLE 3.
Attack rate and handling time (days) (means ± SE) for A. andersoni and N. neoreticuloides feeding on A. pallida adults.
Discussion
This is the first study of the functional responses of the predatory mites A. andersoni and N. neoreticuloides to A. pallida density. Amblyseius andersoni consumed more prey than N. neoreticuloides did across all prey densities. Our results suggest that A. andersoni have an effective predatory potential for the control A. pallida on wolfberry plants in China.
The two phytoseiid species displayed a Type II functional response to A. pallida, similar to that of A. andersoni when fed on Panonychus ulmi Koch (Tetranychidae) (Koveos & Broufas 2000). A type II functional response is common in phytoseiid mites (Afshar & Latifi 2017; Alfaia et al. 2018; Barbosa et al. 2019). Phytoseiid predators with a type II functional response, such as Neoseiulus californicus McGregor, Neoseiulus cucumeris Oudemans, Neoseiulus barkeri Hughes and Amblyseius swirskii Athias-Henriot, were proved to be efficient for control of pest organisms, especially at low prey densities (Koehler 1999; Jafari et al. 2012; van Lenteren 2012; Calvo et al. 2015; Song et al. 2016; Patel and Zhang 2017; Akyazi & Liburd 2019; Bazgir et al. 2020). The attack rate and handling time determine the magnitude of the functional response (Pervez & Omkar 2005). Female A. andersoni response to A. pallida included a higher attack rate (5.961) and a shorter handling time (0.014 d) than other phytoseiid biocontrol agents, such as A. swirskii to Eotetranychus frosti McGregor eggs (0.1142, 0.4858h) (Bazgir et al. 2020), and Neoseiulus womersleyi Schicha to Tetranychus urticae Koch eggs at 25° (5.467, 0.056 d) (Sugawara et al. 2018). The maximum attack rate representing predation capacity is obtained by dividing the experiment time (1 day) on handling time (T/Th) (Fathipour et al. 2017). As the handling time decreased, predation capacity during one day increased. Female A. andersoni had a higher maximum attack rate (57.6) than male A. andersoni (33.5), or female N. neoreticuloides (23.7).
Several factors can affect the functional response and predation rate, including the sex of the predator (Parajulee et al. 1994). Female A. andersoni consumed more prey and showed a higher attack rate and a shorter handling time than did conspecific males. A similar trend was observed for female and male N. cucumeris feeding on different stages of Bemisia tabaci Gennadius (Li et al. 2017). Functional responses can be influenced by physical characteristics of the host plant (Koveos & Broufas 2000; Ahn et al. 2010). Thus we conducted the experiments on leaves of wolfberry, the original habitat of predator and prey. To avoid the deviation of feeding experience (Castagnoli & Simoni 1999; Mendes et al. 2018), both phytoseiid species were uniformly fed with wolfberry mite galls for 30 days prior to bioassay. In addition, the type of functional response is not constant, which can be influenced by the alternative food and the stage of prey (Ganjisaffar & Perring 2015; Li et al. 2018; Fathipour et al. 2020). Due to the small size of A. pallida (Kuang 1983), only adult A. pallida were provided in the present study.
Nevertheless, it is difficult to replicate a natural environment by a simple experimental arena (O'Neil 1989). Therefore, further field studies will be necessary to determine the efficiency of each species of predator under more realistic conditions for offering a reference point by using native predatory mites as biocontrol agents against A. pallida.
Acknowledgements
This study was supported by the National Key R&D Program of China (2017YFD0201000), Science and Technology Project of Inner Mongolia (2020GG0065) and Natural Science Foundation of Inner Mongolia (2019MS03018).