Neoseiulus californicus (McGregor, 1954)

Mikawa, Yuya, Munkhtumur, Mungunzaya, Kawamura, Taichi, Yokoyama, Akio, Mori, Kotaro, K, Masatoshi Toyama, K, Mineaki Aizawa & SonodaK, Shoji, 2025, Effects of pesticide application and Neoseiulus barkeri on spider mite control using commercialized Neoseiulus californicus in a Japanese pear greenhouse, Acarologia 65 (3), pp. 771-781 : 772-775

Commercialized N. californicus

The predator-release device (Miyako Banker®; ISK Bioscience K.K.) consists of a waterproof shelter (Banker-sheet®; ISK Biosciences K.K.) holding a sachet of N. californicus (Biobest Belgium N.V., Antwerp, Belgium) (58 × 72 × ca. 5 mm), a black felt patch (trichome-mimicking material) (50 × 100 × 1 mm), and five water-absorbing polymers (moisturizing agents). The sachet contains N. californicus (ca. 100 individuals per sachet) and its prey, Glycyphagus destructor (Schrank) (Acari: Glycyphagidae ), with bran as a bulking agent. The black felt patch and water-absorbing polymers within the shelter are expected to function respectively as oviposition and resting sites for N. californicus and humidity retainers ( Shimoda et al.

2017, 2019). The manufacturer’s recommendation for the installation was 100 predator-release devices per 1,000 m 2.

Study site

Population surveys of phytoseiid mites and spider mites were conducted during 2019–2021 in a commercial Japanese pear greenhouse with a polyvinyl chloride plastic cover. The greenhouse had windows on the ceiling that opened and closed automatically, with manually operated windows on the north and south sides. Japanese pear trees were trained on overhead horizontal trellis material, with a canopy height of ca. 1.8 m in the greenhouse. The exact age of each tree is unknown, but they were all more than 2.5 m tall; all produced fruits for sale. At the time of fruit thinning, the leaves of the lateral branches partially touched adjacent pear trees.

No noticeable difference in the growth rates of Japanese pear trees was observed during the three surveyed years (data not shown). Several old trees in the greenhouse were replaced with younger trees before the 2019 and 2020 growing seasons. The greenhouse size (ca. 3,800

m 2) of 2019 and 2020 was reduced to ca. 3,500 m 2 in 2021 because of conversion to grape production. In total, 400 predator-release devices were installed in the greenhouse on March

28, 2019 and March 25, 2021. In 2020, 200 predator-release devices were installed on March

26 and another 200 on April 30. One to three predator-release devices were placed around the trunks at ca. 1.0 m above the ground, depending on the tree size. Wild ground cover vegetation in the greenhouse was managed approximately once a month using a glyphosate herbicide (Roundup; Monsanto Co., Creve Coeur, MO, USA) and a mowing machine during the cultivation period each year. The pesticides highly toxic to N. californicus among those used in the greenhouse are presented in Table 1 ( Table S1 provides all pesticides used) ( Yoshimura et al. 2022). The pesticide types and their application schedules at the study site did not change to any considerable degree after introduction of the predator-release devices in 2017 (data not shown).

Sampling procedure

Sampling was conducted in rows 1, 3, 5, 7, 9, and 13 out of the 13 rows from the entrance in

2019, in rows 2, 4, 6, 8, 10, and 12 out of the 13 rows from the entrance in 2020, and in rows 2,

4, 6, 8, 10, and 12 out of the 12 rows from the entrance in 2021. In each row, 15–20 trees were present. In each survey year, 20 leaves were collected randomly from four fixed trees in each row, i.e., a total of 480 leaves were collected for each sampling event. Sampling was conducted at 5-day to 14-day intervals during March 28 – October 30 in 2019, March 26 – October 28 in 2020, and April 5 – October 27 in 2021 (32 samplings in 2019 and 2020 and 28 samplings in

2021).

Leaves were brushed using a brushing machine (Daiki Co., Ltd., Konosu, Japan) to collect mites in each sampling event on a Petri dish filled with 70% ethanol. Using a binocular microscope (SZY7; Olympus Corp., Tokyo, Japan), phytoseiid mites were separated from spider mites. Until DNA extraction, phytoseiid mites separated in each sampling event were stored, irrespective of their sex and developmental stage, in a glass container (4 ml) that had been filled with 3–3.5 ml of 99.5% ethanol. In 2019 and 2020, phytoseiid mites collected in each sampling event were mixed and stored together, but in 2021, the phytoseiid mites collected in each sampling event were mixed for each survey tree and stored. All collected spider mites were determined to be T. urticae based on morphological characterization using the previously described microscope.

DNA extraction

Genomic DNA was extracted from phytoseiid mites collected during the survey period in 2019 (974 individuals), 2020 (1,062 individuals), and 2021 (2,521 individuals) using PrepMan Ultra (Applied Biomaterials, Warrington, UK) or MightyPrep reagent (Takara Bio Inc., Kusatsu, Japan). Briefly, a single phytoseiid mite introduced into 10 μL of the reagent was incubated at 95 °C for 10 min and then at room temperature for 2 min. After centrifugation at 15,000 × g

for 2 min, the supernatant was recovered as a DNA sample. For subsequent PCR amplification, 0.2–0.5 µL of the supernatant was used.

PCR amplification for discrimination of phytoseiid mite species

In 2019 and 2020, ribosomal internal transcribed spacer (ITS) sequences were used to identify N. californicus , N. barkeri , Neoseiulus womersleyi Schicha , Amblyseius eharai Amitai and Swirskii , Amblyseius swirskii Athias-Henriot , and Gynaeseius liturivorus (Ehara) using PCR,

with species-specific primer sets, as described by Mikawa et al. (2019). For samples collected in 2021, species identification was conducted for N. californicus and N. barkeri . Quick Taq

HS DyeMix (Toyobo Co., Ltd., Osaka, Japan) or EmeraldAmp MAX PCR Master Mix (Takara Bio Inc.) was used for PCR. The PCR conditions were the following: 1 cycle of 3 min at 94 °C, followed by 40 cycles of 15 s at 94 °C, 30 s at 60 °C, and 1 min at 72 °C, with final extension at 72 °C for 7 min.

Microsatellite genotyping and data analysis

Genomic DNA extracted individually from adult N. californicus females (417, 278, and

432 individuals, respectively, in 2019, 2020, and 2021) was used for nuclear microsatellite genotyping. Also, 24 naturally occurring samples collected in the greenhouse in 2017 and commercialized N. californicus samples reported by Mikawa et al. (2020) (35, 30, and 75 samples in 2019, 2020, and 2021, respectively) were used as references for the analyses. Adult N. californicus females collected in 2019, 2020, and 2021 were divided respectively into 18, 11, and 18 subpopulations, fundamentally based on the sampling date or month as presented below.

2019: April (April 24 and April 29), May 8, May 15, May 22, May 29, June 5, June 12,

June 19, June 26, July 3, July 10, July 17, July 24, August 14, August 19, August 30, September

4, and September 11;

2020: May 7, May 13, May 20, May 27, June 3, June 10, June 17, June 24, June 30, July 8,

and July 15;

2021: April 12, April 19, April 26, May 12, May 26, June 2, June 9, June 18, June 25, July 7,

July 14, July 20, July 28, August 4, August 11, August 18, August 25, and September/October (September 22, October 6, October 13, October 20, and October 27).

Multiplex PCR using five nuclear microsatellite loci described in an earlier paper ( Mikawa et al. 2020) was conducted using a PCR kit (Type-it Microsatellite; Qiagen Inc., Hilden, Germany). The PCR conditions were the following: 1 cycle of 5 min at 95 °C, followed by 40 cycles of 30 s at 95 °C, 90 s at 63 °C, and 30 s at 72 °C, with final extension for 30 min at 60 °C.

The amplified PCR products were analyzed using a DNA sequencer (3500 Genetic Analyzer; Applied Biosystems). Genotypes were determined using software (GeneMapper v.4.1; Applied

Biosystems).

The number of alleles (N A), observed H (O) and expected H (E) heterozygosities, and the fixation index (F IS) were calculated using GenAlEx v.6.503 ( Peakall and Smouse 2006).

The deviation of F IS from the Hardy–Weinberg equilibrium was tested using FSTAT v.2.9.3.2

( Goudet 2001). The frequencies of null alleles were estimated using Cervus v.3.0.7 ( Kalinowski et al. 2007).

Genetic structure was assessed using Bayesian clustering analysis implemented using

InStruct ( Gao et al. 2007) because inbreeding, along with deviation from the Hardy–Weinberg equilibrium and the presence of linkage disequilibrium, were observed, as described herein in

Results. The results of 10 independent chains (runs) at K = 2, on which naturally occurring and commercialized N. californicus were mostly assigned to different clusters ( Mikawa et al. 2020),

were integrated using CLUMPP v. 1.1 ( Jakobsson and Rosenberg 2007). Commercialized and naturally occurring samples were mostly assigned, respectively to Clusters 1 and 2 (data not shown). The results were presented as vertical bar plots (Excel; Microsoft Corp., Redmond,

WA, USA).

Statistical analysis

To evaluate the effects of N. californicus and N. barkeri on spider mite occurrence, a generalized linear mixed model (GLMM) with Poisson distribution and Gauss–Hermite quadrature method was constructed using the “glmmML» package in R ver.4.3.2 (R Core Team 2023). The number of spider mites was used as the response variable. The number of N. californicus and the number of N. barkeri were used as fixed effects. The trees from which leaves were collected were used as random effects. Analyses were conducted using data on each sampling date showing prominent spider mite occurrences (July 7 – August 18) in 2021.