As a Potential Control Agent of Tetranychus Urticae (Acari: Tetranychidae): Eﬀect of Pest/Predator Ratio on Pest Abundance on Strawberry

Experimental and Applied Acarology (2005) 37:57–66 DOI 10.1007/s10493-005-0067-7 Springer 2005

-1

Neoseiulus californicus (Acari: Phytoseiidae) as a potential control agent of Tetranychus urticae (Acari: Tetranychidae): eﬀect of pest/predator ratio on pest abundance on strawberry

NANCY M. GRECO*, NORMA E. SA´NCHEZ and GERARDO G. LILJESTHRO¨ M Centro de Estudios Parasitolo´gicos y de Vectores (UNLP-CONICET), CEPAVE, 2 No. 584 (1900), La Plata, Argentina; *Author for correspondence (e-mail: [email protected]; phone: +54- 021-233471; fax: +54-21-232327)

Received 21 February 2005; accepted in revised form 20 June 2005

Key words: Conservation biological control, Neoseiulus californicus, Predator/prey ratios, Straw- berry, Tetranychus urticae

Abstract. Neoseiulus californicus (McGregor) is a promising agent for successful Tetranychus urticae Koch control through conservation techniques, in strawberry crops in La Plata (Buenos Aires, Argentina). In prey–predator interaction, initial relative densities have an important effect on system dynamics. The economic threshold level (ETL) used for this pest in the present study was 50 active mites per leaflet. In our laboratory experiments, initial T. urticae to N. californicus ratio had a significant effect on the population abundance of T. urticae at a 7-day period. When pest/predator ratio was 5/1 (at initial pest densities from 5 to 15 females/leaflet) the final number of active T. urticae/leaflet was significantly lower than the ETL, while at 20 females/leaflet this number did not differ from the ETL. At 7.5/1 ratio, the final number of active T. urticae/leaflet, at initial pest densities from 5 to 15 females/leaflet, reached the ETL without surpassing it. At 10/1 and 15/1 ratios, pest densities exceeded the ETL only at 15 initial T. urticae/leaflet. Most greenhouse and field observations were consistent with the predictions of a graphical model based on experimental results. This predator was very effective in limiting pest densities at a 7-day period and within the range of pest–predator ratios and absolute densities used in this study. Conservation of N. californicus promoting favorable pest/predator ratios may result in early control of T. urticae.

Introduction

Predaceous phytoseiid mites are important biological control agents of pest mites aﬀecting many crops in diﬀerent parts of the world (Helle and Sabelis 1985). In some agroecosystems, these natural enemies may drive spider mite numbers below economically damaging levels (Nyrop et al. 1998). Phytoseiulus persimilis Athias-Henriot, Neoseiulus fallacis (Garman) and N. californicus (McGregor) are commonly used through augmentative techniques, to control the two-spotted spider mite Tetranychus urticae Koch, in several countries (Hussey and Scopes 1985; Waite 1988; Raworth 1990; Zalom et al. 1990; Coop et al. 1997; Steinberg et al. 1999; Zalom 2002). 58

In commercial strawberry crops in La Plata (Buenos Aires, Argentina), the two-spotted spider mite is the most serious arthropod pest and N. californicus is its main established predator. It has a broader food range and a higher capacity to stay in patches with low pest density than other phytoseiids (Wei-Lan and Laing 1973; Croft et al. 1997; Pratt et al. 1999). Moreover, in this area, N. californicus was observed exhibiting a high spatial coincidence with T. urticae and a high ability to detect leaflets with prey (Greco et al. 1999). All these attributes suggest that it would be a promising natural enemy for successful T. urticae control through augmentative or conservation techniques. A current control method for this pest consists in the use of miticides on a calendar basis. Many growers have recently started to decrease their use in order to cut costs or with the purpose of obtaining chemical-free production. Conservation biological control would be the most appropriate alternative considering the need to reduce inputs in this region. Indeed, the natural occurrence of N. californicus on strawberry crops in Spain, due to the selective use of pesticides, was sufficient for T. urticae control (Ribbes 1990; Garcıá Mari et al. 1991). In prey–predator interactions, initial relative densities have an important effect on system dynamics (Shaw 1985). The levels of control of T. urticae attained by phytoseiid mites depend on prey to predator ratios favorable to the predator (Strong and Croft 1995), among other factors. Therefore, investi- gating the effect of T urticae/N. californicus ratio on the pest abundance would be very important to develop a biological control protocol for T. urticae in La Plata strawberry crops. The knowledge of effective T. urticae/N. californicus ratios would allow growers to make control decisions based on weekly monitoring schedule of pest and predator, using miticides only when N. californicus predation cannot prevent pest damage. Considering that the pest population has an exponential growth we would expect that different T. urticae/N. californicus ratios be required to achieve a satisfactory control at different initial pest densities. Our objectives were: (1) to determine the effect of initial T. urticae and N. californicus ratio on T. urticae density after a week interval, in laboratory experiments, and (2) to examine the experimental laboratory results in relation to field samples.

Materials and methods

Laboratory experiments

Tetranychus urticae and N. californicus colonies were reared in an experimental greenhouse on strawberry plants (Fragaria X ananassa). The short-day cultivar ‘Camarosa’ was used. Experiments were conducted under controlled conditions of 25 ± 2 C, 60– 70% relative humidity and 14L/10D photoperiod. The experimental unit was a 59 new and completely expanded trifoliate leaf. Each petiole was placed in water- filled tubes (height 7 cm, diameter 2 cm) to keep the leaves turgid. Tubes were placed in plastic containers (500 ml) covered with plastic film. The experiments were planned as a completely randomized design and each treatment was replicated between 8 and 15 times. One day old T. urticae females and recently mated N. californicus females were used for the experiments. Spider mites and predators were transferred to the leaflets with a very fine brush. The experimental initial pest/predator ratios as well as the pest and predator densities assayed are summarized in Table 1. Pest and predator densities used in the experiments ranged between values frequently found in undamaged field crops. Initial density of 20 T. urticae per leaflet was assayed only at 5/1. After 7 days, the active number of pest and predator on each leaflet was recorded and named final density. This period was set because it represented the time customarily interval used by growers for monitoring this pest. The effects of T. urticae/N. californicus ratios on pest and predator abundances after 7 days were tested using MANOVA. A significant MANOVA was followed by a univariate analysis (ANOVA). Data were transformed by square root when appropriate. Differences between active T. urticae densities after 7 days and the economic threshold level (ETL) were analyzed by the two-tailed t-test for differences between a population mean and a hypothesized population mean (Zar 1996). The ETL of this pest has not been determined in Argentina. However, observations made in La Plata area indicated that the yield was not reduced at infestation levels ranging from 50 to 100 active mites per leaflet, during the spring (Tito unpublished data). Taking into account this information and considering that the most common cultivars grown in La Plata are short-day (Rosa Linda, Milsei Tudla, Camarosa, Sweet Charly), we used the ETL of 50 active mites per leaflet reported by Wyman et al. (1979) for another short-day cultivar. A graphical model was constructed with experimental data. Initial active T. urticae and N. californicus densities were represented on the y-axis and x- axis, respectively. Then, a line was drawn joining the greatest number of pests at a given predator density such that pest final pest density would not

Table 1. Initial pest and predator densities and the pest/predator ratios used in the experiments.

Treatments: Initial pest/predator Pest–predator densities per leaﬂet (n)

Experiment 1 Experiment 2 Experiment 3 Experiment 4

5:1 5–1 (10) 10–2 (10) 15–3 (10) 20–4 (10) 7.5:1 5–0.67 (15) 10–1.33 (9) 15–2 (10) 10:1 5–0.50 (8) 10–1 (9) 15–1.5 (8) 15:1 5–0.33 (10) 10–0.67 (8) 15–1 (12) Control 5–0 (24) 10–0 (10) 15–0 (14) 20–0 (9) 60 exceed the ETL. This line separates two regions: the lower one represents the densities of both species at which the pest will never exceed the ETL after a week, while the upper one represents those densities that will exceed the ETL.

Field sampling

During 1999 and 2000 growing seasons, one greenhouse and one open-field commercial strawberry crop (of 600 and 900 m2, respectively) located in La Plata were sampled. Plants of ‘Camarosa’ and ‘Rosa linda’ cultivars were planted in both, greenhouse and field, in autumn. Crops with low or no miticide applications were chosen. During spring season, from September to December, densities (i.e. the number of active mites per leaflet) of both pop- ulations were determined once a week. A leaflet was selected at intervals of 5– 10 m in each row (the number of leaflets in a sample varied from 70 to 160). The number of eggs, immatures and adults of T. urticae and the number of active individuals, immatures and adult stages of N. californicus were recorded by examining each leaflet with a pocket lens of 10 · magnification. The mean density of active mites for both species was determined. To examine the relationship between field samples and experimental results, mean weekly densities were superimposed on the graphical model and were color coded to indicate whether or not pest density the following week was greater or less than the ETL.

Results

At the end of the week interval, MANOVA analysis indicated that prey and predator densities were significantly affected by T. urticae/N. californicus ratios (treatments) (Experiment 1: Wilks’ k = 0.09; F8,102 = 29.67; p < 0.001. Experiment 2 : Wilks’ k = 0.11; F8,80 = 20.55; p < 0.001. Experiment 3: Wilks’ k = 0.08; F8,96 = 30.09; p < 0.001. Experiment 4: Wilks’ k = 0.05; F2,16 = 159.62; p < 0.001). In the absence of the predator, active stages of T. urticae increased 16–24-fold, during a 7-day period. Univariate ANOVA indicated that in all experiments pest/predator ratio had a significant effect in reducing pest density (p < 0.001) (Table 2). Only in experiment 1, at initial predator densities of 0.33 and 0.50, did the final density of T. urticae not differ from the control and the percentages of T. urticae reduction were 9.87 and 18.27. The reduction in pest density in the remaining initial pest/predator ratios in experiment 1 and in the other experiments, ranged between 57 and 95%. Predator densities were positively affected by decreasing pest/predator ratios (p < 0.001) in experiment 3, and a similar pattern was observed in experiment 2(p < 0.005). Table 2. Numbers of active T. urticae/leaflet and N. californicus/leaflet, and pest/predator ratios after 7 days, at different initial conditions.

Initial pest Initial predator Initial pest/predator Pest density at Predator density at Pest/predator ratio n density density ratio 7 days (Mean ± SE) 7 days (Mean ± SE) at 7 days (Mean ± SE)

5 0 0 79.60 ± 8.40 a (9.24 ± 0.46) 0 0 24 5 0.33 15 71.74 ± 8.37 a (8.36 ± 0.49) 4.83 ± 0.36 a (2.30 ± 0.07) 16.01 ± 2.37 10 5 0.50 10 65.06 ± 11.93 a (7.74 ± 0.84) 6.81 ± 0.94 a (2.67 ± 0.16) 10.95 ± 2.82 8 5 0.67 7.5 21.89 ± 4.53 b (4.31 ± 0.50) 4.31 ± 0.61 a (2.12 ± 0.14) 7.79 ± 2.99 15 5 1 5 10.76 ± 2.14 b (3.21 ± 0.31) 4.34 ± 0.65 a (2.16 ± 0.12) 3.03 ± 0.79 10 10 0 0 173.46 ± 17.02 a (13.05 ± 0.61) 0 0 10 10 0.67 15 74.79 ± 8.71 b (8.58 ± 0.47) 6.59 ± 0.87 a (2.76 ± 0.14) 11.57 ± 2.09 8 10 1 10 37.29 ± 11.39 bc (5.44 ± 0.88) 5.30 ± 0.70 b (2.39 ± 0.13) 6.94 ± 1.95 9 10 1.33 7.5 21.74 ± 10.67 c (3.80 ± 0.93) 8.73 ± 1.33 a (3.04 ± 0.19) 3.19 ± 1.68 9 10 2 5 23.30 ± 3.55 c (4.75 ± 0.35) 11.53 ± 1.40 a (3.40 ± 0.22) 3.47 ± 1.63 10 15 0 0 251.52 ± 63.66 a (15.71 ± 0.59) 0 0 14 15 1 15 86.95 ± 36.81 b (9.09 ± 0.60) 9.11 ± 1.02 b (3.04 ± 0.17) 9.11± 1.56 12 15 1.5 10 112.75 ± 61.75 b (10.13 ± 1.13) 10.06 ± 1.36 b (3.20 ± 0.19) 15.50 ± 3.23 8 15 2 7.5 58.90 ± 38.25 b (7.18 ± 0.85) 12.50 ± 1.24 b (3.57 ± 0.16) 5.28 ± 1.17 10 15 3 5 11.90 ± 12.79 c (3.00 ± 0.54) 18.43 ± 1.57 a (4.32 ± 0.17) 0.66 ± 0.22 10 20 0 0 479.41± 76.47 a 0 0 9 20 4 5 49.09 ± 56.94 b 26.40 ± 2.46 2.50 ± 1.33 10

Means of T. urticae/leaflet within each experiment followed by the same letter are not significantly different (p > 0.05, Tukey test). In parentheses are mean and SE of square root transformed data. 61 62

Figure 1. Results of the comparison between active T. urticae densities after 7 days and the ETL (t-test, p < 0.05, at experimental initial pest densities from 5 to 20 females/leaﬂet and for diﬀerent T. urticae/N. californicus ratios. Black, empty and striped bars represent pest densities lower, equal and higher than the ETL, respectively.

When pest/predator ratio was 5/1 (at initial pest densities from 5 to 15 females/leaflet) the final number of active T. urticae/leaflet was significantly lower than the ETL (Figure 1), while at 20 females/leaflet this number did not differ from the ELT. At 7.5/1 ratio, the final number of active T. urticae/leaflet, at initial pest densities from 5 to 15 females/leaflet, reached the ETL without surpassing it. At 10/1 and 15/1 ratios, pest densities exceeded the ETL only at 15 initial T. urticae/leaflet. The graphical model was constructed using initial pest–predator density coordinates 5–0, 10–0.67, 15–2 and 20–4. Most greenhouse and field observations were located under the model line (56.25 and 66.67% in greenhouse in 1999 and 2000, respectively; and 85.71 and 66.67% in open field in 1999 and 2000, respectively) (Figure 2), and did not reach the ETL, so they are consistent with model predictions. From the points laying above the model line, only one having a high initial pest density and a very low initial predator density, exceeded the ETL. At the end of the experiments pest/predator ratios were similar or lower than at the beginning of the experiment. Only in one case the final ratio was higher than the initial (Table 2).

Discussion

Neoseiulus californicus was very eﬀective in limiting pest densities during a 7- day period for most pest/predator ratios and absolute densities used in our 63

Figure 2. Relation between graphical model and observed data (greenhouse and field). The line joins the greatest number of pests, at a given predator density, such that final pest density would not exceed the ETL. Empty dots represent initial field pest–predator combinations in which final pest density did not exceed the ETL. Black dots represent initial field pest–predator combinations in which final pest density surpassed the ETL. experiments. Different T. urticae/N. californicus ratios had a significant effect on the population abundance of the pest. Spider mite-Phytoseiid ratios <10/1 are considered favorable in many cropping systems (Croft and Hoyt 1983; Wilson et al. 1984; Gonza´lez Zamora et al. 1991; Strong and Croft 1995). In the present study, a T. urticae/N. californicus ratio of 5/1 reduced the pest below the ETL at pest initial density from 5 to 15. When ratios ranged from 7.5/1 to 15/1 T. urticae reached the ETL but never surpassed it at a 7-day interval, at initial densities from 5 to 10 active mites per leaflet. At densities from 15 and 20 active mites per leaflet, ratios of 7.5/1 and 5/1, respectively, avoided the pest surpassing the ETL. In these cases, we could expect that at the next week the pest would not reach very high densities, because the final prey/ predator ratios became, in general, more favorable to the predator. In spite of the simplicity of laboratory conditions, under which the graphical model was constructed, 77% of field data were similar to its predictions. This model shows the potential of this predator to be used in an IPM program. The field data below the model line completely fulfilled the predictions. The field data over the model line should have surpassed the ETL. Nevertheless, most of them did not do so. Probably other factors, besides the predator, limit pest growth in the field (i.e. adverse weather conditions, other mortality factors, lower food quality and cultural practices). Neoseiulus californicus, as well as other phytoseiids, has the capacity of ambulatory and aerial dispersal (Croft and Jung 2001). Dispersal behavior 64 would allow an early colonization of the strawberry crop and, once in it, it would allow the predator to disperse in search of the prey. This behavior results in the high temporal and spatial coincidence at leaflet level, as was detected in strawberry crops in the study area (Greco et al. 1999). Also, the presence of the predator in almost all samples and at very low pest densities, would suggest that N. californicus is able to persist within the agroecosystem even at low prey densities. Croft et al. (1997) and Pratt et al. (1999) have found that it may stay in a field with few prey and feed on other sources or wait for the prey to return while minimizing starvation. All these traits explain, at least in part, the capacity of this predator to prevent T. urticae surpassing the ETL in La Plata strawberry crops. Literature reports about ETL of this pest in strawberry are variable. Wyman et al. (1979) determined an ETL of 50 active mites per leaflet for short-day strawberry plants (Tufts) and Oatman et al. (1981, 1982) have shown that 90–100 mites/leaflet had no significant effect either on fruit yield or on size in strawberry. However, Walsh et al. (1998) indicated that detectable yield reductions on day-neutral cultivars (Selva) occurs at population densities higher than 1 T. urticae/leaflet. Zalom (2002) pointed out that the established ETL for the first 4 months following autumn transplant is five mites per mid-tier leaflet, while summer transplants have an ETL of 10 mites per mid-tier leaflet. Plants are less sensitive to mite feeding once harvest begins, and ETL increases to 15–20 mites per mid-tier leaflet. Our findings have relevance in developing an integrated program for this pest in strawberry. Conservation biological control involves the use of tac- tics that include the manipulation of the habitat of natural enemies in order to enhance their performance (Barbosa 1998). Raworth et al. (1994) have suggested that the efficacy of N. californicus could be enhanced in orchards in the south of France, by providing a specific pollen food source. In La Plata area, future studies should identify plant species harboring overwin- tering N. californicus and positively influencing its population increase. Furthermore, miticide applications should be minimized and restricted only to those that do not affect this predator. Conservation of N. californicus promoting favorable pest/predator ratios may result in early control of T. urticae.

Acknowledgements

We thank Fernanda Cingolani and Florencia Tejerina for technical assis- tance and Andrea Guillade for improving the English version of the man- uscript. We also thank two anonymous reviewers for their constructive comments. 65

References

Barbosa P. 1998. Conservation Biological Control. Academic Press, London pp.396. Coop L., Rosetta R. and Croft B. 1997. Release calculator and guidelines for using Neoseiulus fallacis to control two-spotted spider mites in strawberry. Version 1.1. Designed for use at the Web site: http://www.orst.edu/Dept/entomology/ipm/mcalc.html. Croft B.A. and Hoyt S.C. 1983. Integrate management of insect pests of pome and stone fruits. Wiley, New York. Croft B.A. and Jung C. 2001. Phytoseiid dispersal at plant to regional levels: a review with emphasis on management of Neoseiulus fallacis in diverse agroecosystems. Exp. Appl. Acarol. 25: 763–784. Croft B.A., Monetti L.N. and Pratt P.D. 1997. Comparative life histories and predation types: Are Neoseiulus californicus and Neoseiulus fallacis (Acari: Phytoseiidae) similar Type II selective predators of spider mites? Environ. Entomol. 27: 531–538. Garcıá Mari F., Masiello L., Saques J., Ribes A., Soro T. and Laborda R. 1991. Control biolo´gico de aran˜ a roja en cultivos valencianos de fresoń. Agrıćola Vergel. Marzo: 165–170. Gonza´lez Zamora J.E., Orenga S., Garcıá Mari F. and Laborda R. 1991. Liberaciońdeaćaros depredadores para el control de aran˜ a roja en fresoń. Phytoma 32: 20–27. Greco N., Sańchez N. and Liljesthro¨ m G. 1999. Spatial distribution and coincidence of Tetranychus urticae (Acari: Tetranychidae) and the predator Neoseiulus californicus (Acari: Phytoseiidae) on strawberry. Exp. Appl. Acarol. 23(7): 567–580. Helle W. and Sabelis M.W. (eds) 1985. Spider Mites Their Biology, Natural Enemies and Control. Vol. 1B, Elsevier, Amsterdam, pp. 285–297. Hussey N.W. and Scopes N. 1985. Greenhouse Vegetables (Britain). In: Helle W. and Sabelis M.W. (eds), Spider Mites Their Biology, Natural Enemies and Control. Elsevier, Amsterdam, pp. 285– 297. Nyrop J., English-Loeb G. and Roda A. 1998. Conservation biological control of spider mites in perennial cropping systems. In: Barbosa P. (ed.), Conservation Biological Control. Academic Press, pp396 . Oatman E.R., Sances F.V., LaPre´L.F., Toscano N.C. and Voth V. 1982. Effect of different infestation levels of the twospotted spider mite on strawberry yield in winter plantings in southern California. J. Econ. Entomol. 75: 94–96. Oatman E.R., Wyman J.A., Browning H.W. and Voth V. 1981. Effects of releases and varying infestation levels of the twospotted spider mite on strawberry yield in southern California. J. Econ. Entomol. 74: 112–115. Pratt P.D., Schausberger P. and Croft B.A. 1999. Prey-food types of Neoseiulus fallacis (Acari: Phytoseiidae) and literature versus experimentally derived prey-food estimates for five phytoseiid species. Exp. Appl. Acarol. 23: 551–565. Raworth D.A. 1990. Predators associated with the twospotted spider mite, Tetranychus urticae,on strawberry at Abbottsford, BC, and development of non-chemical mite control. J. Ent. Soc. Brit. Columbia 87: 59–67. Raworth D.A., Fauvel G. and Auger P. 1994. Location, reproduction and movement of Neoseiulus californicus (Acari: Phytoseiidae) during the autumn, winter and spring in orchards in the south of France. Exp. Appl. Acarol. 18: 593–602. Ribes A. 1990. Control biolo´gico de la aran˜ a roja en fresoń. Agricultura y cooperacioń 76: 24–25. Shaw P.B. 1985. Simulation model of a predator–prey systems comprised of Phytoseiulus persimilis and Tetranychus urticae. II Model sensitivity to variations in the life history parameters of both species and to variations in the functional response and components of the numerical response. Res. Popul. Ecol. 27: 1–23. Steinberg S., Hadassa C. and Kaminski J. 1999. Product control of Phytoseiulus persimilis (Athias- Henriot): current practice of a commercial producer. In: van Lenteren J.C. (ed.), Integrated Control in Glasshouses, IOBC/WPRS Bulletin: 22(1), 237–240. 66

Strong W.B. and Croft B.A. 1995. Inoculative release of Phytoseiid mites (Acarina: Phytoseiidae) into rapidly expanding canopy of hops for control of Tetranychus urticae (Acarina: Tetrany- chidae). Environ. Entomol. 24(2): 446–453. Walsh D.B., Zalom F.G. and Shaw D.V. 1998. Interaction of the twospotted spider mite (Acari: Tetranychidae) with yield of day-neutral strawberries in California. J. Econ. Entomol. 91: 678– 685. Waite G.K. 1988. Integrated control of Tetranychus urticae in strawberries in southeast Queens- land. Exp. Appl. Acarol. 5: 23–32. Wei-Lan Ma and Laing J.E. 1973. Biology, potential for increase and prey consumption of Amblyseius chilenensis (Dosse) (Acarina: Phytoseiidae). Entomophaga 18: 47–60. Wilson L.T., Hoy M.A., Zalom F.G. and Smilanik J.M. 1984. Sampling mites in almonds: 1. Within-tree distribution and clumping pattern of mites with comments on predator–prey interactions. Hilgardia 52: 1–13. Wyman J.A., Oatman E.R. and Voth V. 1979. Eﬀects of varying twospotted spider mite infestation levels on strawberry yield. J. Econ. Entomol. 72: 747–753. Zalom F.G., Pickel C. and Welch N.C. 1990. Recent trends in strawberry arthropod management for coastal areas of the western United States. In: Bostanian N.J., Wilson L.T. and Dennehy T.J. (eds), Monitoring and Integrated Management of Arthropod Pests of Small Fruit Crops. Intercept, Hampshire, England, pp. 239–258. Zalom F.G. 2002. UC IPM Pest Management Guidelines: Strawberry, Insects and Mites. In: Philips P.A., UC IPM Program, UC Cooperative Extension Ventura Co. N.C. Toscano, Entomology, UC Riverside UC ANR, Publication 3468. www.ipm.ucdavis.edu/PMG/select- newpest.strawberry.html. Zar J.H. 1996. Biostatistical Analysis. Prentice Hall, New Jersey pp. 662.