Soutenance mémoire
Utilisation de plantes-réservoirs en lutte biologique
Potential and cost of a pollen supplement, a banker plant and misting for enhancing biological control of Tetranychus urticae with Neoseiulus fallacis in high tunnel raspberry production in Quebec, Canada
Résumé
La production sous grands tunnels améliore le rendement et la qualité des fruits, mais fournit également un environnement idéal pour la prolifération de Tetranychus urticae Koch (Acari: Tetranychidae). Bien que des acaricides soient disponibles, les délais avant-récolte et le développement de résistance peuvent être problématiques. Le prédateur Neoseiulus fallacis Garman (Acari: Phytoseiidae), est bien connu pour son efficacité à contrôler les tétranyques. Cependant, son utilisation a tendance à être coûteuse, puisque plusieurs introductions sont souvent nécessaires au cours d’une même saison de production. L’objectif principal de cette étude était d’améliorer l’efficacité de la lutte intégrée des ravageurs T. urticae tout en diminuant le nombre d’introductions de phytoséiides dans la culture de framboises sous grands tunnels. Nous avons testé l’efficacité du Nutrimite® (de Biobest), un supplément alimentaire pour acariens prédateurs, une plante-réservoir (Sorbaria sorbifolia (L.) A. Braun (Rosaceae)), et la brumisation afin de réduire le coût de la lutte biologique aux tétranyques en framboisière sous tunnels. Ce projet de recherche a été réalisé dans un contexte qui reflète le plus fidèlement possible la réalité de l’industrie horticole du Québec: les essais ont eu lieu chez deux producteurs de framboises en grands tunnels avec des populations de tétranyques naturelles. Les données recueillies par dépistage ont démontrées que le Nutrimite a eu un effet positif sur l’efficacité prédatrice de N. fallacis, que la brumisation réduit les populations de T. urticae et que S. sorbifolia n’a eu aucun effet comme plante-réservoir.
1 Abstract
2 High tunnels increase yield and improve fruit quality but also provide an ideal
3 environment for Tetranychus urticae Koch (Acari: Tetranychidae) outbreaks. Acaricide
4 application is an option, but pre-harvest interval and development of resistance among
5 spider mite can be problematic. Predatory mites such as Neoseiulus fallacis Garman
6 (Acari: Phytoseiidae) are well known for their efficacy against spider mites, but biological
7 control with phytoseiids can be costly, as several introductions are required throughout the
8 season. The main objective of this study was to test strategies for improving the efficacy of
9 integrated pest management of T. urticae while decreasing the number of phytoseiid mite
10 introductions needed in high tunnel raspberries, and reducing costs. We tested the efficacy
11 of a pollen supplement (Nutrimite® from Biobest) for phytoseiids, banker plants (Sorbaria
12 sorbifolia (L.) A. Braun (Rosaceae)),and misting. Both the pollen supplement and misting
13 were found to reduce T. urticae densities. Our findings may make biological control more
14 attractive to growers.
Introduction
Interest in high tunnel production has increased in recent years in North America. Many berry growers in the northern United States and Canada have established these structures over the last decade (Demchak and Hanson 2013; Xu et al. 2014), and high tunnel raspberry is the main protected crop grown in the province of Quebec nowadays (Villeneuve, 2011; Xu et al., 2014).
High tunnel raspberry cultivation offers several advantages: it protects the crop from rain and heavy winds, and extends the growing season (Wien 2009; Demchak and Hanson 2013; Xu et al. 2014; Castilho et al. 2015). Because the fruits are not exposed to rain, there is a significant reduction of diseases such as gray mold (Botrytis cinerea) (Heidenreich et al. 2012; Demchak and Hanson 2013; Funt et Hall 2013). These attributes allow growers to achieve higher crop yields, up to more than twice those obtained with field grown raspberries (Xu et al., 2014). Compared to field-grown raspberries, those grown under tunnel exhibit a higher level of cosmetic quality and uniformity, as well as better post-harvest conservation (Heidenreich et al. 2012; Demchak and Hanson 2013).
However, protected crops have been found to contain a greater number of detectable pesticides with a higher proportion of samples containing multiple pesticide residues when compared to analogous crops grown in an open field (Allen et al., 2015). High tunnels also provide ideal conditions for the proliferation of spider mites (Sonneveld et al. 1996; Heidenreich et al. 2012; Demchak and Hanson 2013; Castilho et al. 2015). The tunnel environment is hotter and dryer than field conditions (Lemaire, 2012; Wien, 2009; Xu et al., 2014), which favors two-spotted spider mite (Tetranychus urticae) proliferation. This mite pest grows faster and produces more eggs when relative humidity is low (25-30%) rather than high (85-90%) (van de Vrie et al. 1972; Helle and Sabelis 1985a; Wood 1992; Walzer et al. 2007). The survival of immature stages, adult life-span and percentage of egg hatch are also favored by conditions of low relative humidity (Holtzer et al. 1988; Wood, 1992), as is leaf damage (Helle and Sabelis 1985a).
The two-spotted spider mite (TSSM), Tetranychus urticae, is the most common arthropod pest, and is among the most damaging to raspberry grown under high tunnels (Sonneveld et al. 1996; Heidenreich et al. 2012; Demchak and Hanson 2013). While several miticides are registered against TSSM, pre-harvest application intervals and resistance development can be problematic (Van Leeuwen et al. 2010; Yorulmaz Salman and Sarıtaş 2014). Moreover, to date, T. urticae showed resistance to 94 active ingredients listed in 468 cases (Whalon et al., 2016). Among all arthropod pests, the TSSM has the highest prevalence of pesticide resistance (Ferreira et al., 2015; Van Leeuwen et al., 2010; Whalon et al., 2016).
Phytoseiid mites are important natural enemies of spider mites, and are widely used in biological control programs (Helle and Sabelis 1985b; McMurtry and Croft 1997; McMurtry et al. 2013). While the effectiveness of phytoseiid mites for spider mite control in high tunnel raspberries has been demonstrated in the past (e.g. Heidenreich et al. 2012; Lemaire 2012), several introductions can often be needed, making this method expensive (Pratt and Croft 2000; Frank 2010) and therefore less attractive to growers.
Neoseiulus fallacis is a commercially available phytoseiid native to Quebec and an effective natural enemy of TSSM (Bostanian et al., 2010; Roy et al., 2005). These attributes were important in selecting a predator species for introduction in a perennial system such as cultivated raspberry and made N. fallacis an ideal candidate for our tests. According to the classification proposed by McMurtry and Croft (1997), N. fallacis is a Type II predator, meaning that it is a selective predator of tetranychid mites and can also feed and complete their development on other arthropods and pollen (McMurtry and Croft 1997; Pratt et al. 1999; Lundgren 2009; McMurtry et al. 2013). A diet of pollen improves N. fallacis survival by more than 2.5 times compared to a starvation regimen (Pratt et al., 1999). Supplementing pollen to generalist predators is a practice that is becoming more common in biological control programs (e.g., Duarte et al., 2015; Janssen et Sabelis, 2015).
Banker plants are used in many crops to enhance establishment, development and dispersal of beneficial organisms employed in biological control and thereby contribute to long-term suppression of a pest (Pratt and Croft 2000; Frank 2010; Parolin et al. 2010; Huang et al. 2011; Messelink et al. 2014). A study by the Quebec Institute for Expertise in Ornamental Horticulture (IQDHO) suggested that Sorbaria sorbifolia (Rosaceae), a shrub commonly grown in ornamental nurseries, could has potential for use as a banker plant among nursery crops in Quebec (Authier et al., 2012). In Authier et al (2012), S. sorbifolia was found to be an attractive plant for phytoseiid mites, including N. fallacis (Authier et al. 2012; Huot 2012; Lemaire and Simard 2012).
While misting is widely used in greenhouses to reduce temperatures (Baeza et al., 2013; Both, 2008; Connellan, 2002; Suzuki et al., 2015), it is far less common in high tunnels even though a few studies have confirmed the benefits. For instance, a study conducted of tunnels in Switzerland showed better spider mite control with Phytoseiulus persimilis when crops were misted (Linder et al., 2003b), possibly due to the reduction in temperature and increase in relative humidity (Linder et al., 2003). A study conducted in greenhouse cucumber crops showed that fogging favored P. persimilis while simultaneously reducing T. urticae populations (Zhang, 2003). While lethal humidity (LH50) is about 70% at 20°C for N. fallacis eggs (Croft et al., 1993), with a relative humidity of 50%, larval mortality of the species reaches 91.9% for individuals who are fed and 98.1% for those under starvation (Croft et al., 1993). N. fallacis thereby requires cooler and more humid conditions than those created under high tunnels and misting might procure these conditions.
Our aim was to test the efficacy of 1) a pollen supplement, 2) a banker plant, 3) misting and 4) a combination of these strategies for enhancing the performance of predatory mites in high tunnel raspberry crops in Quebec, Canada. A secondary aim was to estimate the cost of each strategy and compare it to the cost of conventional management, with acaricide applications.
Materials and methods
Experiment setting
High tunnel trials took place in 2014 and 2015 at the farms of two important berry growers in the greater Quebec City region in the province of Quebec (Canada): Ferme Onésime Pouliot in Saint-Jean-de-l’Ile-d’Orléans (referred to hereafter as the ‘Pouliot farm’; 46°55’23.3”N 70°57’59.6”W) and Les Productions horticoles Demers Inc. in Lévis (hereafter the ‘Demers farm’; 46°42’16.2”N 71°19’58.3”W). One tunnel covered an area of 896 m2 (8m x 112m) at the Pouliot farm, and 592 m2 (8m x 74 m) at the Demers farm. One tunnel was used at the first farm and two tunnels were used for the tests at the second. Three rows of floricane fruiting red raspberry cv. ‘Tulameen’ were under cultivation in pots in each tunnel. Pouliot farm used 15L pots and Demers farm 10L pots. Sixteen plots of 100 fruiting canes were delimited on each farm (about 27 m2) at Pouliot farm and about 40 m2 at Demers farm). Miticides were applied as needed, because although we wanted to limit the number of chemical interventions as much as possible, these trials were conducted on plants belonging to commercial growers, for whom a high yield is essential.
Based on a previous study (Therriault et al., 2013), prior to trials, we set a threshold of three spider mite motiles/leaflet. For both years of this study, mite population densities were determined by in situ observations during spring and destructive sampling when raspberry plants were sufficiently developed to tolerate it (about June 5-10). From then on, 20 leaflets per plot were collected randomly twice (2014) or once (2015) a week, placed in re-sealable storage bag and stored in a 10°C cooler until counted in the laboratory. The number of TSSM and predatory mite motiles and eggs were then counted with a stereomicroscope. In 2015, a second monitoring, an in situ visual one, was done every week. Application cost of each treatment was also estimated.
The IPM strategies tested during this two-year study were a combination of N. fallacis introductions with one of the following techniques: the use of a cattail pollen based food supplement known as Nutrimite® (Biobest), S. sorbifolia as a banker plant, and misting of high tunnels. The combinations of strategies tested slightly differed depending on the year (2014 and 2015), as explained below.
2014 experiment
Mite populations were monitored from May 29 to September 8, 2014. Four treatments were tested, each replicated four times: #1: predators only (control) (P); #2: predators + supplement (PS); #3: predators + banker plants (PB) and #4: predators + banker plants + supplement (PSB). For treatments with banker plants, the density was two S. sorbifolia bushes per plot of 100 raspberry canes. Sorbaria plants were installed in plots on June 9 at both farms. A miticide, specifically the acaricide Acramite (active ingredient: bifenazate), was applied early in the season (June 11 – Pouliot, June 16 – Demers) for all treatments. Seven to 10 days afterwards, N. fallacis individuals were introduced (rate of 5 individuals /m2). At the same time the pollen supplement Nutrimite (BioBest) was applied for the first time (June 18 – Pouliot, June 26 – Demers), using the hand sprayer « Nutrigun » at a rate of 500g / ha. As the food supplement retains its nutritional quality for two weeks following its application (Biobest Ltd, 2013), second and third applications were made two and four weeks after the phytoseiid release. The spider mite specialist Phytoseiulus persimilis Athias-Henriot (Acari: Phytoseiidae) was released as needed for high densities (hot spots) of T. urticae (Hoy, 2011c; Kazak et al., 2015) at an introduction rate of 25 individuals/ m2.
Statistical analysis (2014)
In 2014, separate analyses were performed for each site because crop management differed between farms: miticide was applied once at the Pouliot farm, three times at the Demers’.
The experimental design was completely randomized with repeated measures using date as the repeated factor. The best covariance structure among observations taken across time on the same experimental unit was chosen based on the AIC criterion. The dependent variables (i.e. mean density of T. urticae and phytoseiids per leaflet) were log-transformed in order to approach the assumptions of the model. When appropriate, the baseline measurement was used as a covariate in the model. The Shapiro-Wilk’s statistic was used to evaluate the normality assumption, while residuals’ plots were used to investigate the homogeneity of variance. Slight departure of the normality assumption was sometimes observed, but since results were identical to those obtained by the Brunner et al. (2002) nonparametric approach for longitudinal data, the usual approach was retained. Analyses were performed using the Mixed procedure of SAS software (release 9.4, SAS Inc., NC) with a significance level of α=0.05.
2015 experiment
In 2015, the experiment was monitored from May 26 to August 31, 2015. Four treatments were compared: #1: predators only (control) (P); #2: predators + supplement (PS); #3: predators + misting (PM) and #4: predators + supplement + misting (PSM). There were two introductions of N. fallacis at a lower rate than in 2014: 1 /m2 on May 27 and June 10. The supplement Nutrimite (500g/ha) was applied twice, at the same time N. fallacis was released. Misting was operational from June 13 and was set to run from 10 a.m. to 4 p.m. on warm and sunny days. The misting system functioned for 25 seconds at intervals of 1 minute as soon as the temperature inside the tunnel rose above 27°C. A six-station irrigation controller (DIG 5006-1 model) with a 24V electrical valve was used to control misting, programmed to function for 25 seconds every minute at the Pouliot farm. As for the Demers farm, it was programmed to function for 3 seconds every minute during harvest time, to avoid heavy calcareous deposits on foliage and fruits. Two misting lines were installed in the tunnels, between and parallel to raspberry rows, at a height of about 3 m. The misting system was set with with 2-way Tee Dan foggers type (Jain irrigation inc.) with 1.8GPH nozzles spaced 3 m apart. Droplet size was 90 microns at 58 PSI (Jain Irrigation, 2010). The foggers were oriented in the same direction as the rows, in order to avoid watering the raspberry plants directly. At both farms, water used for misting came from an irrigation lake. The miticide Acramite 50WS (a.i.: bifenazate) was applied on June 27 on the Pouliot farm (with miticide Apollo SC, a.i.: clofentezine) and June 30 at the Demers farm. Temperature and relative humidity (RH) data were recorded every 15 minutes by three HOBO data loggers in each tunnel and one outside the tunnels, for a total of 7 data loggers per site.
Because the incidence of gray mold caused by Botrytis cinerea Pers. was a concern in misted tunnels, four random samples of 100 fruits per tunnel (n = 800) were taken at the Demers farm between September 10 and September 21, 2015. The fruits collected were examined visually for the presence or absence of B. cinerea (observation of conidiophore sporulation (Agrios, 2004) and/or other signs of the fungus itself).
Statistical analysis (2015)
Data from both growers were combined for the 2015 population analyses. For that experiment, the experimental design was a split-plot design with repeated measures, with misting in the main plots, supplement in the subplots and date as the repeated factor. The best covariance structure among observations taken across time on the same experimental unit was chosen based on the AIC criterion. All analyses were performed with a significance level of α=5%.
For the two-spotted spider mite population analysis, a generalized linear mixed model was fitted to the data using a negative binomial distribution with the default log link function. This was done because there were many zeros in the data (no TSSM observed). This analysis was performed using the Glimmix procedure of SAS software (release 9.4, SAS Inc., NC).
For the phytoseiid population analysis, a normal linear mixed model was fitted to the data. The dependent variables (i.e. mean density of phytoseiids per leaflet) were log-transformed in order to meet the assumptions of the model. The Shapiro-Wilk’s statistic was used to evaluate the normality assumption, while residuals’ plots were used to investigate the homogeneity of variance. Analyses were performed using the Mixed procedure of SAS software (release 9.4, SAS Inc., NC).
Results
2014 experiment
In 2014, the TSSM data showed a significant difference between treatments (F = 3.71; df = 3, 12; p = 0.0425) at the Pouliot farm. The P and PB treatments were equivalent, as were the PS and PSB treatments (Error! Reference source not found.). For the P and PB treatments, the mean density of spider mite motiles was 1.07 (± 1.42) per leaflet; for the PS and PSB treatments, the mean density was 0.61 (± 0.85). The highest number of TSSM was reached in the P treatment (control), with an average of 6.71 (± 11.34) motiles per leaflet on August 5 (Error! Reference source not found.). On that date, two plots under the P treatment had very high mean numbers of T. urticae: 11.90 (± 17.01) and 26.35 (± 18.28) (data not shown). For the other treatments, only PSB exceeded the threshold of 3 motiles/leaflet, but for a very short time, reaching 3.46 (±3.49) motiles/leaflet on August 1, before the mean spider mite population dropped to 3.00 (±5.14) and then to 0.88 (±1.20) motiles/leaflet within the next week (Error! Reference source not found.).
As for phytoseiid densities, there was also a significant difference between treatments (F = 8.97; df = 3, 12; p = 0.0022) with the same pattern as for spider mites; the P and PB treatments were equivalent, as were the PS and PSB treatments (Error! Reference source not found.). For the P and PB treatments, the mean density of phytoseiid motiles was 0.30 (± 0.37) per leaflet; for the PS and PSB treatments, the mean was 0.25 (± 0.27). The highest average of phytoseiid motiles/leaflet was 1.28 (± 2.35), obtained on July 25 in the control treatment. At the Pouliot farm, a total of three P. persimilis introductions were made during the season: July 25, August 1 and August 8. According to TSSM densities, because plots didn’t have the same pest pressure, no P. persimilis was introduced in the PS treatment plots, one introduction was made in PSB treatment plots (July 25) and two introductions in P and PB treatments (August 1 and August 8). The P. persimilis introduction rate was a curative one: 25/m2 (Biobest Canada Ltd, 2015a, 2015b). The average size of mite populations (T. urticae, N. fallacis and P. persimilis) at the Pouliot farm is shown in Error! Reference source not found., which shows that very few P. persimilis individuals were found, with a mean density of 0.01 (± 0.02) motiles per leaflet.
Table 2 shows the estimated cost for each treatment. Plots with the PS treatment harbored the lowest densities of spider mites (t = -7.66; df = 12; p <.0001), compared to treatments P and PB, and also had the lowest estimated cost ($2709 / ha), 24 to 36% lower than that of the other treatments tested ($ 3551, $ 4233 and $4262 / ha) (Table 2).
On the Demers farm, we found no treatment effect on either density of predatory (F = 2.56; df = 3, 12; p = 0.1082) or spider mites (F = 0.48; df = 3, 12; p = 0.7009). For all treatments, the mean density of N. fallacis motiles per leaflet varied from 0.00 to 2.05 (± 0.61). The highest average size of the phytoseiid motile populations was 4.14 (± 3.29), reached on July 22 in the control treatment. As for TSSM, the mean density of motiles per leaflet varied from 0.03 to 9.94 (± 3.41). The maximum average number of TSSM individuals was also reached on July 22, but in the PSB treatment, with 14.77 (± 10.03) motiles per leaflet (data not shown). At the Demers farm, there were also three P. persimilis introductions made on July 10, July 16 and July 29 (data not shown). The P. persimilis population pattern was the same as that found on the Pouliot farm, very few individuals were scouted: the mean density of motiles varied between 0.00 and 0.29 (± 0.10).
|
Table des matières
Introduction générale
Chapitre I : État des connaissances
1.1 Les Tetranychidae
1.1.2 Les tétranyques dans les framboisières
1.1.3 Les moyens de lutte contre les tétranyques
1.2 Les Phytoseiidae
1.2.1 Biologie des phytoséiides
1.2.3 Régime alimentaire (type de prédation)
1.2.4 Utilisation en lutte biologique
1.3 Suppléments nutritifs
1.3.1 Rôle du pollen dans la lutte biologique avec phytoséiides
1.4 Plantes-réservoirs
1.4.1 Utilisation de plantes-réservoirs en lutte biologique
1.4.2 Sorbaria sorbifolia
1.5 Brumisation
1.5.1 Utilisation en serres
1.5.2 Utilisation en lutte biologique
1.5.3 Performance des framboisiers
1.6 Culture du framboisier
1.6.1 Framboisiers non-remontants
1.6.2 Framboisiers remontants
1.6.3 Principaux ravageurs et maladies des framboisières
1.6.4 Culture de framboisiers sous grands tunnels
1.7 Problématique
1.8 Objectifs et hypothèses de recherche
1.8.1 Objectif général
1.8.2 Objectif spécifique 1
1.8.3 Objectif spécifique 2
1.9 Approche méthodologique
Chapitre II: Potential and cost of a pollen supplement, a banker plant and misting for enhancing biological control of Tetranychus urticae with Neoseiulus fallacis in high tunnel raspberry production in Quebec, Canada
Résumé
Abstract
Introduction
Materials and methods
Experiment setting
2014 experiment
2015 experiment
Results
2014 experiment
2015 experiment
Discussion
Phytoseiid mite introductions
Food supplement
Banker plant
Misting
Acknowledgements
References
Appendices
Chapitre III : Conclusion générale
Bibliographie
Télécharger le rapport complet
