Resistance of Wild Solanum Accessions to Aphids and Other Potato Pests in Quebec Field Conditions

Two experiments were done to determine the susceptibility of six wild potato accessions to the aphids Macrosiphum euphorbiae (Thomas) (Hemiptera: Aphididae) and Myzus persicae (Sulzer). Densities of aphid colonies were compared between caged Solanum pinnatisectum Dunal (Solanales: Solanaceae), S. polyadenium Greenmam, S. tarijense Hawkes, S. infundibuliforme Philippi, S. oplocense Hawkes, and S. stoloniferum Schlechted and Bouché, and the commercially cultivated potato plant S. tuberosum L. cv. Désirée. Moreover the susceptibility of S. polyadenium and S. tarijense to the Colorado potato beetle Leptinotarsa decemlineata (Say) (Coleoptera: Chrlysomelidae), the potato flea beetle Epitrix cucumeris (Harris), and the potato leafhopper Empoasca fabae (Harris) (Hemiptera: Cicadellidae) was compared to that of S. tuberosum cv. Désirée in the field. Results indicated that S. polyadenium and S. tarijense were more resistant to M. persicae than S. pinnatisectum and the commercially cultivated S. tuberosum cv. Désirée. Solanum polyadenium was more resistant to aphids than S. tarijense in 2004, but not in 2005. Moreover, S. polyadenium and S. tarijense were more resistant than S. tuberosum cv. Désirée to L. decemlineata, E. cucumeris and E. fabae.

Development of insecticide resistance (Boiteau et al. 1987;van Toor et al. 2008) and public awareness of possible health problems associated with pesticides (Rimal et al. 2001) increased the interest for the development of alternative and sustainable control strategies for potato pests. One strategy that has received much attention is the development of insectresistant potato cultivars (Flanders et al. 1992;Pelletier and Michaud 1995). In addition to the development of genetically modified plants, researchers have focused mainly on hybridization programs between commercially cultivated potato cultivars and related wild Solanum species resistant to several potato pests (Flanders et al. 1992) and diseases (Chen et al. 2003).
Laboratory and field experiments have identified many Solanum accessions as resistant to either one or a few potato pests (Flanders et al. 1992;LeRoux et al. 2007). Resistance mechanisms have been classified as either antibiosis (changes in insect biology and demographic parameters) or antixenosis (changes in insect behavior leading to low or non acceptance of the host plant) (LeRoux et al. 2008). In Solanum spp., antibiosis relies mainly on glycoalkaloids present in leaves (Stürckow and Löw 1961;Gregory et al. 1981;Güntner et al. 1997Güntner et al. , 2000Lachman et al. 2001;Lorenzen et al. 2001). For example, some species such as Solanum pinnatisectum Dunal (Solanales: Solanaceae) and S. polyadenium Greenmam have a high level of -tomatine (Gregory et al. 1981;Sinden et al. 1991;Deahlt et al. 1993), which is known to hinder L. decemlineata growth (Kowalski et al. 2000) and lower M. euphorbiae reproductive rate (Güntner et al. 1997). High levels of -chaconine and -solanine also have negative impacts on M. persicae adults, lowering feeding and fecundity, and increasing mortality (Fragoyiannis et al. 1998). Antixenosis-based resistance may be conferred by glandular trichomes (Gibson 1971(Gibson , 1976aTingey and Gibson 1978;Tingey and Laubengayer 1981;Neal et al. 1990;Pelletier and Smilowitz 1990;Yencho and Tingey 1994;Alvarez et al. 2006;Pelletier and Dutheil 2006). Glandular trichomes are known to alter the ability of many herbivores to colonize, forage, and survive on the plant. For example, glandular trichomes reduce the proportion of L. decemlineata larvae feeding on S. polyadenium and S. berthaultii leaves and increase larval mortality (Gibson 1976b;Neal et al. 1989).
On the other hand, certain plant characters conferring pest resistance are thought to have positive impacts on some natural enemies (Obrycki 1986). For example, the ladybird beetle Coleomegilla maculata (Griffin and Yeargan 2002a, b) lays eggs preferentially on plant species bearing glandular trichomes, and oviposition of the aphidophagous midge Aphidoletes aphidimyza is positively correlated with S. tuberosum trichome density (Lucas and Brodeur 1999). Obrycki and Tauber (1985) also observed positive relationship between ladybird eggs (unidentified species) and trichome abundance. For both ladybirds and A. aphidimyza, the preference for oviposition on trichome bearing plants may be associated with a lower predation risk of the most susceptible life stages on those plants (Lucas and Brodeur 1999;Griffin and Yeargan 2002a, b).
Studying the impact of Solanum spp. candidates on both pests and natural enemies is of importance for breeding programs, as natural enemies are known to contribute to aphid (Obrycki et al. 1983;Karley et al. 2003;Koss and Snyder 2005) and L. decemlineata (Hilbeck et al. 1997;Chang and Snyder 2004;Koss and Snyder 2005) biocontrol. The exclusion of natural enemies from resistant plants could provide an enemy-free space to herbivores adapted to resistant plants (Gassmann and Hare 2005).
Fields experiments are required in many different geographical areas since the expression of resistance characters may be lower in the field than in the laboratory (Tingey and Gibson 1978;Obrycki and Tauber 1984) and vary with environmental conditions (Gianfagna et al. 1992;Nihoul 1993). As such, resistant plants may have a different impact in the field than in controlled conditions, both on pests and on their natural enemies (Obrycki and Tauber 1984). Only a few field studies have tested the impact of resistant Solanum accessions on natural enemies.
To address that question, two experiments were done in Southern Quebec field conditions. The first compared the capacity of the aphids M. persicae and M. euphorbiae to thrive on six caged Solanum accessions and on caged potato plants. The second sampled and compared potato pest and natural enemy occurrence on two Solanum accessions and on potato plants.

Materials and Methods
The experiments were performed from 2004 to 2007 on a commercial farm located at Saint-Paul d'Abbotsford (45.4127°N, 72.8598°W), Quebec, Canada. No insecticides, fungicides, or herbicides were applied to the experimental field.
Plants were previously grown in a greenhouse located at the Horticulture Research and Development Center, Agriculture and Agri-Food Canada, Saint-Jean-sur-Richelieu, Qc, Canada. Wild Solanum species seeds were obtained from the USDA Potato Introduction Project (Sturgeon Bay, Wisconsin, U.S.A.). The seeds were sown in early April in 25 X 50 cm plastic containers. Three weeks later, 30 seedlings of each variety were transplanted into pots (15 cm diameter). Potato plants S. tuberosum cv. Désirée were grown from tubers.
Aphids (M. persicae and M. euphorbiae) were reared on S. tuberosum cv. Désirée. One rearing unit was located at the Horticulture Research and Development Center, and the other was located at the University of Quebec in Montreal. Plantlets were transferred from the greenhouse to the field in early June both years, and were first placed in a shaded area to allow adaptation to field conditions. Three days later, nine plants of each accession and nine potato plants were transplanted in the field. Each plant species was transplanted in the field in three groups (a group consisted of 3 plantlets transplanted on a row). Both row spacing and planting distance on a row were set to 0.90 m. A muslin cage (Height: 1 m, Diameter: 0.60 m) was placed above each plant. The muslin at the base of each cage was buried in the soil at a depth of 20-30 cm and a lateral entry closed by a metal clip allowed access to the plant. Sampling started one week following aphid infestation. In 2004, all Solanum species were sampled weekly for four weeks. Solanum pinnatisectum, S. polyadenium, S. tarijense, and S. tuberosum cv. Désirée were sampled further for two weeks for a total of six weeks. In 2005, all plants were sampled weekly for six consecutive weeks. At each sampling date, the numbers of apterous and alate aphids were counted separately. All other insects found on the plants were removed.

Experiment 1 -Impact of six wild Solanum accessions on Myzus persicae and
Experiment 2 -Impact of two wild Solanum accessions on aphids, Leptinotarsa decemlineata, Epitrix cucumeris, Empoasca fabae, and natural enemies Experiment 2 was performed in 2007. It aimed at evaluating the natural occurrence of potato pests and natural enemies on S. tarijense PI 414150, S. polyadenium PI 230463, and S. tuberosum cv. Désirée in Southern Quebec field conditions. The accessions used were the same as in experiment 1.
Plantlets were transferred from the greenhouse to the field on 18 June 2007 and planted on 20 June 2007. The three accessions were planted in nine monospecific patches (three patches by plant species) laid out in a latin square. Each patch consisted of 7 x 7 plants. The distance between each patch was 1.80 m. The distance between plants corresponded to a typical plantation layout in southern Quebec, i.e. 0.30 m between plants on a row and 0.90 m between rows. Weeds were manually removed once or twice a week.
Leptinotarsa decemlineata survey. On 29 June, 3 July, 6 July, 10July, and 11 July 2007 all plants were inspected and all observed L. decemlineata eggs, larvae, and adults were removed. Eggs and adults were counted.
Destructive sampling. On 27 July 2007, 15 plants per patch were cut, individually enclosed in plastic bags, brought to the laboratory, and put in a freezer until inspection. Plants were carefully inspected and every insect collected was put in 70% alcohol for future identification. Specimens were identified in the laboratory using a dissecting microscope and a field guide.

Statistical analysis
Data was analyzed using the statistical software JMP (SAS Institute 2001). Experiment 1: Data were rank-transformed and a MANOVA for repeated measures was applied followed by contrast analysis (between each pairs). In 2004 the analysis was done for four weeks in order to allow comparisons between the seven accessions tested, and then for six weeks for comparisons between S. tuberosum cv. Désirée, S. pinnatisectum, S. polyadenium, and S. tarijense. In 2005, the analysis was done for six weeks. Plants with missing data were excluded from the analysis (therefore, 1-4 plants/accession were compared for each analysis).
Experiment 2: For the L. decemlineata survey, the numbers of egg clutches and adults per row were compared (rows were used as experimental units). For the destructive sampling the numbers of individuals per plant were compared (individual plants were used as experimental units). This statistical methodology was used since the number of patches by accession was low (3 by accessions). Data were compared between accessions using Kruskal-Wallis tests and Tukey-type posthoc tests for non parametric data (Zar 1999).  Experiment 2 -Impact of two wild Solanum accessions on aphids, Leptinotarsa decemlineata, Epitrix cucumeris, Empoasca fabae, and natural enemies Leptinotarsa decemlineata survey. On most sampling dates, more egg clutches and adults were collected on S. tuberosum cv. Désirée than on S. poyadenium and S. tarijense (Table  1). There were no significant differences aaaaaaa between S. polyadenium and S. tarijense regarding egg clutch and adult densities.

Experiment 1 -Impact of wild
Destructive sampling. Results of the destructive sampling are presented in Table 2 (Kruskal-Wallis: L1 = 27.21, d.f. = 2, P < 0.0001; L2 = 27.03, d.f. = 2, P < 0.0001). The higher proportion of third instar larvae on S. tuberosum cv. Désirée was not as pronounced than for the first and second instar larvae (28 individuals, 89.3% on S. tuberosum and 10.7% on S. polyadenium), and even though there was a global significant difference (Kruskal-Wallis: = 13.79, d.f. = 2, P = 0.0010), posthoc tests found no significant difference between Solanum species. The fourth instar larvae distribution did not differ between the three Solanum species: out of the 11 individuals collected 72.7% were from S. tuberosum cv. Désirée and 27.3% from S. polyadenium (Kruskal-Wallis: = 4.14, d.f. = 2, P = 0.1263). The lower number of third and fourth instar larvae collected is probably the results of the L. decemlineata survey that ended only 16 days before the destructive sampling. During that sampling, all L. decemlineata eggs observed were removed and thus L. decemlineata population had only 16 days to build up again before the destructive sampling.
Natural enemy densities were very low in this study. A total of 74 spiders, 11 predaceous hemipteran nymphs, 3 predaceous hemipteran adults, and 1 coccinellid larva were collected from all three Solanum species.

Discussion
Keeping aphids at low densities on potato plants is of primary importance as Mowry (2001) demonstrated that damages caused to tubers by PLRV are linearly correlated to M. persicae densities. The two experiments performed in this study showed that S. polyadenium and S. tarijense were generally more resistant than S. tuberosum cv. Désirée to the two most important aphid species present in Quebec, i.e. M. persicae and M. euphorbiae. As suggested by laboratory experiments (LeRoux et al. 2007(LeRoux et al. , 2008, S. stoloniferum was also more resistant than S. tuberosum cv. Désirée to M. persicae in the field, but not to M. euphorbiae. On the other hand, S. pinnatisectum was found to be as susceptible to, or more susceptible than, S. tuberosum cv. Désirée to aphids. This contrast with Pelletier and Clark (2004) that demonstrated, in laboratory conditions, that the same accession of S. pinnatisectum was resistant to M. euphorbiae. This difference is possibly due to the impact of the environment on some resistance factors (Tingey and Gibson 1978; Obrycki and Tauber 1984;Gianfagna et al. 1992;Nihoul 1993) and highlights the importance of field experiments when evaluating pest resistance.
Solanum tarijense and S. polyadenium also had negative impacts on potato flea beetle and leafhopper populations. Apart from showing that S. tarijense is resistant to the flea beetle E. cucumeris, the results indicate that the observed resistance of S. polyadenium to E. cucumeris, and of S. polyadenium and S. tarijense to the leafhopper E. fabae (Sleesman 1940;Tingey and Gibson 1978;Flanders et al. 1992;Pelletier and Michaud 1995) also occurs in Quebec field condition.
More importantly, Colorado potato beetle, L. decemlineata, laid significantly less eggs on both S. tarijense and S. polyadenium than on S. tuberosum cv. Desirée. Again, the resistance of many Solanum species to L. decemlineata has previously been reported in other geographic areas, notably in New Brunswick, Canada (Pelletier et al. 1999;Pelletier and Dutheil 2006).  reported that the resistance mechanism of S. polyadenium PI 230463 (the same accessions) was mainly antibiosis as L. decemlineata laid more eggs on this species than on S. tarijense and other species. However, we observed significantly more L. decemlineata egg clutches on S. tuberosum cv. Désirée than on S. polyadenium, suggesting an antixenosis-based resistance in the wild S. polyadenium. Both experiments were done in field conditions but in different geographic areas, so the difference observed could be due to different growing or field conditions inducing differences in resistance factors.
For both S. tarijense and S. polyadenium, the difference in pest density was mainly conferred by the plant resistance characteristics since natural enemies' density was very low. This low density is not surprising as previous studies demonstrated that the density of most natural enemies in potato fields closely follows that of M. persicae and M. euphorbiae (Karley et al. 2003;Kabaluk et al. 2006