The multicolored Asian ladybeetle Harmonia axyridis ( Coleoptera : Coccinellidae ) in Quebec agroecosystems ten years after its arrival

The multicolored Asian ladybeetle, Harmonia axyridis (Coleoptera: Coccinellidae), was first reported in southern Canada in 1994. In this study, we examined the status of the ladybeetle in Quebec agroecosystems seven to ten years after its arrival. We integrated the findings from eight independent field studies carried out in pome fruit, grapes, medicinal crops, field corn, sweet corn, sweet pepper, lettuce, and soybean. Nine aphidophagous and 6 coccidophagous coccinellid species were sampled on these crops. Among these aphidophagous species, three invasive (H. axyridis, Coccinella septempunctata and Propylea quatuordecimpunctata) as well as one indigenous species (Coleomegilla maculata lengi) were consistently present. Higher richness was observed in perennial crops, where the 6 coccidophagous coccinellids were almost exclusively present. Higher abundance of aphidophagous coccinellids was observed in field and sweet corn. The multicolored Asian ladybeetle was a dominant species of the coccinellid assemblage in all crops monitored. 737 * Corresponding author; e-mail : genevievelabrie@yahoo.ca tering sites, benefiting from a cold-free space (Labrie et al., 2007). Harmonia axyridis is thought to affect indigenous ladybeetle species either by exploitative competition or by intraguild predation (Hironori & Katsuhiro, 1997; Yasuda & Ohnuma, 1999; Dixon, 2000; Kajita et al., 2000; Yasuda et al., 2001; Michaud, 2002; Snyder et al., 2004; Lucas et al., 2007). Its establishment was followed by important decreases in natural populations of several local ladybeetle species such as Brachiacantha ursina (Fabricius), Cycloneda munda (Say) and Chilocorus stigma (Say) (Colunga-Garcia & Gage, 1998). In recent years, Quebec agroecosystems have been invaded by two other coccinellids, Propylea quatuordecimpunctata L., probably in the 1960s (Gordon, 1985) and Coccinella septempunctata L. in 1973 (Larochelle, 1979). These ladybeetles, particularly C. septempunctata, have also earlier affected coccinellid assemblages, reducing abundance of indigenous species such as Coccinella transversoguttata richardsoni Brown, Adalia bipunctata L., Coccinella novemnotata Herbst and Hippodamia convergens Guérin-Méneville (Wheeler & Hoebeke, 1995; Elliott et al., 1996; Ellis et al., 1999; Evans, 2004). Seven to ten years after its arrival in Eastern North America, this study examines the relative abundance of the multicolored Asian ladybeetle and other coccinellid species in three perennial and five annual crops of southern Quebec. MATERIAL AND METHODS


INTRODUCTION
Biological invasions usually occur in three successive stages, (1) transport of the invader to the target area, (2) establishment and growth of invasive populations, and (3) dispersion of the invader into adjacent areas (Shea & Chesson, 2002).The complexity of trophic interactions in the recipient ecosystem is one of the main opposition factors of the ecosystem to the invasion.This is because it is more difficult to get established in a complex food web than in a simpler one (Hewitt & Huxel, 2002;Shea & Chesson, 2002;Stachowicz et al., 2002).Based on the availability of plant resources and the limited number of trophic links, agro-ecosystems are highly susceptible targets for biological invasions.
Seven to ten years after its arrival in Eastern North America, this study examines the relative abundance of the multicolored Asian ladybeetle and other coccinellid species in three perennial and five annual crops of southern Quebec.

Pome fruit
The study was carried out in 1999 and 2000 in an organic apple orchard (planted in 1986) in Henryville (45°8´N, 73°11´W), Quebec, Canada.In both years, adult coccinellid populations were assessed from May to August by white sticky traps and by tapping apple trees on the third and sixth peripheral rows of the orchard.Tapping was done weekly over a 1 m 2 cloth placed under a series of 20 apple trees.For each tree, eight branches were tapped three times.Branches were selected in the four cardinal points at 50 cm and 150 cm from the ground.Ten white sticky traps were also placed at these two distances from the ground on apple trees.Once a week, traps were observed and coccinellid species identified.

Grapes
The study was carried out from 2000 to 2002 in two commercial vineyards (both of 0.7 ha and planted in 1983-84) in southern Quebec, Canada.Located in Dunham (45°07´N, 72°51´W), the L'Orpailleur vineyard comprised only vines of the cultivar White Seyval.The Dietrich Jooss vineyard was located in Iberville (45°16´N, 73°11´W), and it comprised De Chaunac, Cayuga White and White Seyval cultivars.A two-liter plastic container was held below a berry cluster and the cluster was tapped twice.Coccinellids fallen into the container were identified, counted, and then released in the vineyard.Each week from June to mid-September, 100 grapevines (50 "De Chaunac" and 50 "Cayuga White") were sampled at random at the Dietrich Jooss vineyard and 100 "White Seyval" grapevines were sampled at random at L'Orpailleur.Coccinellid adults were sorted out in the laboratory, identified and tallied.Voucher specimens were deposited at the Eastern Cereal and Oilseeds Research Centre, Agriculture and Agri-Food Canada, Ottawa, Canada.Except for fungicides, no insecticides were used in the plot where coccinellids were collected (Bostanian et al., 2003).

Medicinal crops
From the end of May to mid-September 2003, four medicinal crops, Achillea millefolium Linnaeus, Valeriana officinalis Linnaeus, Hypericum perforatum Linnaeus and Echinacea purpurea (Linnaeus) Moench, were sampled weekly in fields located at L'Assomption (45°50´N, 73°25´W), Quebec, Canada.The crops had been planted in 2002.Each field was divided in 36 sections, 5 m apart of each other and of other plant species.In each section, two plants were examined visually each week for coccinellid eggs, larvae and adults.

Sweet corn
During the summer 2000 and 2001, a corn field of the D'Artagnan cultivar was sampled from the beginning of July to the end of August.The field was located in Saint-Hyacinthe (45°37´N, 72°57´W), Quebec, Canada.It was divided into eight sections with 10 sampling stations in each section.Visual observations of coccinellids were done weekly on five plants, near the 10 sampling stations.Adults, larvae and eggs of coccinellids were identified and counted.

Field corn
The study was done in three commercial corn fields in Farnham and l'Ange-Gardien (45°24´N, 72°44´W), Quebec, Canada, in 2002 and2003.During the summer, the fields were sampled seven times from the beginning of August to mid-October.In the center of each field, 30 plants were randomly selected and inspected for the presence and abundance of aphids and coccinellids.Eggs and early instars of coccinellids were brought to the laboratory and reared to the adult stage in an incubator (25°C, 65% R.H. and 16L : 8D photoperiod) to be identified to the species.

Sweet pepper
The study was done in a field located in Laval (45°30´N, 73°32´W), Quebec, Canada.It was sampled two times per week from early July to mid-August 2002.Each sample consisted of 50 plants randomly selected and visually inspected.The presence and abundance of aphids and coccinellids were recorded.

Lettuce
From mid-May to mid-September 2002, 25 fields of head lettuce, romaine and leaf lettuce were sampled.The fields were located in Sherrington (45°10´N, 73°31´W), Quebec, Canada.Each two weeks, five samples per 0.4 ha were randomly selected and visually examined for coccinellid abundance.

Soybean
The study was carried out in 10 soybean fields in two agricultural areas near Quebec City (from 45°11´N to 46°44´N, from 70°57´W to 74°23´W), Quebec, Canada, between June and August 2002 and 2003.Samples were collected every two weeks by one to two hauls of sweep net in 10 zones in each field on the upper vegetation.Insects collected were brought to the laboratory for identification.Only adult coccinellids were identified.

Statistics
As the eight crops under study had different agronomic and ecological conditions, only intra-crop statistical analysis were performed.In crops where more than one year of data was available, a contingency Table (G test) compared the relative proportions of the different species of coccinellids as well as the proportion of invasive versus indigenous species between the different years (SAS Institute, 1996).When required, subsequent two-group G-tests were carried out.

Sweet pepper
During July and August 2002, 109 adult coccinellids from four different species were collected on more than 1800 plants; C. maculata represented 34% of samples, H. axyridis 42.2%, P. quatuordecimpunctata 14.7% and C. septempunctata 9.1% (Fig. 2).No sampling was carried out in September because the grower applied pesticides following an aphid outbreak.The percentage of introduced coccinellids was 66.1%.

DISCUSSION
The results of this study indicate that seven to ten years after its arrival, H. axyridis is clearly one of the dominant coccinellid species in most crops of Eastern Canada.This invasive species was observed in three perennial and five annual monitored crops (Table 1).Other unpublished data from the same area indicated that this coccinellid is also present in raspberry, canola, and celery.In raspberry for example, H. axyridis represented 1.1% of the coccinellids collected in 1995 and 18.7% in 1996, out of a total of 374 and 916 specimens collected, respectively (Roy, unpubl.data).In soybean, the arrival of this coccinellid coincided with the arrival of the soybean aphid, Aphis glycines Matsumura (Fox et al., 2004;Mignault et al., 2006).Our study reports results from several agroecosystems, where the species is well adapted.However, its eurytopic habits include forest and other natural systems (Tedders & Schaefer, 1994).The species has also been observed on aspen, birch, willow and fir (Berthiaume et al., 2000).It has also been observed in swamp ecosystems and urban areas (e.g.Montreal, Canada).Geographically, this invader is reported throughout southern Quebec and as far North as Abitibi (48°30´N, 77°47´W).This colonization reflects a huge success for this invasive species.
In all monitored crops, nine aphidophagous and six coccidophagous species were observed (Table 1).Highest richness was observed in perennial crops (pome fruit, grape), where the coccidophagous species were almost exclusively observed (except for soybean, where one individual of B. ursina was observed in 2003).Among the aphidophagous species, three invasive (H.axyridis, C. septempunctata and P. quatuordecimpunctata) were consistently present, as well as one indigenous species (C.maculata) (Table 1).While for a given crop, relative proportion of coccinellid species varied between years (Fig. 1, 2), these invasive species actually represented between 12 to 89% of the coccinellids encountered.Some annual crops presented very high proportion of invasive species, such as soybean, with 89 and 87% of invasive species, in 2002 and 2003 respectively (Fig. 2D).These three invasive species are polyphagous predators attacking a great diversity of prey including aphids, mites, scales, and lepidopteran and coleopteran larvae (Hodek & Hon k, 1996).They are also eurytopic species, exploiting a great diversity of habitats.Consequently, they may have caused considerable changes in various ecosystems.For example, in 1978, indigenous species represented 100% of the coccinellid assemblage in southern Quebec field corns.The coccinellids comprised 74.5% Hippodamia tredecimpunctata tibialis (Say) and 24.1% C. maculata (Coderre, unpubl. data).For over five years now, H. tredecimpunctata tibialis has completely disappeared from the corn fields.While in our study the highest abundance of aphidophagous coccinellids was observed in sweet and field corn, three invasive (H.axyridis, C. septempunctata and P. quatuordecimpunctata) and only one indigenous (C.maculata) species were present, suggesting high incidence of invasive coccinellid species in corn fields.Brown &Miller (1998) andLucas et al. (2002) showed that a species may be more susceptible to impact by an invader than other species because of their common spatial niche in the ecosystem.Other indigenous species such as C. maculata were thought to be affected by the presence of H. axyridis (Musser & Shelton, 2003;Hoogendoorn & Heimpel, 2004).However, our study showed the continued presence of the indigenous C. maculata along with the three introduced ladybeetles in each agroecosystem examined.Coleomegilla maculata is found on the lower parts of corn plants and avoids competition with H. axyridis which is found on the upper parts of plants (Musser & Shelton, 2003).As a result of this niche partitioning, C. maculata is able to coexist with H. axyridis in corn fields, suggesting that this species may not suffer adversely from the presence of these invasive species in the future.
The multicolored Asian ladybeetle is now well established in most crops and agro-ecosystems in Canada and USA.Several crops monitored in our study presented a predominant proportion of H. axyridis in the different years of the study (Figs 1, 2).The arrival of the multicolored Asian ladybeetle has been associated with significant changes in native guild structures (Tedders & Schaeffer, 1994;LaMana & Miller, 1996;Brown & Miller, 1998;Michaud, 2002;Snyder et al., 2004).The consequences of such an invasion may also adversely affect other natural enemies of aphids such as hover flies, lacewings, gall midges, and hemipteran predators by exploitative competition and intraguild predation (Horn, 1991;Koch et al., 2003;Lucas et al., 2007).High abundance and presence of the multicolored Asian ladybeetle and two other invasive species suggest their adverse impact on indigenous species.As agroecosystems may be subjected to several dynamic processes, including potential invasion by other coccinellid species, our results should be revisited in the context of a longer term study (over 20 years after invasion).

TABLE 1 .
Presence/absence and richness of coccinellid species in Quebec perennial and annual crops.