Alfalfa fields promote high reproductive rate of an invasive predatory lady beetle
- First Online:
- Cite this article as:
- Kajita, Y. & Evans, E.W. Biol Invasions (2010) 12: 2293. doi:10.1007/s10530-009-9644-8
Invasive insect species often may attain high fecundity in agricultural habitats, thereby contributing to their establishment in new geographic regions and their displacement of similar native species. Such may be true for predatory lady beetles (Coleoptera: Coccinellidae) that have been introduced to North America in recent decades, raising concerns of adverse impact on native lady beetles. In northern Utah, Coccinella septempunctata L. first appeared in 1991, and is now predominant among lady beetles especially in alfalfa fields. We assessed the suitability of alfalfa fields as breeding habitat for females of C. septempunctata and the native, similarly sized Coccinella transversoguttata richardsoni Brown. The timing and amount of egg production differed significantly between C. septempunctata and C. transversoguttata as populations of aphids increased through spring and early summer. Reproduction by both species conformed to the egg window hypothesis, with populations of the predators producing most eggs before aphid numbers peaked. But consistently among fields and years, females of C. septempunctata produced more eggs, and did so earlier in the spring, than C. transversoguttata females even at low prey density. Furthermore, C. septempunctata females were more successful than females of C. transversoguttata in approaching their maximum body weights and reproductive output as measured in the laboratory under ideal conditions. The strong reproductive success of C. septempunctata may contribute to its displacement of C. transversoguttata in irrigated alfalfa in the generally arid Intermountain West of North America and to its establishment as an abundant species in this region of North America.
KeywordsInvasive speciesReproductionNative speciesPredatory lady beetlesCoccinella septempunctata
Agricultural fields are often highly suitable habitats for invasive insect species. These fields can provide invaders with favorable physical conditions (e.g., as provided by irrigation) and abundant food resources, including both host plants for herbivores and pest insects for natural enemies (e.g., Evans and Lanham 1960; Duelli et al. 1990; McKone et al. 2001; Rand and Louda 2006; Rand et al. 2006). Insects introduced for biological control may spread from such habitats to become well-established in natural habitats as well (e.g., Henneman and Memmott 2001; Holt and Hochberg 2001; Sheppard et al. 2004).
We address here the suitability of agricultural habitat in promoting the establishment of an invasive lady beetle Coccinella septempunctata L., and its displacement of a native congener Coccinella transversoguttata richardsoni Brown in the Intermountain West of North America (Evans 1991, 2004). Recent, high-profile invasions of North America by species of predatory lady beetles (Coleoptera, Coccinellidae) have raised widespread concern regarding potentially adverse effects on native North American lady beetles (Howarth 1991; Simberloff and Stiling 1996; Elliott et al. 1996; Obrycki et al. 1998a; Louda et al. 2003). In the generally arid landscape of the Intermountain West, irrigated alfalfa often supports high populations of the pea aphid (Acyrthosiphon pisum, Hemiptera: Aphididae) and provides a favorable, mesic environment for aphidophagous lady beetles. Since its arrival and initial establishment in the early 1990s, C. septempunctata has consistently maintained large populations in alfalfa fields while at the same time the densities of the formerly abundant C. transversoguttata and other native lady beetles have declined (Evans 2004; Elliott et al. 1996; Turnock et al. 2003). C. septempunctata occurs as well in many other (both native and disturbed) habitats, although most often in lower numbers (personal observations; see also Evans 2000; Hesler and Kieckhefer 2008). The high abundance of C. septempunctata in alfalfa reflects its use of this same habitat with pea aphids in its native (Palearctic) range (e.g., Honěk 1982, 1985). Alfalfa in the Intermountain West, as the most widely planted crop in the region, may serve as an important habitat promoting regionally high numbers of C. septempunctata.
The numerical dominance of C. septempunctata over native lady beetles in alfalfa suggests that this habitat is especially suitable for this invasive lady beetle in comparison with native competitors. We test this hypothesis here by comparing the ability of overwintering females of C. septempunctata and the similarly sized C. transversoguttata to reproduce in this habitat. Superior reproductive success of the invasive species in this habitat might account in part for its displacement of its native congener. We compare the seasonal reproductive patterns of both lady beetle species with the temporal dynamics of aphid populations in alfalfa. Such comparison is made in part with reference to the egg window hypothesis (Hemptinne et al. 1992; Kindlmann and Dixon 1993; Dixon 1997, 2000) and its key prediction that aphidophagous lady beetles lay most eggs prior to the attainment of peak numbers in local populations of their aphid prey. In addition, the comparison is made in light of the oft-stated hypothesis that high fecundity can be a key life-history trait promoting the success of invasive species (e.g., Baker 1974; Crawley 1986; Hogmire et al. 1992; Rejmanek and Richardson 1996; Boivin et al. 2008; Hodek and Michaud 2008; Kajita 2008).
Materials and methods
Lady beetle and aphid sampling
Lady beetles and pea aphids were sampled each week in alfalfa fields of northern Utah, near Logan in Cache County, from mid April through early July in 2004 and 2005 (lady beetles engaged in little reproduction from mid summer on; Kajita 2008). Two fields were sampled in each year. In 2004, these fields were located at the Utah State University (USU) Animal Science and Caine Dairy Farms. In 2005, fields were located at the USU Cache Junction and Wellsville Farms. The alfalfa in these fields was first cut for hay on June 2, 2004, and June 24, 2005. A strip (15 × 50–60 m) of alfalfa was left uncut in each field. This strip was sampled thereafter for the next 2–3 weeks until the cut alfalfa had re-grown to 30 cm height and could once again be sampled for both lady beetles and pea aphids.
Densities of pea aphids were determined weekly by sampling 20 alfalfa stems at each of 5 randomly selected locations at each census in each field in 2004, and 10 alfalfa stems at each of 25 random locations in each field in 2005. Stems were shaken within a white bucket (30 cm in diameter × 35 cm in height) and the dislodged pea aphids were counted (previous use of this technique was shown to dislodge more than 97% of the aphids from the stems; Evans, personal observations).
To concentrate lady beetle adults in alfalfa fields for collection of large numbers, 10 m2 plots within each field were sprayed with a sugar solution (15% sucrose) each week during the growing season (e.g., Evans and England 1996). Sugar acts as an arrestant (but not an attractant) for lady beetles (Evans and Swallow 1993 and references therein). Consumption of the sugar does not result in egg production (and oviposition) in C. septempunctata and C. transversoguttata, which instead is stimulated by aphid consumption (Richards and Evans 1998). Lady beetles were collected from plots during the next 2–5 days following sugar spraying, and brought to the lab immediately. The females’ wet weights were measured, and then each female was placed individually in a Petri dish (5.5 cm in diameter, 1.5 cm in height) with one drop of sugar water (15% sucrose) on the lid. Dishes were kept in an incubator (22°C, 16L:8D) for 48 h. Eggs laid in the dishes were counted at 24 and 48 h. In those relatively few instances in which females cannibalized some of their own eggs, the residue (or presence of eggshell) from each cannibalized egg (a spot where the egg had been attached to the substrate) was noted and included in counts of eggs produced (Hironori and Katsuhiro 1997). Maintaining females in this fashion provided a non-lethal method to estimate the number of mature or nearly mature eggs that a female had at the time of collection. Females held without prey will proceed to lay such eggs (rather than resorb them) over a 48-h period upon removal from prey (Richards and Evans 1998; Honĕk et al. 2008; Kajita and Evans 2009). Sample sizes for different periods of the season, as defined below in Data Analyses, are presented in the Results.
Sweep-net sampling (10 sweeps × 10 samples) was conducted in 2004 to examine the number of lady beetle larvae occurring in the alfalfa fields after the first cut (i.e., when larvae became abundant enough to sample) from June 11 to July 14. Sampling was initiated near the center of the fields, with successive samples taken along a transect towards the edge of the field. Thereafter, visual censuses (of 60 person-minutes) were conducted for pupae along the edge of each field (where pupae could be found most readily). All collected larvae and pupae were placed in a vial, and immediately brought back to the lab to distinguish species. To confirm species identity, each larva or pupa was kept individually in a Petri dish with prey (pea aphids) until it molted into an adult.
In 2005, visual censuses were conducted weekly to examine the population dynamics of adult predatory lady beetles in fields. These censuses were conducted in areas other than in the sucrose plots. As with sweep sampling in 2004, each census (of 60 person-minutes) began in the center of the field and continued along a straight transect toward the edge of the field. Adult lady beetles were collected from the alfalfa foliage, placed in a vial, and identified to species thereafter in the lab.
Periods during which adult lady beetles and aphids were sampled in alfalfa fields in 2004 and 2005, with periods identified both by numbered sampling weeks and by dates
1–4 (April 19–May 16)
4–6 (May 9–May 30)
5 (May 17–24)
7 (May 31–June 6)
6–7 (May 25–June 7)
8–9 (June 7–20)
8 (June 8–14)
10 (June 21–27)
9–13 (June 15–July 19)
11–14 (June 28–July 25)
Differences in weather conditions between years resulted in different seasonal timing of lady beetle reproductive activity between 2004 and 2005. In particular, the spring of 2005 was cooler and wetter than the spring of 2004, and correspondingly lady beetle reproductive activity occurred later in 2005 (Table 1).
To compare between years, the first week (“week 1”) for the field studies in each year was considered to be the third week of April (beginning April 19 and 18 in 2004 and 2005, respectively). This is the week that sampling of alfalfa fields began in 2004, whereas the first week of sampling in the cooler spring of 2005 was week 4 (May 9). As indicated in Table 1, data were pooled from individual females (collected in the field and evaluated in the lab) according to weeks of collection that correspond to the seasonal periods (and sub-periods) of reproductive activity as described above.
The reproductive activity of females of C. septempuctata and C. transversoguttata was compared during the different periods of the season by conducting three sets of analyses. All analyses were performed by using SAS 9.1 (SAS Institute 2002).
First, a comparison was made of the percentages of field collected females that were gravid when collected and produced eggs when held subsequently for 48 h in the laboratory. This comparison was made for females of the two species collected during the early and late periods (as defined above), and during each of the 3 peak reproductive sub-periods (early, mid, and late). Because seasonal reproductive patterns (including seasonal percentages of females that were gravid) were very similar overall between the two fields sampled in a given year, these patterns were compared for C. septempunctata and C. transversoguttata by pooling data from the two fields. To account for small sample sizes (particularly of C. transversoguttata females) in comparisons, an exact test (Monte Carlo Estimate, n = 10,000) was used for the analyses (sample sizes are presented below with the results of the statistical tests).
Second, the reproductive activity of C. septempunctata and C. transversoguttata was also compared for the number of eggs laid during 48 h after collection. For this comparison, only those females that laid eggs (one or more) were included. A two-way analysis of variance (ANOVA) was used to compare between the two species and the two years, the number of eggs (transformed by square root) laid. The analysis was focused on the mid peak time period each year (weeks 6 and 7 in 2004, and weeks 8 and 9 in 2005), when reproductive activity overall was at its height (patterns between species and years during mid peak were apparent also at lower overall rates of reproduction both earlier and later in the year; Kajita 2008).
Third, rates of egg production of gravid females of C. septempunctata and C. transversoguttata collected from the field at mid peak were compared to the potential rates that females could achieve under ideal conditions. Thus, the rates of field collected gravid females were estimated as the number of eggs laid in the 48 h following collection divided by the mean number of eggs laid in 48 h by a reproductively active, overwintered female when maintained in the laboratory and provided with excess pea aphids (estimated as 162.9 for C. septempunctata, and 116.3 for C. transversoguttata; Kajita 2008). Data were transformed by square root, and then analyzed by two-way ANOVA (species, year, species × year).
An index was calculated for the total number of eggs laid each week by the populations of C. septempunctata and C. transversoguttata females during 2005 (when visual censuses were conducted to determine seasonal patterns of adult abundance). The number of eggs per female (for all females combined, including those that did not lay eggs) was multiplied times relative abundance (i.e., the number of adults counted per 60 person-minutes of visual censusing).
Three-way ANOVA (species × year × period) was used to compare changes in wet weight of all females of C. septempunctata and C. transversoguttata collected in alfalfa fields between early and peak periods. Wet weights of females collected in alfalfa fields in 2004 and 2005 during the peak period were also compared by two-way ANOVA after dividing them by the mean wet weight of a fully fed female (estimated as 63.0 mg for C. septempunctata and 55.4 mg for C. transversoguttata; Kajita 2008).
In 2004, adults of C. septempunctata were observed to be the most abundant lady beetles present in the alfalfa fields, and they were collected in greatest numbers from sugar plots throughout the season. In 2005, weekly visual censusing of adults revealed that C. septempunctata was again the most common species present throughout the spring. At the two fields, 72.7% of adults in early weeks and 55.5% in peak weeks were C. septempunctata, while 7.3 and 10.7% were C. transversoguttata. Almost 100% of adults at both fields in late weeks were C. septempunctata.
Results of three-way ANOVA (species, year and period) for wet weight of females of Coccinella septempunctata and C. transversoguttata collected from alfalfa fields during early and peak periods in 2004 and 2005
Species × year
Species × period
Year × period
Species × year × period
Field collected females of both species during the peak period in 2004 and 2005 weighed less than females provided with excess aphids in the lab (Kajita 2008). Mean weights of field collected females of C. septempunctata were 79 ± 1% of maximum potential weight in both 2004 and 2005, compared with relatively lower mean weights of females of C. transversoguttata (73 ± 2 and 75 ± 1% of maximum weight in 2004 and 2005, respectively; two-way ANOVA: effect of species F1,620 = 13.38, P = 0.0003; effect of year F1,620 = 0.29, P = 0.59; interaction of species × year F1,620 = 0.49, P = 0.49).
During the cool spring of 2005, very few females (less than 3%) of both C. septempunctata and C. transversoguttata laid eggs during the early period (weeks 4–6) (Fig. 4b). However, as the temperatures warmed by week 7, once again a much greater percentage of C. septempunctata females (50%) than of C. transversoguttata females (12%) became reproductively active and laid eggs. A greater percentage of C. septempunctata females continued to lay eggs thereafter (including during the late period); however, there were not significant differences between C. septempunctata and C. transversoguttata during the late peak and late period (week 10, and weeks 11–14, Fig. 4b).
Consistently among fields and between years, gravid C. septempunctata females laid more eggs during the mid-peak period than gravid C. transversoguttata females (two-way ANOVA: effect of species F1,398 = 34.47, P < 0.0001, interaction of year × species F1,398 = 1.15, P = 0.18). Thus, gravid C. septempunctata versus C. transversoguttata females laid 54.6 ± 2.4 (±SE) versus 30.8 ± 4.3 eggs when held for 48 h upon collection from alfalfa fields in 2004, and 31.7 ± 1.3 versus 19.2 ± 1.5 eggs in 2005. Overall, the average egg production of both species was greater in 2004 than in 2005 (effect of year F1,398 = 96.87, P < 0.0001).
In both years, the ratio of egg production of field-collected females to mean egg production of females maintained on an excess diet of pea aphids in the laboratory was greater for C. septempunctata than for C. transversoguttata during the mid-peak period (2004: 0.33 ± 0.01 versus 0.26 ± 0.04; 2005: 0.26 ± 0.01 versus 0.16 ± 0.02; two-way ANOVA: effect of species F1,398 = 6.04, P = 0.014, effect of interaction of year × species F1,398 = 1.09, P = 0.29). For both species, the field rate of reproduction more closely matched the maximum potential rate in 2004 than in 2005 (effect of year F1,398 = 35.37, P < 0.0001).
Most larvae (89%) and pupae (98%) that occurred in the Animal Science and Caine Dairy alfalfa fields in 2004 belonged to C. septempunctata. A few individuals of other species, Hippodamia convergens Guerin, H. quinquesignata (Kirby), and H. sinuata crotchi (Casey), were collected (see Kajita 2008 for details). Interestingly, no larvae or pupae of C. transversoguttata occurred in the samples.
The invasive C. septempunctata has much greater success in reproducing in the alfalfa habitat than does the native C. transversoguttata. While producing greater numbers of eggs throughout the season, females of C. septempunctata began producing eggs earlier in the season and also more nearly reached their potential egg production (as measured in the laboratory) than did females of C. transversoguttata. Further reflecting that the invasive species was particularly successful in reproducing in this habitat, most of the larvae and pupae recovered from fields were individuals of C. septempunctata. These results support the generalization that invasive species are often characterized by unusually high fecundity (e.g., Baker 1974; Crawley 1986; Hogmire et al. 1992; Rejmanek and Richardson 1996; Boivin et al. 2008; Hodek and Michaud 2008).
The much greater reproductive activity of C. septempunctata than C. transversoguttata in alfalfa was also readily apparent in the analysis of seasonal population patterns as measured in 2005. Interestingly, females of C. septempunctata and C. transversoguttata collectively (i.e., as populations) produced most eggs before pea aphid numbers peaked, as predicted by the egg window hypothesis for aphidophagous lady beetles (Hemptinne et al. 1992; Kindlmann and Dixon 1993; Dixon 1997, 2000). The hypothesis states that the optimal reproductive strategy for a female lady beetle is to lay her eggs early in the development of a local aphid population (i.e., before aphid numbers peak), such that her offspring can complete development before aphid numbers decline severely and immature lady beetles become highly vulnerable to starvation, cannibalism and intraguild predation. Other field studies have also supported the egg window hypothesis (Hemptinne et al. 1992; Dixon 2000; Osawa 2000).
Why is the invasive C. septempunctata able to reproduce more successfully in alfalfa fields than the native C. transversoguttata? Throughout the spring, females of C. septempunctata in alfalfa fields achieved and maintained higher body weights as well as higher rates of egg production (relative to maxima as established in the laboratory) than did females of C. transversoguttata. Such success may reflect greater foraging ability by C. septempunctata not only for preferred aphids which occurred at low density, but also for a variety of other prey that aphidophagous lady beetles are known to consume (Hagen 1987; Pemberton and Vandenberg 1993; Hodek and Honěk 1996; Davidson 2008). Such prey include larvae of the alfalfa weevil that reach very high densities every spring and that can serve as alternative prey when pea aphid populations are low (Evans 2004; Evans and Toler 2007; Davidson 2008).
Differences in life-cycle phenology also may contribute to the earlier and greater reproductive activity of C. septempunctata versus C. transversoguttata, as has been suggested in other comparisons of invasive and resident insects (e.g., Boivin et al. 2008). North American populations of C. septempunctata have stronger tendency to be univoltine than do populations of C. transversoguttata (Obrycki and Tauber 1981; Phoofolo et al. 1998; Kajita 2008; Hodek and Honěk 1996). Among overwintering adults therefore, those of C. septempunctata may have more time in the summer and fall to prepare for winter diapause, and consequently may emerge the following spring with relatively high fat reserves with which to initiate reproduction (e.g., Hodek and Honěk 1996; Barron and Wilson 1998; Dixon 2000). In addition, as observed in visual censuses in the present study, overwintered females of C. septempunctata tend to arrive in alfalfa fields earlier in the spring than do females of C. transversoguttata.
The highly successful invasion of North America in recent decades by C. septempunctata and other introduced lady beetles (especially Harmonia axyridis Pallas) has led to widespread concern about their effects on native lady beetles (e.g., Louda et al. 2003). In alfalfa fields of the Intermountain West, populations of C. septempunctata have increased greatly, while at the same time populations of C. transversoguttata and other native lady beetles have declined (Evans 2000, 2004). Similar patterns in various habitats in other parts of North America (e.g., Elliott et al. 1996; Turnock et al. 2003; Hesler and Kieckhefer 2008; Brown and Miller 1998; Brown 2003) have stimulated great interest in potential causes of such species replacement.
Intraguild predation (IGP) has been widely hypothesized as one mechanism for the successful establishment of C. septempunctata (e.g., Obrycki et al. 1998a; Kajita et al. 2000, 2006; Snyder et al. 2004; Yasuda et al. 2004). IGP by C. septempunctata adults on C. transversoguttata eggs or larvae could have contributed to the observed greater reproductive success of C. septempunctata in alfalfa fields. This was not addressed in the present study, however, and in general the importance of such IGP in accounting for replacement of native lady beetles by C. septempunctata has yet to be demonstrated in the field (e.g., Obrycki et al. 1998b).
Exploitative competition may also be important in accounting for the displacement of native lady beetles by C. septempunctata in alfalfa fields studied here. Aphid populations in alfalfa during the present studies were low in comparison to densities recorded prior to the introduction of C. septempunctata (Evans 2004). In succeeding well in reproducing and reaching high numbers in alfalfa, C. septempunctata may now be limiting the population growth of pea aphids in this habitat. Consequently, native lady beetles may have abandoned alfalfa fields to forage elsewhere. As indicated by the results of the present study, females of the native C. transversoguttata, in contrast to females of C. septempunctata, achieve relatively little reproductive success while foraging in alfalfa fields with current, low aphid populations.
In summary, irrigated alfalfa is a distinctive habitat in the arid Intermountain West of North America. Following the introduction of this crop to the region in the mid nineteenth century, alfalfa fields have provided a widespread, very favorable, mesic environment for aphidophagous lady beetles (Evans 2000). The field data presented here indicate that alfalfa is a much more suitable habitat for reproduction by C. septempunctata than by the native species C. transversoguttata. The large degree of reproductive success achieved by C. septempunctata in this habitat may intensify its competitive effect on C. transversoguttata and thereby contribute to its displacement of the native congener. Our results further suggest that alfalfa, as an abundant habitat of the western landscape, may often produce large numbers of C. septempunctata that can then disperse widely as adults among many other habitats as well (including habitats that can be less favorable for reproduction). Further study, including study of reproductive success of this invasive and other lady beetle species in habitats other than alfalfa, is required to evaluate whether such a process may contribute to adverse effects on numbers of native lady beetles across the landscape (Evans 2004; Hesler and Kieckhefer 2008).
We thank N. Davidson, M. Anderson, J.-S. Lee, N. Pitigala, B. Lowry, V. Weerasekera, S. Davidson, A. Jones, B. Hadfield, B. Hammond, and R. Capener for assisting in the field and lab; D. Knudsen of the Utah Agricultural Experiment Station (USU) for accommodating our use of alfalfa fields; and D. Alston, F. Messina, T. Pitts-Singer, R. Schmidt, and two anonymous reviewers and the editor, Dr. Simberloff, for their very helpful comments on the manuscript. This research was supported by the Utah Agricultural Experiment Station, and the Ecology Center, Utah State University, and complied with the current laws of the United States.