The caterpillars of Utetheisa ornatrix, the ornate bella moth, feed on host plants in the genus Crotalaria (Fabales: Fabaceae), which, in Florida, encompasses 4 native and 10 introduced species. In the laboratory, Utetheisa ornatrix male larvae developed faster and resulted in larger adults, when raised on Crotalaria incana vs. Crotalaria lanceolata (2 species native to Africa). In the wild, this can potentially give a selective advantage to individuals that develop on C. incana. The presence of these and other exotic host plants, some of which are quite common (e.g. C. pallida, C. spectabilis, C. retusa), may be skewing the evolution of U. ornatrix, as well as its phenology and distribution. A constant temperature of 60 °F (15.6 °C) caused U. ornatrix to develop twice as slowly as it did at 72 °F (22.2 °C), but it was sufficiently high for successful development. Sentinel larvae placed in cages in the field during Oct–Dec, when temperatures averaged ≈ 62 °F, developed successfully. Regardless of the rearing conditions and the host plants used in the laboratory, male larvae developed slightly more slowly than female larvae. This translated in their emerging from the pupae later than the females, which is supported by field observations.
The ornate moth, Utetheisa ornatrix L., is found in the eastern and midwestern regions of the United States and throughout the Neotropics. Its caterpillars feed on host plants in the genus Crotalaria (rattlebox plants) (Fabales: Fabaceae). These moths are brightly, aposematically colored, with shades of yellow, red, pink, orange, and white adorning their wings, and white, spotted heads, thoraxes, and abdomens - a coloration that advertises their demonstrated distastefulness to predators (Eisner & Eisner 1991). The chemical ecology of this species has been studied in depth (Conner 2008 and references therein). For instance, pheromone production in males has been determined to directly depend on their diet as caterpillars, and only occurs when the diet contains Crotalaria pyrrolizidine alkaloids (PAs) (Conner et al. 1981). This is important, because the genetic quality of males correlates with the pheromone signal (Kelly et al. 2012), larger males contain more PAs than the smaller males (Conner et al. 1990), and evidence suggests that larger ones may be favored by sexual selection (LaMunyon & Eisner 1993; Iyengar et al 2002). The complex interactions between PA-containing plants and Lepidoptera in general are summarized by Boppré (1990) and will not be reviewed here. Additionally, a large body of knowledge also exists on the interaction of U. ornatrix specifically with dietary PAs extracted from plants and fed to the larvae (e.g., Cogni et al. 2012).
The main focus of the present paper is to report the observed differences between 2 introduced host plants of the genus Crotalaria, which have become naturalized diets of U. ornatrix in the southeastern United States, and to discuss how these differences may be important in the evolution of this species.
Many species of Crotalaria have been intentionally introduced to North America for use as green manure and soil improvement. Some of these species are considered invasive and have even been proven to be toxic to cattle, with their toxicity varying depending on their PA content (Williams & Molyneux 1987). In Florida, there are 14 different species of Crotalaria, 10 of which are introduced (Wunderlin & Hansen 2011) (Table 1). Utetheisa ornatrix, which formerly fed on just 4 host plant species native to Florida, presumably is now using all of these Crotalaria, some of which are more abundant than the native ones. For instance, the local population of U. ornatrix larvae used in the present study naturally feeds on 3 species that are native to Africa (Crotalaria incana L., C. pallida Ait. and C. lanceolata E. Mey.) and one species native to Asia (C. spectabilis Roth), and native C. pumila Ortega; of these C. lanceolata is the most prevalent. Other populations that we observed in north-central Florida may have only a single (native or exotic) Crotalaria species available to them.
In the present study, we compared Crotalaria incana and C. lanceolata as host plants for U. ornatrix during laboratory rearing. Specifically, we determined how the growth rate of larvae and the size of the resulting adult moths would vary when raised on one plant vs. the other. Utetheisa ornatrix, though mostly a tropical species, is also found in climates where it becomes relatively cold, so we tested how larval and pupal development are affected by temperature.
MATERIALS AND METHODS
Even though U. ornatrix females can mate multiple times, eggs are mostly fertilized by the sperm of a single male (LaMunyon & Eisner 1993). Thus, to reduce genetic variability within a sample group, offspring of a single wild-collected female were used for each of the 2 experiments. Two female moths were netted behind the Florida Museum of Natural History, Gainesville, Florida. The first female, whose offspring were used in the temperature-dependence experiment, was caught on 15 May 2012, while the host plant-quality experiment was conducted with eggs from a female caught on 1 Oct 2012. Females were fed 5% sugar solution twice per wk. They laid eggs in the cups in which they were held, and, when the neonates hatched, they were provided with host plant leaves.
During the temperature-dependence experiment, larvae were initially raised gregariously in 16 oz. plastic cups at a room temperature of approximately 72 ± 2 °F (22.2 °C). Leaves of Crotalaria lanceolata were replaced 3 times per wk at which time the cups were also cleaned. During molting into the ultimate instar, larvae were randomly removed and divided into 2 groups. One group was raised further with only one larva to a cup at ≈ 72 ≈F (22.2 ≈C), and the second group was raised with one larva per cup at approximately 60° ± 2 °F (15.6 °C) (in the collections room, McGuire Center, Florida Museum of Natural History).
To test whether the results of our laboratory rearing at a constant temperature of ≈60 °F (15.6 °C) translate to the environmental conditions of Oct–Dec in north-central Florida, where temperatures fluctuated between the low 30's °F and low 80's °F and averaged at ≈ 62°F (NOAA 2012), we placed a dozen penultimate instars in 2 mesh cages outside and supplied them with plant material of C. lanceolata. Cages were placed in a partially shaded area in the habitat of U. ornatrix and were restocked weekly with either green seed pods (cage 1) or stems with leaves (cage 2).
For the experiment that involved raising larvae on Crotalaria incana and C. lanceolata, neonate larvae were split randomly into 2 groups. At first, larvae in both groups were raised gregariously in 16 oz. cups, but beginning with the penultimate instar, they were kept individually in 2 oz. plastic cups. Three times per week, the leaves and the cups were replaced and the progress of larval development was recorded. This experiment (Trial 1) was repeated using another egg batch laid by the same female starting 2 wk later (Trial 2) with a slight modification: larvae that were raised on C. incana were switched for a period of 5 days (while in their penultimate instar) to C. lanceolata and then returned to feeding on C. incana. A sample of host plant leaves was dried, ground-up into powder, and analyzed for nitrogen content using an Eager 200 CHN analyzer.
SPECIES OF CROTALARIA FOUND IN FLORIDA.
Adult moths resulting from all trials were frozen upon emergence and then mounted on spreading boards. Their forewing length (from the base of the wing to the tip) was measured with electronic calipers. All voucher specimens were labeled and deposited in the collection of the McGuire Center for Lepidoptera Research, Florida Museum of Natural History, Gainesville. Statistical analyses comparing the duration of larval and pupal development and forewing length of the different groups were conducted by a two-way ANOVA and an unpaired t-test.
The 2 groups of larvae that were raised on different host plants developed at different rates (Table 2; Fig. 1). The larvae raised on C. lanceolata took significantly longer to reach maturity. In the second trial, where larvae feeding on C. incana were switched during their penultimate instar for 5 days to feeding on C. lanceolata and then were switched back to feeding on C. incana, the larvae pupated with a 5–7-day delay in comparison with the larvae on C. incana in the first trial. The larvae raised on C. lanceolata in trial 2, developed at about the same rate. There was a difference in the nitrogen content of the 2 host plants (C. incana: 3.4–3.6%, C. lanceolata: 2.8–2.9%).
In the first trial, the mean size (forewing length (FW)) of male adult moths that resulted from feeding on C. incana significantly exceeded that of male moths raised on C. lanceolata (P < 0.03), while the sizes of females raised on C. incana were not significantly different from females raised on C. lanceolata (Fig. 2). In the second trial, caterpillars raised on C. incana/C. lanceolata/C. incana also resulted in males which were significantly larger than those raised on C. lanceolata (P < 0.03). Similarly to Trial 1, the difference observed between females was not statistically significant (P < 0.098).
During laboratory rearing at different temperatures, the last instars' and pupal developments were twice as fast at ≈ 72 °F (22.2 °C) than at ≈60 °F (15.6 °C) (Table 2). Low temperature in this study led to development of smaller adult moths, however this study had small sample size and the latter results need to be confirmed by additional replications. The sentinel larvae in the outdoor cages successfully developed into adult moths that emerged between 31 Dec and 7 Jan, developing at approximately the same rate as the laboratory population held at ≈ 60 °F (15.6 °C).
Male vs. Female Rate of Development
In all of our experiments, male larvae developed more slowly than their female counterparts. This statistically significant difference in development time varied from 1 to 5 days depending on the trial (e.g., P < 0.05—individually reared moths ≈72 °F (22.2 °C); P < 0.04—moths reared on C. incana in Trial 1). The difference was quite obvious because females were always the first to emerge. This corresponds to our observations in the wild, where females were observed to initially be the more numerous sex present when the fall generation appeared in late Sep.
The observed difference between the rates of development on C. incana and C. lanceolata cannot be explained just by the slight differences that we found in the nitrogen content of the 2 host plants. Based on preliminary data produced by similar experiments, C. pallida, which had an even higher nitrogen content (3.8–4.2%), did not spur larval growth the way that C. incana did (Sourakov, unpublished). Because it has been shown that arctiid larvae do not directly absorb plant toxic alkaloids and utilize them for defense, but rather deactivate them first in the gut, which requires energy (Hartman 2008), it is possible that feeding on C. incana leads to faster larval development as a result of its lower toxicity. Analyses of seeds from many species of the genus (Williams & Molyneux 1987) indicate that C. incana has 3-times lower PA content (0.07%) than C. lanceolata (0.21%). This theory is supported by a recent laboratory study that showed that increased PAs in the artificial diet lead to slower larval development in U. ornatrix (Cogni et al. 2012). It would be interesting to assess if moths that developed on C. incana are more palatable to predators due to their lower alkaloid content. If such variation in adult moth toxicity (and consequently, in palatability) exists, it can be a potential source for automimicry (Bowers 2008) under which condition, the individuals within a population that develop on more toxic plants may serve as models for the ones that developed on less toxic plants. Testing the PA contents and palatability of the U. ornatrix adults raised on different Crotalaria species is the next logical step in determining if this is the case.
Unlike the larvae of many Lepidoptera, which stay on a single individual plant throughout their development, woolly bears (arctiid larvae) are known to travel between host plants, and U. ornatrix larvae are no exception. In the wild, a single larva may travel between host plants of different species, which is now not only possible, thanks to the introduction of many Crotalaria species to Florida habitats, but is also likely, because they may grow in close proximity, as in the case of our source population. Development was slower for larvae fed on C. incana/C. lanceolata/C. incana compared with that of larvae reared exclusively on C. incana. It may be that larvae that were first adapted to feeding on a host plant with lower toxicity may have a difficult time adapting to a more toxic host plant. Alternatively, feeding on a higher PA-level plant without growth could have been a sign of larvae trying to satisfy their PA requirements: in recent laboratory experiments (Hoina et al. 2012), U. ornatrix larvae first fed an artificial low-PA diet later preferred a high-PA diet when offered the choice, even though it led to prolonged development.
LABORATORY BIOLOGY OF UTETHEISA ORNATRIX: DEVELOPMENT TIME AND WING SIZE OF RESULTING MOTHS REARED ON DIFFERENT HOST PLANS AND UNDER DIFFERENT TEMPERATURES.
Males that resulted from feeding on C. incana exceeded in size males raised on C. lanceolata. Being larger can indicate a higher PA load carried by a male (Conner et al. 1990). This can potentially give these males an advantage during sexual selection, as was shown by LaMunyon & Eisner (1993), and can give advantages to the offspring in the form of accelerated oviposition by the female they mate and by the larger eggs she lays (Iyengar & Eisner 2002). However, elaborate experiments, similar to those described by these authors, are required to test whether larger males developed on one host plant would be preferred over smaller males developed on another.
The various plants of the genus Crotalaria, appear to have different phenologies in north central Florida. As we observed in 2012, by Dec–Jan, C. spectabilis and C. pallida in some of the sites were greatly affected by the lightest freezing, C. incana showed yellowing and loss of leaves, while C. lanceolata and the native C. pumila, on the contrary, appeared to be present in all stages (from young sprouts, to flowering and seeding plants). As late as the end of January, we found on them eggs and larvae of U. ornatrix. It is possible to speculate, based on these observations and on our laboratory rearing described above, that the phenology and the evolution of U. ornatrix in southeastern U.S. have and will be affected by the introduction of so many diverse host plants.
It is noteworthy that male larvae developed more slowly than female larvae. In most other species of Lepidoptera, the reverse is normally the case, with males being both smaller than females and emerging earlier. In most butterfly populations, for example, males are usually on the wing for a few days before females emerge and actively compete for mating opportunities by patrolling the area or even visiting female pupae. It has been shown by other studies that during sexual selection in U. ornatrix, males have a lot to lose by mating to a “substandard” female, since they become less desirable by females after the first mating (Bezzerides et al 2005). Conner et al. (1980, 1981) described in detail the courtship in which females call the males, and it was shown that males transfer ca. 10% of their body weight together with the spermatophore to females, during mating (LaMunyon & Eisner 1993). These alkaloids and other nutrients that males lose during mating are collected by the feeding larvae and are, therefore, irreplaceable. Even though that the males must choose carefully to whom they transmit these “nuptial gifts” (González et al. 1999), it has been suggested that the benefits are accrued by females, as males mate on an opportunistic basis (Iyengar & Eisener 2004). Females of U. ornatrix are capable of mating multiple times to increase their fecundity (LaMunyon 1997; Iyengar & Reeve 2010), and therefore have less at stake when choosing a mate and more at stake in doing so as early as possible and as many times as possible. Perhaps their faster development and early emergence is a mechanism for them to compete for males rather than not vice versa.
Utetheisa ornatrix is capable of developing on both Crotalaria lanceolata (a species distributed in the USA from Louisiana to North Carolina and south to Florida) and C. incana (found from Oklahoma to North Carolina to Florida) (USDA plant database) both of which have been introduced to the United States as exotic plants that are native to Africa.
The biology of U. ornatrix in the laboratory was markedly affected by the choice of the host plant. Larvae developed faster and resulted in larger adults (males only) when raised on Crotalaria incana rather than on Crotalaria lanceolata.
Constant temperatures of as low as ≈ 60 °F (15.6 °C) are not an obstacle to the successful development of these moths, even though it takes them twice as long to develop as they do at ≈ 72 °F (22.2 °C). The rate of laboratory development at ≈ 60 °F (15.6 °C) corresponds to that of these moths in the wild in north-central Florida during Oct–Jan.
We thank the staff of the Department of Chemistry, University of Florida, for analyzing the nitrogen content of host plant leaves. Alexandra Sourakov proofread the manuscript of this paper and offered helpful suggestions. Walter Judd helped to identify the host plants and brought important references to our attention. Patti and David Locascio provided logistical support. Faculty and students of the Entomology and Nematology Department (University of Florida) are acknowledged for maintaining a habitat suitable for U. ornatrix on campus. We are greatly indebted to the anonymous reviewers for their input, which greatly improved this paper.
NOAA. 2012. National Climatic Data Center ( http://www.ncdc.noaa.gov) last accessed 18 Jan 2013.
USDA. Plant Database ( http://plants.usda.gov/java/profile?symbol=CRIN5) last accessed 2 Jan 2013.