Temperature Effects on Development and Survival of Two Stable Flies, Stomoxys calcitrans and Stomoxys niger niger (Diptera: Muscidae), in La Réunion Island

Abstract

Two stable fly species, Stomoxys calcitrans (L., 1758) and Stomoxys niger niger Macquart, 1851, co-occur in La Réunion, where they are important pests of cattle. The survival and developmental rate of the immature stages were compared at five constant temperatures from 15 to 35°C. In both species, immature survival was highest at 20–25°C and markedly decreased at 15 and 35°C. At the lower temperatures, mortality was observed mainly for S. calcitrans larvae and S. niger eggs. At the higher temperatures, mainly pupae of both species died. At all temperatures, S. calcitrans survived better than S. niger. Developmental time was highly similar in both species, decreasing from 71 d at 15°C to 13 d at 30°C in S. calcitrans and from 69 d at 15°C to 14 d at 30°C in S. niger. Developmental times increased slightly at 35°C. Surprisingly, the tropical S. niger developed slightly faster than the cosmopolitan S. calcitrans at 15–20°C; the reverse was found at higher temperatures. Temperature summation models confirmed that S. niger had a lower developmental threshold than S. calcitrans (11.3 versus 12.2°C) and higher day-degree (DD) requirements to complete development (251 versus 225 DD). Overall, the results suggest that S. calcitrans is better adapted than S. niger, in terms of adult production from eggs, in the temperature range of La Réunion.

stable flies, temperature, developmental rate, stage-specific survival, La Réunion Island

Stable flies (Stomoxys spp.) (Diptera: Muscidae) are blood-sucking insects associated with livestock and wildlife throughout the world. Occasionally, they also bite humans. They represent a serious nuisance because of their painful bites, blood predation, and transmission of pathogens (Zumpt 1973, D'Amico et al. 1996). Their economic impact can be considerable (Campbell et al. 2001).

Among the 18 Stomoxys species described, two occur in La Réunion Island: the cosmopolitan Stomoxys calcitrans (L., 1758), and Stomoxys niger niger Macquart, 1851, which is also found in Africa, Madagascar, and Mauritius (Zumpt 1973). Both species occur from sea level to high-altitude plains (1,600 m) and are pests of cattle, particularly in dairy barns. During the wet and warm season (November-April), large pullulations are observed, with daily catches of up to 5,000 flies per Vavoua trap (unpublished data). In addition, stable flies are considered mechanical vectors of anaplasmosis, the primary agent of mortality in dairy cattle. They also could play an important part in the transmission of the bovine leukosis virus (Buxton et al. 1985).

To control stable fly populations and reduce their pathogenic and economic impacts in La Réunion, a research project has been set up to study both species in the field and in the laboratory. Adequate methods of control require thorough knowledge of the biology and ecology of the flies (Batra 1982). When closely related species exploit the same resources, the biological and ecological characteristics of each species must be determined, and, ultimately, how the two species interact. Otherwise, measures against the most abundant species (S. calcitrans in La Réunion) might favor the less common species.

Many studies have been conducted with the cosmopolitan S. calcitrans, mainly in North America and South Africa. Seasonal changes in abundance were described in relation to temperature and rainfall (Mullens and Meyer 1987, Greene 1989, Lysyk 1993). Experimental studies clarified the influence of temperature on immature survival, developmental time, and adult longevity and reproduction (Berry and Kunz 1977, 1978; Kunz et al. 1977; Sutherland 1979; Lysyk 1998). Using those results, Lysyk (1998) calculated the intrinsic rate of increase of S. calcitrans at different temperatures. Such data are not available, however, for S. niger. Most reports on this species relate to feeding habits. Some studies examined seasonal changes in abundance and their relationships with field temperature and rainfall, in Africa (Kangwagye 1974) and Mauritius (Kunz and Monty 1976). Ramsamy (1979) developed rearing methods suitable for S. niger, but there have been no experimental studies of the influence of environmental factors on the life stages.

In the ongoing research in La Réunion, a systematic comparison between the two species has been undertaken to gather not only new information on S. niger but also further information on S. calcitrans. Traits may vary between populations of a cosmopolitan species, especially in islands (Hedrick 1984). In our study, the influence of temperature on the development of both species was determined in the laboratory. The survival and developmental rate of three life stages (eggs, larvae, and pupae) were compared between S. calcitrans and S. niger.

Materials and Methods

Study Area.

La Réunion, the largest volcanic island of the Mascarene archipelago (2,507 km²), lies 800 km east of Madagascar (21° 20′ S, 55° 15′ E). This island rises to 3,069 m and has a tropical climate. Mean annual temperatures are 23–26°C at sea level. Annual rainfall ranges from 700 to 1,200 mm on the west coast to 3,000–5,000 mm on the east coast. La Réunion has >36,000 heads of cattle, found mainly in moderately elevated areas.

Biological Material.

Insects were from recent stock colonies of S. calcitrans and S. niger (first to third generation in the laboratory). Colonies were established in January 2002 during the annual pullulation period from flies trapped at a dairy farm in the southeastern part of the island (910-m elevation). They were maintained at 25 ± 1°C and 70 ± 10% RH under a photoperiod of 12:12 (L:D) h, in the CIRAD-3P Laboratory, La Réunion. Cages (30 by 30 by 30 cm) containing ≈1000 adult flies were placed above an oviposition substrate composed of elephant grass, Pennisetum purpureum Schumacher, crumbs. The same substrate was used as larval development medium. The flies were fed daily with bovine blood collected twice a week from a slaughterhouse and citrated at 6 g liter^-1. For this, sponges dampened with blood and warmed at 38°C in a water bath were laid on the top of each cage for 45 min. The oviposition substrate was changed daily. When pupae were observed, they were removed from the rearing medium by flotation, dried, and transferred to new cages.

Developmental Time and Survival of Life Stages.

Experiments were conducted in climatic chambers (MLR-350, Sanyo, Japan) at five constant temperatures (15, 20, 25, 30, and 35°C) and 70 ± 10% RH, under a photoperiod of 12:12 (L:D) h. These conditions were controlled daily by using a thermo-hygrograph (Jules Richards Instruments, Paris, France), a mercury thermometer, and a digital thermo-hygrometer.

After the oviposition substrate was changed in the stock cultures, freshly laid eggs were collected with a brush and transferred onto moistened pieces of dark blotting-paper (3 by 3 cm). At least 750 eggs of each species were collected over several days. The time of transfer was considered the time of egg laying. The pieces of blotting paper were put into dishes (10 cm in diameter by 10 cm in height) containing P. purpureum crumbs and covered with mosquito tulle to prevent larvae from exiting. Eggs were then observed under a binocular microscope at 4-h intervals during the day (8–10-h intervals at night) to count and remove empty chorions. This approach provided the number of young larvae hatched in each dish in the time interval. To assess the egg stage duration, hatching was assumed to occur at the middle of the time interval. After each observation, the unhatched eggs remaining on blotting paper were placed in a new dish, so that all dishes contained only larvae hatched in the same time interval. Eggs that did not hatch within 10 d were considered dead. The developmental time and survival of larvae were determined in the same way, by counting newly formed pupae in the medium. These were transferred to plastic boxes (8 by 4 by 1.5 cm) containing a slightly moistened sponge. Observations were continued to record the number of newly emerged adults in each box, which made it possible to determine the developmental time and survival of pupae. All stage durations were thus estimated, with errors ranging between 0 and 5 h (≈0.2 d). The duration of development for each individual was obtained by summing its developmental times as egg, larva, and pupa.

Temperature Summation Models.

In both species, the linear regressions were established between developmental rate (1/developmental time, in days) and temperature. When the regression was significant, the lower threshold for development (T₀), i.e., the theoretical temperature at which the developmental rate is zero, was calculated as the x intercept. The inverse of the slope of the regression line was the thermal constant (K), i.e., the number of heat units above T₀, expressed as day-degrees (DD), that are required to complete development (Honek 1996, Danks 2000).

Statistical Analyses.

All analyses were conducted using SAS (SAS Institute 2001). The effects of temperature and species on the proportions of survivors were tested using PROC GENMOD. Two-way analysis of variance was used to test the effects of temperature and species on the developmental time (PROC GLM). In both types of analysis, pairwise mean comparisons were made with the Bonferroni significance level adjustment. Linear regressions of developmental rate on temperature were calculated using PROC REG.

Results

Survival.

The total survival of the immature stages, from egg laying to adult emergence, differed significantly among temperatures (χ² = 321.40, df = 4, P < 0.001) and species (χ² = 54.88, df = 1, P < 0.001). There was a significant temperature-species interaction (P < 0.001), because the differences in survival between species were not exactly the same at all temperatures. Although S. calcitrans consistently showed a higher survival than S. niger (Fig. 1), the differences were significant only at 20°C (χ² = 10.69, df = 1, P < 0.01), 25°C (χ² = 8.96, df = 1, P < 0.01), and 35°C (χ² = 41.30, df = 1, P < 0.001).

Fig. 1

Total survival of the immature stages in S. calcitrans (▲) and S. niger (■) reared at five constant temperatures. Proportions are given with 95% confidence intervals.

In both species, survival was highest at 20 and 25°C, without any significant difference between these temperatures. At 15°C, survival decreased significantly (χ² = 40.66, df = 1, P < 0.001 for S. calcitrans;χ² = 23.43, df = 1, P < 0.001 for S. niger). Above 25°C, survival decreased slightly at 30°C, with the difference being significant in S. calcitrans only (χ² = 3.92, df = 1, P < 0.05), but there was a dramatic decrease in both species at 35°C (χ² = 48.11, df = 1, P < 0.001 for S. calcitrans;χ² = 129.19, df = 1, P < 0.001 for S. niger). At this temperature, only 35% of the S. calcitrans eggs and 6% of the S. niger eggs reached the adult stage (Fig. 1).

Survival data for the life stages (Table 1) served to specify the least resistant stage to low and high temperatures in each species. At 15°C, mortality was observed mainly in larvae for S. calcitrans and in eggs for S. niger. In contrast, at 35°C, mainly pupae of both species died.

Table 1

Proportions of survivors at different life stages, in S. calcitrans and S. niger reared at five constant temperatures

Developmental Time.

The developmental time from the egg to the adult stage differed significantly among temperatures (F = 13172.5; df = 4, 880; P < 0.001) and species (F = 11.09; df = 1, 880; P < 0.001), and there was again a significant interaction between the two factors (F = 25.58; df = 4, 880; P < 0.001).

In both species, developmental time decreased for each 5°C increase in temperature between 15 and 30°C (Fig. 2). In contrast, developmental time increased slightly when temperature increased from 30 to 35°C. All differences were significant, except for the increase in developmental time at 35°C in S. calcitrans (see tests in Table 2).

Fig. 2

Duration of development from the egg to the adult stage in S. calcitrans (A) and S. niger (B) reared at five constant temperatures.

Table 2

Mean developmental times (in days) at different life stages, in S. calcitrans and S.niger reared at five constant temperatures

At all temperatures, developmental time differed significantly between the two species, but there was an obvious temperature-species interaction: at 15 and 20°C, development was significantly shorter in S. niger, whereas at 25, 30, and 35°C it was slightly but significantly shorter in S. calcitrans (Table 2). Data for the different life stages (Table 2) show that the fast development of S. niger at 15 and 20°C resulted only from the performance of the larvae.

Temperature Summation Models.

The durations of total development in both species resulted in highly significant linear relationships between developmental rate and temperature within the 15–30°C range (Fig. 3). The linear relationships between the developmental rate of each life stage and temperature were also highly significant (Table 3).

Fig. 3

Relationships between temperature and developmental rate (1/developmental time) in S. calcitrans (▲, Sc) and S. niger (■, Sn) based on the durations of total development at four constant temperatures.

Table 3

Regression equations of developmental rate on temperature (T) in S. calcitrans and S. niger

The lower temperature thresholds (T₀) were calculated both for the development of each life stage and for total development (Table 3). T₀ for egg development did not differ significantly between species, but there were significant differences for larval, pupal, and total development, with lower values in S. niger. T₀ for total development was 11.3°C in S. niger versus 12.2°C in S. calcitrans. The day-degree requirement to complete development (K) was higher in S. niger (250.6 DD) than in S. calcitrans (224.7 DD).

Discussion

Most of the results obtained on S. calcitrans in this study are consistent with previous results on the effects of temperature on the species' survival and developmental time (Kunz et al. 1977, Sutherland 1979, Lysyk 1998). Mean developmental times observed in La Réunion are similar to those reported for North American populations reared at constant temperatures. Minor differences may be because, in the current study, the developmental time of each life stage was based only on individuals that actually reached adulthood, excluding individuals that died before completing their development. For the latter, developmental times were generally longer than in successful individuals. The greatest difference between the results of this study and those previously conducted with S. calcitrans concerns the pupal survival at high temperatures, which was higher in La Réunion. At 30°C, 99% of pupae survived in La Réunion versus 60–70% in most North American studies (Kunz et al. 1977, Lysyk 1998). The difference may reflect local adaptation to tropical climatic conditions. This observation is new for the cosmopolitan S. calcitrans, which generally shows little differentiation between widely distant regions, at least in terms of survival and developmental rate (Lysyk 1998). Note, however, that life history parameters have generally been measured from stock cultures maintained under standard laboratory conditions for many generations, which makes it difficult to compare populations (Tauber et al. 1986). Results on local adaptation that are more significant might be obtained by comparing samples recently collected in the field, as in this study.

For the first time, life history traits were compared between S. niger and S. calcitrans under controlled laboratory conditions. Developmental time is highly similar in both species: it ranges from 69 d at 15°C to 14 d at 30°C in S. niger, compared with 71 d at 15°C and 13 d at 30°C in S. calcitrans. The pattern of immature survival in relation to temperature is also similar in both species, with the highest survival at 20–25°C and a decrease at lower and higher temperatures. At low temperature, the least resistant stage differs between species (the egg in S. niger and the larva in S. calcitrans), but at high temperature the least resistant stage in both species is the pupa. In Diptera, this life stage is obviously the most susceptible to heat, probably because of harmful effects on the metamorphosis process (Bayoh and Lindsay 2004). The most obvious difference between the two species concerns the mean survival at each temperature, which was consistently lower in S. niger. Yet, it is difficult to know whether this difference reflects actual differences in the field or whether this relates to the experimental conditions. Although the rearing medium used (elephant grass) has been reported to favor S. niger (Kunz and Monty 1976, Ramsamy 1979), there may have been an undetected problem acting against the species in the laboratory. In such a case, survival differences between S. calcitrans and S. niger might be smaller in the field. This point should be reexamined under a variety of rearing conditions.

Another difference between the two species concerns the larval development at low temperatures. At 15°C, the larvae of S. niger develop faster and survive much better than the larvae of S. calcitrans. S. niger may temporarily take advantage of this trait in late winter, which should be taken into account in any attempt to explain how the two species coexist in La Réunion. The developmental thresholds and day-degree requirements of the two species also suggest that S. niger is better adapted than S. calcitrans for development at low temperatures. When T₀ and K vary inversely in directly comparable species, the species that exhibits the lower T₀ and the higher K may have a competitive advantage in cold environments (Trudgill 1995, Honek 1996). Surprisingly enough, the tropical S. niger performs better below 20°C, whereas the cosmopolitan S. calcitrans performs better at higher temperatures, at least regarding developmental rates.

In conclusion, the current study suggests that S. calcitrans is better adapted than S. niger in terms of adult production over the range of temperatures in La Réunion. Despite the slightly shorter developmental time of S. niger at low temperatures, this species' advantage may be offset by its higher mortality during development when temperature reaches 20°C. At higher temperatures, S. calcitrans consistently shows a higher developmental rate and a better survival than S. niger. These results are consistent with the high incidence of S. calcitrans in La Réunion, although some aspects remain unexplained. For example, the current study cannot account for the relatively high incidence of S. niger in the warm coastal area of the island, and its even higher incidence in Mauritius (Monty 1972, Kunz and Monty 1976), where the climate is warmer than in La Réunion. Further data are needed to clarify the effects of temperature on the population growth potential of both species, especially data on the reproductive pattern and lifetime fecundity of females at different temperatures.

Acknowledgements

We thank Gary Burkhart for critical reading of the manuscript. We are grateful to E. Tillard and K. Le Roux (Centre de coopération Internationale en Recherche Agronomique pour le Développement [CIRAD], La Réunion) and to the "Groupement Régional de Défense Sanitaire du Bétail de La Réunion" for assistance. This research was supported by funds from the CIRAD-3P, the Conseil Régional of La Réunion Island and the Centre National de la Recherche Scientifique (UMR 5175).