Department of Biology, William Paterson University, Wayne, NJ. 07470. New Bolton Center, School of Veterinary Medicine, University of. Pennsylvania, 382 ...
Air Temperature and Relative Humidity Effects on Behavioral Activity of Blacklegged Tick (Acari: Ixodidae) Nymphs in New Jersey STEPHEX G. VAIL'
AND
GARY SMITH'
J. Med. Entomol. 35(6): 1025-1028 (1998)
ABSTRACT Air-temperature and relative humidity data were used to explain variation in behavioral activity of lxodes scapularis Say nymphs. We estimated behavioral activity as the residual variation in drag-sample data after seasonal changes in population density were removed by regression. The seasonal decline in drag samples between June and August 1995 on field plots at Morristown National Historical Park, NJ, can be described by a simple negative exponential function. Residuals around a fitted exponential were significantly correlated with temperature and with relative humidity measured at the leaf-litter surface, and explained 34 and 44% of the variance, respectively. Multiple regression on temperature and relative humidity explained 51%of the variance. These regressions estimated the explanatory power of microclimate, independent of seasonal correlations,and might provide a basis for day-to-dayprediction ofhuman exposure to Lyme disease.
KEY WORDS Zxodes scapularis, Lyme disease, microclimate, behavioral activity, questing
A B TO PREDICT ~ the activitv of Ixodes sca~ularisSav nymphs might have considerable value in programs for prevention of Lyme disease. Nymph activity, as measured by drag sampling, depends on population densities of nymphs and on variation in host-seeking behavior. Population densities of nymphs are likely to vary with microclimate in a complex way because they depend on rates of development, mortality, dispersal, and host finding success (Vail e t al. 1994). In contrast, microclimatic conditions may have a strong and relatively direct effect on host-seeking behavior (Daniel and Dusbabek 1994). Our purpose in this study was to investigate the effects of temperature and relative humidity o n behavioral activity of I. scapularis nymphs in a natural population in New Jersey. lxodes ticks are exceptionally sensitive t o temperature and relative humidity, compared with other ticks (Knulle and Rudolph 1982, Sonenshine 1991). 1.scapularis larvae and n ,v m* ~ h sunder controlled t e m ~ e r a tures and humidities in the laboratory experience high mortality even at moderately low humidities (Stafford 1994) and moderately high temperatures (unpublished data). Adult host-seeking behavior also reflects this sensitivity to conditions. During t h e winter, field studies indicate that adults are active only above a threshold ambient air temperature of -4°C (Duffy and Campbell 1994) and that their activity is positively correlated with temperature above that threshold (Daniels e t al. 1989). For reasons that are unclear, t h e temperature threshold for coordinated adult activity in laboratory experiments is somewhat higher at -9Department of Biology, William Paterson University, Wayne, NJ 07470. New Bolton Center, School of Veterinary Medicine, University of Pennsylvania, 382 West Street Road, Kennett Square, PA 19348.
11°C (Clark 1995). In Mississippi, Goddard (1992) was able to account for 23% (in 1989-1990) and 41% (in 1990-1991) of t h e variance in sampled adult abundance using a regression model with habitat, temperature, daylength, and "weather type" as predictors. Loye and Lane (1988) found that temperature and relative humidity accounted for 31-66% and 40-66% of t h e variance in adult activity of Ixodes pucijicus Cooley & Kohls in California, respectively. Lane and Stubbs (1990), however, found temperature and humidity to b e unreliable predictors of activity for adult I. paci$cus. Little information exists relative t o t h e activity of lxodes nymphs as a function of micrometeorological conditions. Lane et al. (1995) investigated t h e summer activity of I. p a c i j m nymphs as a function of relative humidity and temperature at t h e soil surface. They found a significant positive correlation with humidity and a significant negative correlation with temperature, accounting for -29 and -30% of t h e variance, respectively. W e found no published studies of I. scapularis nymph activity as a function of microclimate. Here w e report correlations of I. scapularis nymph behavioral activity with temperature and relative humidity measured at t h e leaf-litter surface.
Materials and Methods Our general approach was to estimate behavioral activity as the residual variation i n drag-sample data after seasonal changes in population density were removed by regression. All field sampling of ticks was performed by standard dragging methods using a square ( 1 by 1 m ) of white corduroy fabric mounted on apole and dragged by a r o p e over t h e surface of t h e
-258519811025-1028$02.00/0O 1998 Entomological Society of America
J O C ~ AOFL PY~EDIcAL ENTOMOLOGY
1026
ground for a selected distance (Risley and Hahn 1994).A grid of 9 contiguous 10by 10-m field plots was established in open understory of deciduous forest at the Jockey Hollow unit of Morristown National Historical Park in Morristown, NJ. Approximately once every 5 d between 6 June and 9 August 1995 each plot was sampled, with replacement, by dragging the entire surface systematically, stopping every 20 m to count and place all ticks in a specimen vial. Upon completion of each plot (10 by 10 m), all ticks were released by scattering them uniformly throughout the plot. Sampling was generally performed between 1000 and 1700 hours and was not attem~tedin the rain. Tem~erature and relative humidity were measured in each plot (10 by 10 m) at the surface of the leaf litter on sampling days using a certified-calibrated portable digital thermistor/ hygrometer (model 11-661-7B, Fisher, Springfield, NJ). The probe was placed directly on the leaflitter surface but not in direct sunlight. We did not attempt to measure temperature or humidity profiles within the leaf litter. Means of temperature, humidity, and drag-sample data were computed across all 9 plots for each sampling day. Mean nymph densities (nymphs sampled per ~ O m2) O were plotted as a function of date and fit with a simple exponential function by nonlinear regression (Tilling et al. 1992). An exponential function was used because it provided a good fit in previous studies at the same site (Vail and Smith 1997) and corresponds to the simplest biological assumption of a constant per capita disappearance rate (Vail et al. 1994). Drag-sample data were adjusted for season by computing residuals (i.e., deviations from the fitted function [Sokal and Rohlf 19951) . These residuals (seasonally adjusted data) were then regressed on mean temperature and on mean relative humidity by simple linear regression. Because temperature and relative humidity are typically correlated (Lane et al. 1995),we also computed a multiple regression on both variables to estimate their combined explanatory power (FOX et al. 1994).Finally, we computed regressions of unadjusted nymph totals (raw drag-sample data, not deviations from the seasonal trend) on temperature and humidity for comparison with the adjusted results.
EXPONENTIAL DECLINE OF NYMPH ABUNDANCE 16
DATE Fig. 1. Mean densities of unfed I. scapularis nymphs on experimental plots ( N = 9) versus date. The curve is a simple exponential function fitted by nonlinear regression ( Y =
10.14e-.OW,R2 = 0.58).
Simple linear regression revealed no significant effect of relative humidity on mean nymph density (df = 10, r = 0.34, P = 0.27). Temperature had a significantly negative effect on mean nymph density (df = 10, r = 0.66, P = 0.02, slope = -0.59). Deviations of nymph density from the fitted exponential (i.e., residuals) were separately regressed on temperature and relative humidity. Temperature had a significantly negative effect (df = 10, r = 0.59, P = 0.04, slope = -0.34, Fig. 2) and humidity had a significantly positive effect (df = 10, r = 0.67, P = 0.02, slope = 0.13, Fig. 3) on residual density. Temperature and humidity explained 34 and 44% of the residual
NYMPH ACTIVITY vs. TEMPERATURE 6
Temperature varied from 18 to 32°C between 6 June and 9 August 1995 and relative humidity varied from 55 to 89%. Simple linear regression revealed no significant relationship between either mean temperature or mean humidity and date (df = 10,PTmp = 0.15, P, = 0.93). Mean nvmuh densitv. , ,, as estimated bv drag sam~les. declined steadily from a value of -101i00 m2 in early June to a value of -21 100 m2in early August 1995 (Fig. '1. A fitted simple functionaccounted for 58%ofthetotalvarianceindensit~( N = 12,R2 = 0.58). Comparable data from the same plots in 1994 explained 51% of the variance (Vail and Smith 1997).
-
--
INTERCEPT=8.55 SLOPE=-0.339 R2=0.34 PC.05
a
> o z~ J
Results
Vol. 35, no. 6
s
2-
=) V)
w n
-2
-
16
18
20
22
24
26
28
30
32
34
TEMPERATURE ( O C ) Fig. 2. Residualnymph density, asdeviation of mean nymph density from fitted exponential, versus mean temperature at leaf-litter surface. Linear regression is significant (2' = 8.85-0.339X,P < 0.05,RZ = 0.34).
VAL AND S M ~T E : M~TUR HUMID^, E,
November 1998
NYMPH ACTIVITY vs. HUMIDITY
50
60
70
80
90
100
RELATIVE HUMIDITY (%) Fig. 3. Residualnymph density,computed as deviation of mean nymph density from fitted exponential, versus mean temperature at led-litter surface. Linear regression is significant ( Y = -9.29 +.134X, P < 0.05, R2 = 0.44). variance, respectively (RZTmp = 0.34, RZAH= 0.44, Figs. 2 and 3 ) . Multiple linear regression including temperature and humidity revealed n o significant effect on unadjusted nymph density (R2 = 0.44, 0.05 < P < 0.10). Multiple linear regression revealed significant effects of temperature and humidity on nymph residual density and explained 51% of t h e variance (R2 = 0.51, P = 0.04).
AND I.
scapularis A
m
1027
midity had no significant effect on raw (unadjusted) drag data (P = 0.27), but it had a significant effect on adjusted data (P = 0.02, R2, = 0.44). Our method also eliminates t h e possibility of a spurious correlation with season. In other experiments, w e have discovered spurious correlations of behavior with temperature and humidity over die1 time (unpublished data). Spurious correlations are possible, in principle, if temperature, humidity, and tick activity all vary with seasonal time for independent reasons. Statistical control of seasonal variation, using a procedure like ours, mitigates this possibility. T e m ~ e r a t u r eand relative humiditv accounted for approximately half of the variation among drag samples after seasonal changes were removed. Although we are encouraged that 2 simple environmental measures can account for this much of the variation, it is important to emphasize that much remains to b e explained. It is possible that solar radiation (Atwood and Sonenshine 1967), soil temperature (Lane et al. 1995), or other environmental conditions (Daniel and Dusbabek 1994) contribute to t h e unexplained variation. Acknowledgments Bob Masson (National Park Service) provided access to field sites at Morristown National Historical Park. Ellen See and Linda Rohonyi (William Paterson University) provided able assistance in the field. This work was supported by aNational Institutes of Health grant to G.S. (No. AI-32059),by National Park Service contract to S.V. (No. 1443px183095038), and by grants to S.V. from the William Paterson University Center for Research and Assigned Release Time program. This manuscript is TechnicalReport No. 112 of the William Paterson University Center for Research.
Discussion Previous studies demonstrated that microclimate affects adult behavioral activity of various Ixodes species (Loye and Lane 1988, Daniels e t al. 1989,Lane and Stubbs 1990, Duffy and Campbell 1994, Clark 1995) and nymph activity of I. pacijificus (Lane et al. 1995). Our data extend this conclusion to I. scavularis nymphs as well. Temperature and relative humidity alone explained 33 and 44%,respectively, of t h e variation in seasonally adjusted drag samples (Figs. 2 and 3 ) . Together they accounted for 51% of the variance. These regressions estimate t h e extent to which microclimatic data can explain short-term variation in behavioral activity and might provide a basis for dayto-day prediction of human exposure to risk of Lyme disease. W e analyzed residual variation in drag-sample data after seasonal changes in population density were removed by exponential regression. This is essentially the reverse of a procedure used by Daniels e t al. (1989) t o remove day-to-day variation in adult activity from estimates of population density. Because our nymph data were adjusted for t h e seasonal decline in abundance, they provide a relatively pure measure of the influence of microclimate on host-seeking behavior, as distinct from effects o n demographic rates and tick densities. It is interesting in this regard that hu-
References Cited Atwood, E. L., and D. E. Sonenshine. 1967. Activity of the American dog tick, Damacentar cariabilis (Acarina: Ixodidae), in relation to solar energy changes. Ann. Entomol. Soc. Am. 60: 354-362. Clark, D. D. 1995. Lower temperature limits for activity of several ixodid ticks (Acari:Ixodidae): effects of body size and rate of temperature change. J. Med. Entomol. 32: 449 - 452. Daniel, M., and F. Dusbabek. 1994. Micrometeorological and microhabitat factors affecting maintenance and dissemination of tick-borne diseases in the environment, pp. 91-138. In Daniel E. Sonenshine and Thomas N. Mather [eds.],Ecological dynamics of tick-borne zoonoses. Oxford University Press, New York. Daniels, T. J., D. Fish, and R. C. Falco. 1989. Seasonal activity and survival of adult lxodes dammini (Acari: Ixodidae) in southern New York State. J. Med. Entomol. 26: 610-614. Duffy, D. C., and S. R. Campbell. 1994. Ambient air temperature as a predictor of activity of adult Irodes scapul a d (Acari: Ixodidae). J. Med. Entomol. 31: 178-180. Fox, E., J. Kuo, L. Tilling and C. Ulrich. 1994. Sigmastat for Windows user's manual. Jandel, San Rafael, CA. Goddard, J. 1992. Ecological studies of adult lxodes scapularis in central Mississippi: questing activity in relation to time of year, vegetation type, and meteorological conditions. J. Med. Entomol. 29: 501-506.
1028
JOURNAL OF
MEDICALENTOMOLOGY
Knulle, W., and D. Rudolph. 1982. Humidity relationships and water balance of ticks, pp. 43-70. In F. R. Obenchain and R. Galun [eds.], Physiology of ticks. Pergamon, New York. Lane, R. S., and H. A. Stubbs. 1990. Host-seeking behavior of adult lxodes pacijcus (Acari: Ixodidae) as determined by flagging vegetation. J. Med. Entomol. 27: 282-287. Lane, R. S., J. E. Kleinjan, and G. B. Schoeler. 1995. Die1 activity of nymphal Damacentor occidentalis and Ixodes pacificus (Acari: Ixodidae) in relation to meteorological factors and host activity periods. J. Med. Entomol. 32: 290 -299, Loye, J. E., andR. S. Lane. 1988. Questing behavior of Ixodes pacijicus (Acari: Ixodidae) in relation to meteorological and seasonal factors. J. Med. Entomol. 25: 391-398. Risley, L. S., and S. Hahn. 1994. Lyme disease in northern New Jersey: distribution of black-legged ticks (Ixodes scapularis) and prevalence of the spirochete Borrelia burgdo&&. Bull. N.J. Acad. Sci. 39: 17-21.
Vol. 35, no. 6
Sokal, R. R., and F. J. Rohlf. 1995. Biometry, 3rd ed. Freeman, New York. Sonenshine, D. E. 1991. Biology of ticks, vol. 1.Oxford University Press, New York. Stafford, K. 1994. Survival of immature Ixodes scapularis (Acari: Ixodidae) at different relative humidities. J. Med. Entomol. 31: 310-314. Tilling, L., J. Kuo, and E. Fox. 1992. Sigmaplot for Windows user's manual. Jandel, San Rafael, CA. Vail, S., G. Smith, and C. Lord. 1994. Population biology of lxodes scapularis, the vector of Lyme disease in the easte m and north central United States, pp. 263-277. In M. E. Scott and G. Smith [eds.], Parasitic and infectious diseases. Epidemiology and ecology. Academic, New York. Vail, S., and G. Smith. 1997. Density-dependent seasonal dynamics of blacklegged tick (Acari: Ixodidae) nymphs. J. Med. Entomol. 34: 301-306. Receivedforpublication 9Apri11997; accepted 25 June 1998.
