Chain reactions linking acorns to gypsy moth outbreaks and Lyme disease risk

Science

Volumn 279, Issue 5353, 1998, Pages 1023-1026

  Jones, Clive G ^a   Ostfeld, Richard S ^a   Richard, Michele P ^a   Schauber, Eric M ^b   Wolff, Jerry O ^c  

^a CARY INSTITUTE OF ECOSYSTEM STUDIES   (United States)

^b UNIVERSITY OF CONNECTICUT   (United States)

^c OREGON STATE UNIVERSITY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

ACORN ABUNDANCE; BLACK-LEGGED TICK; DEFOLIATION; DISEASE INCIDENCE; GYPSY MOTH; INTERSPECIFIC INTERACTION; LYME DISEASE; MOUSE; OAK FOREST;

ANIMAL EXPERIMENT; ANIMAL MODEL; ARTICLE; CONTROLLED STUDY; FOREST; LYME DISEASE; MOTH; MOUSE; NONHUMAN; POPULATION DENSITY; PRIORITY JOURNAL; RISK FACTOR; TICK;

ANIMALS; ARACHNID VECTORS; BORRELIA BURGDORFERI GROUP; DISEASE RESERVOIRS; ECOSYSTEM; FORESTRY; HUMANS; IXODES; LARVA; LYME DISEASE; METAMORPHOSIS, BIOLOGICAL; MOTHS; PEROMYSCUS; POPULATION DYNAMICS; PUPA; RISK FACTORS; TREES;

USA, (EAST);

ACARI; ANIMALIA; BORRELIA; BORRELIA BURGDORFERI; CERVIDAE; FAGACEAE; IXODES PACIFICUS; IXODES SCAPULARIS; LEPIDOPTERA; LYMANTRIA DISPAR; LYMANTRIIDAE; PEROMYSCUS; PEROMYSCUS LEUCOPUS; QUERCUS;

EID: 0032512672     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.279.5353.1023     Document Type: Article

References (49)

1. R. S. Ostfeld, C. G. Jones, J. O. Wolff, Bioscience 46, 323 (1996).
2. R. G. Lalonde and B. D. Roitberg, Am. Nat. 139, 1293 (1992).
3. V. L. Sork, J. Bramble, O. Sexton, Ecology 74, 528 (1993).
4. J. O. Wolff, J. Mammal. 77, 850 (1996).
5. G. O. Batzli, Am. Midl. Nat. 97, 18 (1977); M. V. Price and S. H. Jenkins, in Seed Dispersal, D. R. Murray, Ed. (Academic Press, Sydney, Australia, 1986), pp. 191-235.
6. G. O. Batzli, Am. Midl. Nat. 97, 18 (1977); M. V. Price and S. H. Jenkins, in Seed Dispersal, D. R. Murray, Ed. (Academic Press, Sydney, Australia, 1986), pp. 191-235.
7. J. S. Elkinton et al., Ecology 77, 2332 (1996).
8. H. A. Bess, S. H. Spurr, E. W. Littlefield, Harv. For. Bull. 22, 1 (1947).
9. R. W. Campbell and R. J. Sloan, Environ. Entomol. 6, 315 (1977).
10. _, ibid., p. 323.
11. H. R. Smith, Wild. Soc. Bull. 13, 166 (1985).
12. E. H. Forbush and C. H. Fernald, The Gypsy Moth (Wright and Potter, Boston, MA, 1896).
13. J. S. Elkinton and A. M. Liebhold, Annu. Rev. Entomol. 35, 571 (1990).
14. R. W. Campbell and H. T. Valentine, USDA Forest Service Research Paper No. NE-236 [U.S. Department of Agriculture Forest Service (USDAFS), Washington, DC, 1972); _, R. J. Sloan, For. Sci. Monogr. 19 (1977); K. W. Gottschalk, in Proceedings of the Workshop on Southern Appalachian Mast Management, C. E. McGee, Ed. (USDAFS and University of Tennessee, Knoxville, TN, 1990), pp. 42-50.
15. R. W. Campbell and H. T. Valentine, USDA Forest Service Research Paper No. NE-236 [U.S. Department of Agriculture Forest Service (USDAFS), Washington, DC, 1972); _, R. J. Sloan, For. Sci. Monogr. 19 (1977); K. W. Gottschalk, in Proceedings of the Workshop on Southern Appalachian Mast Management, C. E. McGee, Ed. (USDAFS and University of Tennessee, Knoxville, TN, 1990), pp. 42-50.
16. R. W. Campbell and H. T. Valentine, USDA Forest Service Research Paper No. NE-236 [U.S. Department of Agriculture Forest Service (USDAFS), Washington, DC, 1972); _, R. J. Sloan, For. Sci. Monogr. 19 (1977); K. W. Gottschalk, in Proceedings of the Workshop on Southern Appalachian Mast Management, C. E. McGee, Ed. (USDAFS and University of Tennessee, Knoxville, TN, 1990), pp. 42-50.
17. W. J. McShea and J. H. Rappole, VA J. Sci. 43, 177 (1992).
18. W. J. McShea and G. Schwede, J. Mammal. 74, 999 (1993).
19. A. G. Barbour and D. Fish, Science 260, 1610 (1993).
20. D. Fish, in Ecology and Management of Lyme Disease, H. Ginsberg, Ed. (Rutgers Univ. Press, New Brunswick, NJ, 1993), pp. 25-42.
21. R. S. Lane, J. Piesman, W. Burgdorfer, Annu. Rev. Entomol. 36, 587 (1991).
22. L. P. Hansen and G. O. Batzli, Can. J. Zool. 56, 230 (1978).
23. Three pairs of open grids (165 by 165 m and 180 by 110 m for one grid) 100 to 250 m apart, with pairs separated by 1 to 3 km, were located in upland oak forests (57 to 70% oak relative basal area) at the Institute of Ecosystem Studies. Grids were 11-by-11 arrays (12 by 10 for one grid) of trap stations 15 m apart, with two Sherman live traps per station at the centers of 15-by-15-m grid cells used for gypsy moth sampling. A wooden nest box, containing cotton batting nesting material replaced twice a year, was attached at 1.3 m to a large tree [26 to 58 cm in diameter at breast height (DBH)] in each of 20 grid cells; boxes were stratified to maximize intercell distances. These nest boxes have no effect on mouse densities [J. O. Wolff and D. S. Durr, J. Mammal. 67, 409 (1986); D. T. Krohne, J. F. Merritt, S. H. Vessey, J. O. Wolff, Can. J. Zool. 66, 2170 (1988)] and were colonized by mice only. Mouse densities, measured as minimum number known alive (MNKA) per hectare, were estimated from (i) monthly mark-recapture over 2-day periods from the 120 or 121 trap stations on all grids, April-November 1995 and 1996, and July 1994 for two control grids; (ii) trapping of new and recaptured animals during the removal period every 2 to 4 days on experimental grids, 26 June-30 July 1995; and (iii) monthly mark-recapture from nest boxes, February-April 1996. Traps were baited daily with oats, set 2 to 3 hours before dusk, and checked the following two mornings. Traps were not operated during the day or in December-March to avoid risks of overheating or hypothermia of animals. During mark-recapture, all mice were given numbered metal eartags at first capture. Trap station or nest box, tag number, gender, body mass, and age and reproductive condition (for females, vaginal patency and evidence of pregnancy and lactation; for males, whether testes were descended) were recorded at all captures, and animals were released at the point of capture. During the removal period, the same variables were recorded, but tagged and untagged animals were moved to a site 4.5 km away. One hundred forty, 230, and 239 mice were removed from the three experimental grids, respectively. None of the tagged relocated animals were recaptured on grids. Care of animals was in accordance with institutional guidelines.
24. Three pairs of open grids (165 by 165 m and 180 by 110 m for one grid) 100 to 250 m apart, with pairs separated by 1 to 3 km, were located in upland oak forests (57 to 70% oak relative basal area) at the Institute of Ecosystem Studies. Grids were 11-by-11 arrays (12 by 10 for one grid) of trap stations 15 m apart, with two Sherman live traps per station at the centers of 15-by-15-m grid cells used for gypsy moth sampling. A wooden nest box, containing cotton batting nesting material replaced twice a year, was attached at 1.3 m to a large tree [26 to 58 cm in diameter at breast height (DBH)] in each of 20 grid cells; boxes were stratified to maximize intercell distances. These nest boxes have no effect on mouse densities [J. O. Wolff and D. S. Durr, J. Mammal. 67, 409 (1986); D. T. Krohne, J. F. Merritt, S. H. Vessey, J. O. Wolff, Can. J. Zool. 66, 2170 (1988)] and were colonized by mice only. Mouse densities, measured as minimum number known alive (MNKA) per hectare, were estimated from (i) monthly mark-recapture over 2-day periods from the 120 or 121 trap stations on all grids, April-November 1995 and 1996, and July 1994 for two control grids; (ii) trapping of new and recaptured animals during the removal period every 2 to 4 days on experimental grids, 26 June-30 July 1995; and (iii) monthly mark-recapture from nest boxes, February-April 1996. Traps were baited daily with oats, set 2 to 3 hours before dusk, and checked the following two mornings. Traps were not operated during the day or in December-March to avoid risks of overheating or hypothermia of animals. During mark-recapture, all mice were given numbered metal eartags at first capture. Trap station or nest box, tag number, gender, body mass, and age and reproductive condition (for females, vaginal patency and evidence of pregnancy and lactation; for males, whether testes were descended) were recorded at all captures, and animals were released at the point of capture. During the removal period, the same variables were recorded, but tagged and untagged animals were moved to a site 4.5 km away. One hundred forty, 230, and 239 mice were removed from the three experimental grids, respectively. None of the tagged relocated animals were recaptured on grids. Care of animals was in accordance with institutional guidelines.
25. Unless otherwise noted, response variables were examined for grid pair effects, using an appropriate parametric or nonparametric comparison, followed by treatment comparisons using paired tests if P < 0.05 for grid pair effects and unpaired tests if P > 0.05 for grid pair effects. Where one-tailed treatment comparisons were used because there were a priori hypotheses, this is reported; otherwise, comparisons were two-tailed. A single time point was used to compare densities before mouse removal; interpolated densities were used for comparisons at the midpoint of the mouse removal period. Mouse densities were In-transformed before grid effect analyses of variance (ANOVAs) and t tests.
26. Two hundred forty to 242 trees, greater than 7 cm DBH and greater than 2 m in height, without nest boxes, one tree pair per grid cell, were used for moth sampling on trees. Half the trees were banded with slitted, folded, burlap skirts (30 cm) tied at 1.3 m. One-third of all tree pairs were oak-oak, one-third were oak-non-oak, and one-third were non-oak-non-oak. Banded and unbanded tree pairs were alternated across each grid. Late-stage larvae use burlap bands as daytime refuges (11), and their development was monitored until fourth and fifth instars were prevalent. The number of living larvae of all instars on or under burlap bands on all banded trees was counted during dry days between 8 and 15 June and was expressed as the mean number of larvae per tree, with a tree-to-tree within-grid variance estimate. Larval, pupal, and most egg mass density data were analyzed by Kruskal-Wallis ANOVA by ranks and showed no grid pair effects (P > 0.05). Treatment effects were analyzed with Mann-Whitney U tests unless otherwise noted.
27. Pupation, eclosion, mating, and egg-laying occur in the daytime resting locations of late-stage larvae, and the number of flightless females determines egg mass density. Female pupae took a mean (±SE) of 12.7 (±0.4) days (n = 42) to develop to adults, before immediately mating and laying egg masses. The time taken for female pupae to eclose was determined by monitoring individuals every other day from 30 June to 7 August on grid cells with banded oak-oak tree pairs (n = 34 monitored survivors). Because no monitored pupae survived on control grids, additional female (sixth-instar) larvae (n = 8) were enclosed in small wire mesh cages as they began to pupate, to prevent predator access. Caged adults and egg masses were removed and are excluded from density estimates. The fate of individual native female pupae on a given date (uneclosed or eclosed, based on characteristic splitting and adult emergence holes), and characteristic signs of mortality agents on attacked pupae (predation, insect parasitism, fungal hyphae, and other or unknown), were recorded every other day on banded oak-oak tree pairs. Predation on freezedried female pupae was measured with a technique modified from Smith (10). Mass-reared female pupae (USDA Animal and Plant Health Inspection Service, MA) were freeze-dried and affixed with beeswax, which retains an imprint of the species-specific tooth marks of vertebrate predators, in groups of five onto burlap panels 20 by 15 cm. Twenty or 21 panels per grid were stapled under the bands of one of the pair of oak-oak banded trees on 10 July. Pupae were monitored daily to day 9 (except day 6) and then every other day to day 18 or until all pupae had been attacked. Judging by tooth marks and damage characteristics, we recorded attacks as being due to vertebrates, invertebrates, both, or an unknown agent. (10). The total number of successfully eclosing pupae was determined from counts on all banded trees at the end of the eclosion period and was expressed as the mean number per tree, with a tree-to-tree within-grid variance estimate.
28. New egg masses on or under burlap bands were counted on all banded trees and expressed as the mean number per tree, with a tree-to-tree within-grid variance estimate. Twelve randomly selected 15-by-15-m grid cells per grid were also censused for egg masses on or under bands, at heights less than 2 m on banded and unbanded trees, on small saplings, dead trees, woody debris, litter, and rocks. Firm new egg masses were distinguished from spongy old egg masses with visibly hatched eggs by gentle pressing. Mean (±SE) new egg mass densities per grid cell were 0.028 (±0.028) on control grids and 0.305 (±0.121 ) on experimental grids (P = 0.036, one-tailed Mann-Whitney U test). Most egg masses are laid at heights less than 2 m in low-density moth populations [P. Skaller, Environ. Entomol. 14, 106 (1985)].
29. -1, were 2.59 (±0.36) in July 1994 and 38.81 (±9.24) in July 1995 (one-tailed paired t test on In-transformed data, P = 0.045). Mice attacked a mean (±SE) of 1.06 (±0.44) freeze-dried pupae per day in 1994 and 36 (±14) in 1995 (one-tailed paired t test on In-transformed data, P = 0.003). Mean (±SE) egg mass densities per tree on or under burlap bands were 0.11 (±0.02) in 1994 and 0.004 (±0.004) in 1995 (one-tailed paired t test, P = 0.0862).
30. J. Piesman and J. S. Gray, in Ecological Dynamics of Tick-borne Zoonoses, D. E. Sonenshine and T. N. Mather, Eds. (Oxford Univ. Press, New York, 1994), pp. 327-350.
31. M. L. Wilson, S. R. Telford, J. Piesman, A. Spielman, J. Med. Entomol. 25, 224 (1988).
32. J. Piesman, J. G. Donahue, T. N. Mather, A. Spielman, ibid. 23, 219 (1986).
33. J. R. Levine, M. L. Wilson, A. Spielman, Am. J. Trop. Med. Hyg. 34, 355 (1985); T. N. Mather, M. L. Wilson, S. I. Moore, J. M. C. Ribeiro, A. Spielman, Am. J. Epidemiol. 130, 143 (1989).
34. J. R. Levine, M. L. Wilson, A. Spielman, Am. J. Trop. Med. Hyg. 34, 355 (1985); T. N. Mather, M. L. Wilson, S. I. Moore, J. M. C. Ribeiro, A. Spielman, Am. J. Epidemiol. 130, 143 (1989).
35. R. S. Ostfeld, M. C. Miller, K. R. Hazier, J. Mammal. 77, 266 (1996).
36. R. S. Ostfeld, K. R. Hazier, O. M. Cepeda, J. Med. Entomol. 33, 90 (1996).
37. R. S. Ostfeld, Am. Sci. 85, 338 (1997).
38. -2 of canopy). To calculate the amount of acorns to add per grid, we used the mean DBH of masting oaks in 1994 to calculate average crown radius (in meters) {1.71 + 7.92 × DBH [C. D. Canham, A. C. Finzi, S. W. Pacala, D. H. Burbank, Can. J. For. Res. 24, 337 (1994)]} and the average number of oaks masting in 1994. Based on these calculations, we added an average of 270,525 (1172 kg) acorns [primarily red oak, locally collected and donated as well as purchased (Sheffield Seed Co., Locke, N Y)] to each of the three experimental grids. Acorns were added in three events between late October and late November 1995 and were evenly scattered in a 4.45-m average crown radius circle around the center of each of the inner 80 grid cells. The 20 nest boxes on each experimental grid received acorn supplements of 10 per box every month from midSeptember to December 1995; then 5 per box every 2 weeks to early June 1996.
39. 2 white corduroy cloth, supported by a wooden dowel at one end and lead weights at the other end [R. C. Falco and D. Fish, Exp. Appl. Acarol. 14, 165 (1992)] along 15-by-30-m transect segments per grid, with transects randomly selected at each sample time. Every 30 m, ticks were removed and preserved in 70% ethanol for identification. Mean larval tick density (ticks per square meter), averaged over transect segments with a segment-to-segment within-grid variance estimate, were In-transformed (grid effect ANOVA, t test). The numbers of larval ticks attached to live-trapped mice were counted at each trap session. Mean larval ticks per mouse, averaged over each grid, with differences among mice giving the within-grid variance estimate, were In-transformed (grid effect ANOVA, t test). Data for August 1996 are reported because this is the peak period of host-seeking activity by larval ticks, as determined by 7 years of monitoring data (30, 31).
40. E. A. Desy and G. O. Batzli, Ecology 70, 411 (1989).
41. The abundance of infected host-seeking nymphal ticks is the primary determinant of Lyme disease risk (15). Because acorn addition caused higher densities of host-seeking larval ticks, higher densities of ticks on spirochete-infected mice, and higher densities of mice, Lyme disease risk is expected to be substantially increased 2 years after masting. This inference was supported by our observation that the average density of nymphs in June 1997 was 73.9% higher and in July was 31.3% higher on experimental (acorn-supplemented) grids than on control grids. However, because of high variability among sites, a drought causing small sample sizes of nymphs, and possible density-dependent emigration by mice from our experimental grids during the 1996 period of larval infestation, neither of these differences was statistically significant (paired t tests, P = 0.21 and 0.15 for June and July, respectively).
42. R. W. Campbell, For. Sci. Monogr. 15, 1 (1967); C. C. Doane, Invertebr. Pathol. 15, 21 (1970); J. R. Gould, J. S. Elkinton, W. E. Wallner, J. Anim. Ecol. 59, 213 (1990); A. E. Hajek et al., Proc. Natl. Acad. Sci. U.S.A. 87, 6979 (1990).
43. R. W. Campbell, For. Sci. Monogr. 15, 1 (1967); C. C. Doane, Invertebr. Pathol. 15, 21 (1970); J. R. Gould, J. S. Elkinton, W. E. Wallner, J. Anim. Ecol. 59, 213 (1990); A. E. Hajek et al., Proc. Natl. Acad. Sci. U.S.A. 87, 6979 (1990).
44. R. W. Campbell, For. Sci. Monogr. 15, 1 (1967); C. C. Doane, Invertebr. Pathol. 15, 21 (1970); J. R. Gould, J. S. Elkinton, W. E. Wallner, J. Anim. Ecol. 59, 213 (1990); A. E. Hajek et al., Proc. Natl. Acad. Sci. U.S.A. 87, 6979 (1990).
45. R. W. Campbell, For. Sci. Monogr. 15, 1 (1967); C. C. Doane, Invertebr. Pathol. 15, 21 (1970); J. R. Gould, J. S. Elkinton, W. E. Wallner, J. Anim. Ecol. 59, 213 (1990); A. E. Hajek et al., Proc. Natl. Acad. Sci. U.S.A. 87, 6979 (1990).
46. S. L. Pimm, The Balance of Nature? Ecological Issues in the Conservation of Species and Communities (Univ. of Chicago Press, Chicago, IL, 1991); J. T. Wootton, Annu. Rev. Ecol. Syst. 25, 443 (1994); R. D. Holt and J. H. Lawton, ibid., p. 495.
47. S. L. Pimm, The Balance of Nature? Ecological Issues in the Conservation of Species and Communities (Univ. of Chicago Press, Chicago, IL, 1991); J. T. Wootton, Annu. Rev. Ecol. Syst. 25, 443 (1994); R. D. Holt and J. H. Lawton, ibid., p. 495.
48. S. L. Pimm, The Balance of Nature? Ecological Issues in the Conservation of Species and Communities (Univ. of Chicago Press, Chicago, IL, 1991); J. T. Wootton, Annu. Rev. Ecol. Syst. 25, 443 (1994); R. D. Holt and J. H. Lawton, ibid., p. 495.
49. We thank G. Bernon for pupae; S. Clarke for transporting acorns; many local residents for acorn donations; R. Manz and local scout troops for help in distributing acorns; C. Canham, G. Lovett, and H. Smith for technical advice; C. Becker, D. Bulkeley, L. Christensen, R. Collett, K. Eisenhart, J. Evans, D. Fargione, J. Hart, K. Hazler, L. Hebdon, R. Krusic, E. Merwin, J. Planke, K. Scholl, C. Slagle, C. Stanis, and J. Vega for technical assistance; and F. Keesing for critical comment. Financial support was providec by NSF, NIH, the Plymouth Hill Foundation, the General Reinsurance Corporation, and the Mary Flagler Gary Charitable Trust.