International Journal for Parasitology 30 (2000) 1481±1485 www.parasitology-online.com Research note The role of oxygen availability in the embryonation of gallinarum eggs

L.M. Saunders*, D.M. Tompkins, P.J. Hudson

Institute of Biological Sciences, University of Stirling, Stirling FK9 4LA, UK Received 23 August 2000; received in revised form 13 September 2000; accepted 13 September 2000

Abstract The importance of oxygen availability in the embryonation of the infective egg stages of the gastrointestinal parasite was studied in the laboratory. Unembryonated H. gallinarum eggs were kept under either aerobic conditions by gassing with oxygen, or anaerobic conditions by gassing with the inert gas nitrogen, under a range of constant temperatures. Oxygenated eggs embryo- nated at a rate in¯uenced by temperature. Conversely, eggs treated with nitrogen showed no embryonation although when these eggs were transferred from nitrogen to oxygen gas after 60 days of treatment, embryonation occurred. This demonstrated that oxygen is an essential requirement for H. gallinarum egg development, although undeveloped eggs remain viable, even after 60 days in low oxygen conditions. The effects of climate on the of free-living stages studied under constant laboratory conditions cannot be applied directly to the ®eld where climatic factors exhibit daily cycles. The effect of ¯uctuating temperature on development was investigated by including an additional temperature group in which H. gallinarum eggs were kept under daily temperature cycles between 12 and 228C. Cycles caused eggs to develop signi®cantly earlier than those in the constant mean cycle temperature, 178C, but signi®cantly slower than those in constant 228C suggesting that daily temperature cycles had an accelerating effect on H. gallinarum egg embryonation but did not accelerate to the higher temperature. These results suggest that daily ¯uctuations in temperature in¯uence development of the free-living stages and so development cannot be accurately predicted on the basis of constant temperature culture. q 2000 Australian Society for Parasitology Inc. Published by Elsevier Science Ltd. All rights reserved.

Keywords: Parasite egg development; Heterakis gallinarum; Oxygen and development; Temperature and development

Climatic conditions are known to have a major impact on worms within the gut, pass out in the faeces and the development and transmission of nematode parasites of undergo development within the faecal pats on the ground terrestrial hosts and consequently play a signi®cant role in before becoming infective. Once outside the host the free- their epidemiology [1]. Many studies have investigated the living eggs may be exposed to anaerobic conditions, either roles of temperature and relative humidity on the develop- in faecal pats [7], or in surface water of poorly drained soils ment and survival of the free-living infective stages of where oxygen levels may be depleted by bacterial respira- nematode parasites although the majority have undertaken tion [8]. The roles of anaerobic conditions and variations in these trials in ideal conditions [1±3]. Non-motile infective temperature on the development rate of parasite eggs have stages in which the infective larva remains in a resistant egg been little studied. stage are frequently delivered into water or soil conditions Boisvenue [9] studied the effect of aeration on Ascaris where oxygen is limited and even though temperature suum eggs and demonstrated that eggs developed faster and conditions are ideal development may be prevented. had enhanced infectivity when exposed to high aeration Transmission of Heterakis gallinarum (Schrank, 1788) to rate, although he did not test whether eggs were able to its avian host, domestic or in the wild the ring- develop in the absence of oxygen. Low oxygen concentra- necked Phasianus colchicus, is via ingestion of tions may hinder or even stop development of the resistant infective eggs from the soil or other ' faeces when infective egg stages of such as A. suum or H. feeding [4±6]. Unembryonated eggs, produced by adult gallinarum, although development may recommence when oxygen levels are suitable. Alternatively, these infective stages may have evolved an adaptation to continue devel- * Corresponding author. Tel.: 144-1786-467832; fax: 144-1786- opment at a slower rate. To test between these hypotheses a 464994. laboratory study was undertaken in which the development E-mail address: [email protected] (L.M. Saunders).

0020-7519/00/$20.00 q 2000 Australian Society for Parasitology Inc. Published by Elsevier Science Ltd. All rights reserved. PII: S0020-7519(00)00139-9 1482 L.M. Saunders et al. / International Journal for Parasitology 30 (2000) 1481±1485 of H. gallinarum eggs in both aerobic and anaerobic condi- sampling did not in¯uence development, a further ®ve tions was monitored, under a range of both constant and tubes in each treatment group were put aside on day 60 diurnally cycling temperature regimes. and left undisturbed except for the usual gassing procedure. Unembryonated H. gallinarum eggs were obtained from When embryonation had reached 50% in the re-sampled fresh caecal faeces collected from infected captive phea- tubes the eggs in the corresponding undisturbed tubes sants and kept at 48C until required (,4 days). The caecal were examined to check for a difference in development faeces were mixed thoroughly and the number of eggs per between the two sets. gram estimated using the modi®ed McMaster technique A Gompertz growth model [12] was ®tted to the percen- [10]. Distilled water was added to make a suspension tage H. gallinarum egg embryonation data for each containing approximately 200 eggs/ml. temperature/gas treatment group A total of 480 replicate tubes were set up, each tube W ˆ Aexp 2bexp 2kt†† containing 1 ml H. gallinarum egg suspension (approxi- mately 200 eggs) and 2 ml distilled water, and sealed with where W is the percentage embryonation at time t, A is the a rubber bung. The replicates were divided into four groups asymptotic percentage embryonation, b is the zero of time of 120 tubes, which were kept in each of four temperature and k is the growth-rate constant. groups for the duration of the experiment. The ®rst three The Gompertz equation was ®tted using the constrained temperature groups consisted of eggs at constant 12, 17 and non-linear regression program in SPSSe and the parameters 228C. Eggs in the fourth temperature group were kept in an b and k were estimated. Asymptotic percentage embryona- environmental chamber (Sanyo Gallenkamp) programmed tion, A, was constrained to 100% in all models since this is with daily cycles of temperature. A daily cycle comprised 6 the maximum possible value for egg development. Fitted hat128C, 6 h ramping between 12 and 228C, 6 h at 228C and models were used to explore differences among treatment 6 h ramping between 22 and 128C. groups with respect to time, as described by the growth rate Commencing on day 5 of the experiment, 60 replicate parameter k, and the time required for the onset of develop- tubes were treated with oxygen, to aerate the medium, and ment. Model parameters were bootstrapped by sampling 60 with nitrogen, to create controls that received similar with replacement for 1000 iterations. The resulting boot- agitation without aeration, every 2 days. After being opened, strap estimates were used to calculate the mean and 95% each tube was gassed by slow bubbling of the appropriate gas con®dence interval ranges for both k and the number of days into the suspension through a 1 ml disposable pipette required for the onset of embryonation, where the number of connected to the gas cylinder. When the bubbles reached days was obtained by solving the Gompertz equation for t the top of the tube the pipette was rapidly withdrawn, the when W ˆ 0:01%  tube sealed and returned to the appropriate incubator [11]. 1 1 W Four control tubes containing only water, of which two were t ˆ 2 ln 2 £ ln 0:01% k b A gassed with oxygen and two with nitrogen, were included in each temperature group. On each sampling date the oxygen To test for differences between each temperature/gas content of the water was measured with a probe and dissolved treatment group in development rate k and the number of oxygen meter (YSI model 57, Clandon). days required for the onset of embryonation, bootstrap esti- The development of H. gallinarum eggs was recorded mates were compared among models. Models were consid- after 10, 20 and every 5 days thereafter until day 60. Five ered signi®cantly different from one-another at the 5% level replicate tubes were removed from each of the eight if the 95% con®dence interval range of the differences temperature/gas treatment groups and eggs were examined between bootstrap estimates did not span zero. P values in a Petri dish under a binocular microscope ( £ 40). There were estimated as the point at which the distribution of are several stages in egg development: single-celled, bootstrap differences exceeded zero. Differences in oxygen segmentation, morula, tadpole, vermiform and infective concentration measurements for the aerobic and anaerobic larva stages [4] but only eggs in the vermiform and infective groups, and the effects of re-sampling on egg development, larva stages were classed as embryonated. The numbers of were analysed non-parametrically using Mann±Whitney U- unembryonated (single-celled to tadpole stages) and tests. embryonated (vermiform and infective larva stages) eggs The re-sampling of tubes from day 60 onwards had no were counted until a total of 50 eggs had been recorded effect on H. gallinarum embryonation: there was no signi®- per sample. Samples were then discarded. On day 60, cant difference between percentage embryonation of eggs in tubes in the nitrogen groups were transferred to oxygen the re-sampled groups and eggs in the corresponding undis- gassing for the remainder of the experiment to check turbed samples (Mann±Whitney U ˆ 1609:5, P ˆ 0:923). whether the unembryonated eggs were still viable. After Gassing with oxygen produced a signi®cantly higher 60 days, insuf®cient replicates remained so egg counts oxygen concentration in water than gassing with nitrogen: were performed on a set of ®ve marked tubes in each treat- mean oxygen concentration (^SD) for `aerobic' tubes of ment group which were replaced and re-sampled every 5 water was 19.09 ^ 2.13 ppm, compared with 4.79 ^ 1.43 days until the end of the experiment. To check that re- ppm for `anaerobic' tubes (Mann±Whitney U ˆ 9640, L.M. Saunders et al. / International Journal for Parasitology 30 (2000) 1481±1485 1483 P , 0:001). Gassing treatment signi®cantly in¯uenced egg (0.415, 0.429) in aerobic 178C, P , 0.001; 0.183 (0.178, development. Eggs gassed with oxygen embryonated, 0.191) in anaerobic 228C vs. 0.489 (0.477, 0.503) in aerobic whereas those treated with nitrogen did not embryonate 228C, P , 0.001; 0.178 (0.175, 0.183) in anaerobic daily until transferred to oxygen (Fig. 1). cycles vs. 0.498 (0.483, 0.514) in aerobic daily cycles, Once transferred to oxygen, unembryonated eggs gassed P , 0.001). previously with nitrogen required signi®cantly fewer days Temperature in¯uenced both the time to onset and the for the onset of embryonation than eggs that had been in rate of H. gallinarum egg embryonation. There was a aerobic conditions from the beginning of the experiment decrease in time to embryonation with increasing constant (Fig. 1; 6.33 (4.00, 7.34) (mean ^ 95% C.I.) in anaerobic temperature (Fig. 1; 56.24 (46.41, 64.00) (mean ^ 95% 178C vs. 30.93 (30.27, 31.74) in aerobic 178C, P , 0.001; C.I.) for 128C oxygen, 30.93 (30.27, 31.74) for 178C 1.96 (0.87, 2.22) in anaerobic 228C vs. 23.41 (22.89, 23.85) oxygen, 23.41 (22.89, 23.85) for 228C oxygen, P , in aerobic 228C, P ,0.001; 2.49 (2.25, 4.57) in anaerobic 0.001; 6.33 (4.00, 7.34) for 178C nitrogen then oxygen, daily cycles vs. 24.94 (24.29, 25.81) in aerobic daily cycles, 1.96 (0.87, 2.22) for 228C nitrogen then oxygen, P ˆ P , 0.001). However, once transferred to oxygen, eggs 0.06), and an increase in rate parameter k, with increasing previously gassed with nitrogen developed at a lower rate constant temperature (Fig. 1A; 0.069 (0.054, 0.084) (k) compared with those that had been in aerobic conditions (mean ^ 95% C.I.) for 128C oxygen, 0.412 (0.415, 0.429) from the beginning of the experiment (Fig. 1; 0.183 (0.180, for 178C oxygen, 0.489 (0.477, 0.503) for 228C oxygen, 0.188) (mean ^ 95% C.I.) in anaerobic 178C vs. 0.421 P , 0.001). Daily cycles of temperature, ¯uctuating between 12 and 228C, accelerated both the onset and the subsequent varia- tion in egg development. Under both gassing treatments, eggs exposed to daily cycles began development earlier than eggs at constant 178C in the same gassing treatment (Fig. 1; 24.94 (24.29, 25.81) (mean ^ 95% C.I.) for daily cycles in oxygen vs. 30.93 (30.27, 31.74) for 178C oxygen, P , 0.001; 2.49 (2.25, 4.57) for daily cycles in nitrogen then oxygen vs. 6.33 (4.00, 7.34) for 178C nitrogen then oxygen, P ˆ 0.012). Those in daily cycles treated with oxygen developed later than eggs at constant 228C in oxygen (Fig. 1A; 24.94 (24.29, 25.81) for daily cycles in oxygen vs. 23.41 (22.89, 23.85) for 228C oxygen, P ˆ 0.010), whereas those in daily cycles gassed with nitrogen then oxygen began development at the same time as eggs at constant 228C treated with nitrogen then oxygen (Fig. 1B; 2.49 (2.25, 4.57) for daily cycles in nitrogen then oxygen vs. 1.96 (0.87, 2.22) for 228C nitrogen then oxygen, P ˆ 0.140). Furthermore, eggs exposed to cycles in aerobic conditions developed at a higher rate than eggs at constant 178C in aerobic conditions (Fig. 1A; 0.498 (0.483, 0.514) for daily cycles in oxygen vs. 0.421 (0.415, 0.429) for 178C oxygen, P ˆ 0.004), and at a similar rate to eggs at constant 228C in aerobic conditions (Fig. 1A; 0.498 (0.483, 0.514) for daily cycles in oxygen vs. 0.489 (0.477, 0.503) for 228C oxygen, P ˆ 0.422). However, when transferred to oxygen, eggs previously in anaerobic conditions developed at a rate that was independent of temperature (Fig. 1B; 0.183 (0.180, 0.188) for 178C vs. 0.183 (0.178, 0.191) for 228C, P ˆ 0.818; 0.183 (0.180, 0.188) for 178C vs. 0.178 (0.175, 0.183) for daily cycles, P ˆ 0.078; 0.183 (0.178, 0.191) for 228C vs. 0.178 (0.175, 0.183) for daily cycles, P ˆ 0.138). The availability of the free-living infective stages can Fig. 1. Embryonation of H. gallinarum eggs gassed with (A) oxygen and have a major in¯uence on the dynamics of macroparasite (B) nitrogen but transferred to oxygen on day 60 (solid line W, 12±228C systems [13,14]. Parasites that have time delays in their diurnal cycles; dotted and dashed line L, constant 128C; dotted line P, constant 178C; dashed line X, constant 228C). Plots were generated by development, either in terms of arrested development or ®tting mean parameter estimates into the Gompertz growth model for lengthy survival of the free-living infective stages, are each treatment group. more likely to generate oscillations in the abundance of 1484 L.M. Saunders et al. / International Journal for Parasitology 30 (2000) 1481±1485 both host and parasite populations [15]. For example, the [8] demonstrated the ovostatic action of bacteria on A. suum, free-living infective eggs of can remain Ascaridia galli and Aspiculuris tetraptera eggs, the devel- viable for long periods of time and such long lived species opment of which is dependent on the availability of oxygen make cyclic ¯uctuations in host populations more likely [23,24]. Actively developing bacteria in a liquid medium [1,16]. In addition, population models have illustrated acted ovostatically because bacterial respiration caused a how seasonal variations in the development and survival complete consumption of dissolved oxygen, thus preventing of Trichostrongylus tenuis infective larvae can have impor- eggs from starting their cleavage or continuing embryona- tant consequences for their transmission and persistence in tion. red host populations [16,17]. For species that achieve As demonstrated in previous studies with other nema- transmission via a resistant infective egg stage such as the todes [1,25±28], the rate at which H. gallinarum eggs gastrointestinal nematode H. gallinarum, infective stages embryonated increased as the constant temperature in are not mobile and so remain within the faeces in low- which they were maintained increased. However, investi- oxygen conditions until the faeces degrade. In addition, gating the importance of environmental factors under eggs deposited on poorly drained soil in wet habitats may constant conditions is unrealistic since the results cannot be exposed to anaerobic conditions if trapped in water. be applied directly to the ®eld where climatic factors exhibit Hence, the prevention of development in low oxygen condi- daily cycles and stochastic variation. Those workers that tions could have important consequences for the availability have previously examined the effects of ¯uctuating tempera- of infective stages. ture generally found that the development of infective The research conducted here demonstrates that oxygen is larvae under variable temperature was no different to that indeed essential for the development of H. gallinarum infec- at the constant mean temperature [27,29,30]. In contrast, tive eggs, showing that while eggs are exposed to anaerobic this study has provided evidence to suggest that ¯uctuating conditions they are unable to undergo development and temperatures accelerate H. gallinarum infective egg devel- hence remain non-infective. However, our ®ndings also opment. Not only was the onset of development under daily suggest that eggs remain viable under anaerobic conditions cycles earlier but the subsequent rate of development was and are able to develop fully when exposed to suitable levels higher than that of eggs at the constant mean cycle tempera- of oxygen. Although the embryonation rate of such eggs is ture of 178C. Furthermore, these effects were observed for reduced, the length of time to the initiation of embryonation, eggs both treated with oxygen from the onset of the experi- once moved into aerobic conditions, was actually shortened. ment and maintained in anaerobic conditions prior to devel- This was presumably due to the fact that the nitrogen opment in oxygen. gassing treatment did not result in completely anaerobic One possible explanation for the observed temperature conditions ± there was still a low concentration of oxygen effects is that if the eggs are showing a short-term adaptive present. response to rapid changes in temperature [31], then under The most likely explanation for the observed lack of H. ¯uctuating temperatures the development rate of eggs gallinarum egg development in the absence of oxygen is moving from a lower to a higher temperature may, for a that oxygen is a physiological requirement for the process relatively short period of time, be greater than that charac- of embryonation of the infective larvae within the egg stage, teristic for the higher temperature. This phenomenon has as suggested by the increased rate of respiration that accom- been described for the hatching times of Ostertagia circum- panies cell division in other nematode eggs [18]. Indeed, at cincta eggs [26]. An alternative explanation is that as very low oxygen concentrations the conversion of lipid, the temperature increases the corresponding development main energy reserves of parasitic nematode infective stages, increases exponentially, so that the rate of development at to carbohydrate is prevented [19] and reduced lipid utilisa- the temperature above the mean has increased so much that tion has been reported for the phytoparasitic nematode it is not completely nulli®ed by the lower development rate Meloidogyne naasi in waterlogged soils [20]. These proper- at the temperature below the mean. The effect of ¯uctuating ties could have an important in¯uence on the geographical temperature will be investigated further in a series of labora- distribution of H. gallinarum in relation to wet and dry tory experiments. habitats. Infective eggs of the nematode parasites Ascaridia The effects of both oxygen availability and ¯uctuating compar and Heterakis tenuicauda of rock require temperature on the development of parasite infective stages relatively dry conditions for prolonged survival [21] and may have important consequences for host-parasite have been implicated in the cyclic ¯uctuations in rock dynamics and should thus be incorporated into disease partridge abundance associated with relatively dry habitat prediction models in order to improve their accuracy. [22]. The present study suggests that H. gallinarum egg Some epidemiological models for nematode parasites have development in regions with wet soils, due to either high assumed that mean temperature cultures can be used to rainfall or poor drainage, may be prevented if the eggs are accurately predict the development time of infective stages exposed to anaerobic conditions. The effects of low oxygen [32±34]. However, this assumption has been based on a levels in water may also be exacerbated by the presence of limited number of experimental studies and the research bacteria in waterlogged soil. Przyjalkowski and Klekowski conducted here suggests that, at least for species like H. L.M. Saunders et al. / International Journal for Parasitology 30 (2000) 1481±1485 1485 gallinarum with an infective egg strategy, development population interactions. II. Destabilising processes. J Anim Ecol times are signi®cantly underestimated by such methods. 1978;47:249±67. As far as we know oxygen dependent development rates [16] Dobson AP, Hudson PJ. Regulation and stability of a free-living host- parasite system: Trichostrongylus tenuisin red grouse. II. Population have yet to be incorporated into any model of nematode models. J Anim Ecol 1992;61:487±98. infection. [17] Dobson AP, Hudson PJ. Population biology of Trichostrongylus tenuis in the red grouse, Lagopus lagopus scoticus. In: Scott ME, Smith G, editors. Parasitic and infectious diseases: epidemiology Acknowledgements and ecology. San Diego, CA: Academic Press, 1994. [18] Passey RF, Fairbairn D. The respiration of Ascaris lumbricoideseggs. We would like to thank Florence Troup at Drumdruills Can J Biochem Physiol 1955;33:1033±46. estate, Bridge of Allan, Stirlingshire for supplying [19] Passey RF, Fairbairn D. The conversion of fat to carbohydrate during and Pauline Monteith for assistance in the laboratory. embryonation of Ascaris eggs. Can J Biochem Physiol 1957;35:511± 25. L.M.S. was supported by a NERC studentship and D.M.T. [20] Ogunfowora AO. Factors affecting emergence, survival and infectiv- by a NERC standard grant. ity of Meloidogyne naasi. Nematologia 1979;24:72±80. [21] Rizzoli A, Hudson PJ, Manfredi MT, Rosso F, Rosa R, Cattadori IM. Intensity of nematode infection in cyclic and non-cyclic rock References partridge (Alectoris graeca saxatilis) populations. Parasitologia 2000, in press. [1] Anderson RC. Nematode parasites of vertebrates: their development [22] Cattadori IM, Hudson PJ, Merler S, Rizzoli A. Synchrony, scale and and transmission. Wallingford: CAB International, 1992. temporal dynamics of rock partridge (Alectoris graeca saxatilis) [2] Rogers WP, Somerville RI. The infective stage of nematode parasites populations in the dolomites. J Anim Ecol 1999;68:540±9. and its signi®cance in . Adv Parasitol 1963;1:109±71. [23] Przyjalkowski Z, Jakoski BJ. Action of intestinal bacteria on Ascaris [3] Stromberg BE. Environmental factors in¯uencing transmission. Vet suum eggs. Bull Acad Pol Sci, Ser Sci Biol 1968;16:211±3. Parasitol 1997;72:247±64. [24] Przyjalkowski Z. Aspiculuris tetraptera Nitsch, 1821, in germfree and [4] Clapham PA. On the life history of Heterakis gallinoe. J Helminthol conventional mice. Germfree Research. New York: Academic Press, 1933;11:67±86. 1973. pp. 447±9. [5] Lund EE, Chute AM. The reproductive potential of Heterakis galli- [25] Lee DL, Atkinson HJ. Physiology of nematodes. London: Macmillan narum in various species of galliform birds: implications for survival Press, 1976. of H. gallinarum and Histomonas meleagridisto recent times. Int J [26] Young RR, Anderson N, Overend D, Tweedie RL, Malafant KJ, Parasitol 1974;4:455±61. Preston GAN. The effect of temperature on times to hatching of [6] Draycott RAH, Parish DMB, Woodburn MIA, Carroll JP. Spring eggs of the nematode Ostertagia circumcincta. Parasitol survey of the parasite Heterakis gallinarum in wild-living pheasants 1980;81:477±91. in Britain. Vet Record 2000, in press. [27] Gibson ME. The effect of constant and changing temperatures on the [7] Holter P. Concentration of oxygen, carbon dioxide and methane in the development rates of eggs and larvae of Ostertagia ostertagi. J Therm air within dung pats. Pedobiol 1991(35):381±6. Biol 1981;6:389±94. [8] Przyjalkowski Z, Klekowski R. Investigations on the mechanism of [28] Shaw JL, Moss R, Pike AW. Development and survival of the free- ovostatic action of intestinal bacteria on the development of eggs of living stages of Trichostrongylus tenuis, a caecal parasite of red certain intestinal nematodes. Acta Parasitol Pol 1974;22:451±7. grouse Lagopus lagopus scoticus. Parasitol 1989;99:105±13. [9] Boisvenue RJ. Effects of aeration and temperature in in vitro and in [29] Hsu C, Levine ND. Degree-day concept in development of infective vivo studies on developing and infective eggs of Ascaris suum.J larvae of Haemonchus contortus and Trichostrongylus colubriformi- Helminthol Soc Washington 1990;57:51±56. sunder constant and cyclic conditions. Am J Vet Res 1977;38:1115±9. [10] MAFF. Manual of veterinary investigation laboratory techniques, Part [30] Salih NE, Grainger JNR. The effect of constant and changing 7, Parasitology. Reference Book No. 368. Ministry of Agriculture, temperatures on the development of the eggs and larvae of Ostertagia Fisheries and Food. Her Majesty's Stationery Of®ce, 1978. circumcincta. J Therm Biol 1982;7:35±38. [11] Jorgensen WK, Waldron SJ, McGrath J, Roman RJ, de Vos A, [31] Grainger JNR. First stages in the adaptation of poikilotherms to Williams KE. Growth of Babesia bigemina parasites in suspension temperature change. In: Prosser CL, editor. Physiological adaptation. cultures for vaccine production. Parasitol Res 1992;78:423±6. Washington DC: American Physiological Society, 1958. pp. 79±91. [12] Richards FJ. A ¯exible growth function for empirical use. J Exp Bot [32] Gettinby G, Paton G. The role of temperature and other factors in 1959;10:290±300. predicting the pattern of bovine Ostertagia spp. infective larvae on [13] Anderson RM, May RM. Infectious diseases of humans: dynamics pasture. J Them Biol 1981;6:395±402. and control. Oxford: Oxford University Press, 1991. [33] Grenfell BT, Smith G. Population biology and control of ostertagiasis [14] Roberts MG, Smith G, Grenfell BT. Mathematical models for macro- in ®rst year grazing calves. Proc Soc Vet Epidemiol Prevent Med parasites of wildlife. In: Grenfell BT, Dobson AP, editors. Ecology of 1983:54±61. infectious diseases in natural populations. Cambridge: Cambridge [34] Smith G. The population biology of the free-living phase of University Press, 1995. Haemonchus contortus. Parasitology 1990;101:309±16. [15] May RM, Anderson RM. Regulation and stability of host-parasite