AN ABSTRACT OF THE THESIS OF
Claire A. Fuller for the degree of Doctor of Philosophy presented on October 25, 1994. Title: Host Parasite Relationships Between Deer Mice (Peromyscus maniculatus) and
Their Eimerian Parasites (Protozoa). Redacted for Privacy Abstract approved: Andrew R. Blaustein
Hosts and their parasites can affect each other's population dynamics in numerous ways. Parasites may affect host reproduction or survival, and they may alter host behavior in such a way as to increase transmission from one host to the next. Hosts can affect parasite population dynamics if they develop immunity to their parasites or avoid contracting infections through behavioral mechanisms.
Here, I examine the population dynamics of the parasites, Eimeria arizonensis and E. delicata and their deer mouse host, Peromyscus maniculatus.
Deer mice shed up to 106 E. delicata oocysts over the course of a 9 day infection. They rapidly developed immunity to E. delicata under laboratory conditions. In addition, free-living adult deer mice were infected less frequently than free-living juveniles. This pattern suggests that deer mice also developed immunity to E. delicata in the field. The prevalence of E. delicata was not correlated with temperature or rainfall during its free-living stage.
Deer mice shed up to 107 E. arizonensis oocysts over the course of a 10-12 day infection. They developed immunity to E. arizonensis under laboratory conditions, although not as rapidly as to E. delicata. However, several lines of evidence suggest that acquired immunity to E. arizonensis was more variable in the field: 1) the prevalence did not decrease with deer mouse age, 2) animals frequently became reinfected and the intensity of infections (number of oocysts shed) decreased with subsequent infections only in some years and 3) acquired immunity to experimental infections in enclosures and the field was more variable than under laboratory conditions. The prevalence of E. arizonensis was strongly correlated with temperature during its free-living stage, suggesting that this factor may be more important than immunity in its population dynamics.
Eimeria arizonensis affected both deer mouse recruitment
(females) and over-winter survival (males). Eimeria arizonensis and E.delicata population dynamics were primarily affected by different factors, environmental temperature and immunity, respectively. Because deer mice rapidly developed immunity to E. delicata, adults were probably able to escape adverse effects. In contrast, many animals did not develop immunity to E. arizonensis and it negatively affected their survival. Host Parasite Relationships Between Deer Mice (Peromyscus maniculatus) and Their Eimerian Parasites (Protozoa)
by
Claire A. Fuller
A THESIS
submitted to
Oregon State University
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
Completed October 25, 1994 Commencement June 1995 Doctor of Philosophy thesis of Claire A. Fuller presented on October 25, 1994
APPROVED: Redacted for Privacy
MaDor / Professor, representing Zoology-
Redacted for Privacy
Chair of Department 6f Zoology
Redacted for Privacy
Dean of Gr
I understand that my thesis will become part of the permanent collection of Oregon State University libraries. My signature below authorizes release of my thesis to any reader upon request.
Redacted for Privacy Claire A. Fuller, Author ACKNOWLEDGEMENTS
I am greatly indebted to my graduate committee, Andrew
R. Blaustein (chair), who was supportive throughout this venture, Robert G. Anthony, Mark A. Hixon, Philippe A.
Rossignol, David McIntire (graduate representative, retired) and Karen Timm (grad. rep.) for many helpful discussions and for reading many drafts of my manuscripts. Sarah Fryer served as an unofficial committee member and had all of my ideas bounced off her.
This work could not have been completed without the contribution of many people. First and foremost are Dan Edge and Jerry Wolff. They not only allowed me to use their enclosures at a critical juncture in my research, they also were a source of field assistants and equipment. In addition, I had numerous undergraduate assistants without whom the work would not have been possible, or nearly as fun. Two of these became co-authors on chapter 2. Two people assisted with species identification: Martin Adamson and Don
Duszynski. Don also helped me throughout this project by reading manuscripts and allowing me to work in his laboratory. The work was funded by the OSU Zoology Dept.,
OSU IACUC, Sigma Xi and the American Society of Mammalogists. Friday Harbor Laboratories provided space and peace of mind for writing. My friends at OSU made my time there great fun and helped in many aspects of the research. I'd especially like to thank Dave
Booth, Sarah Fryer, Maureen Larson, Dede Olson, Diane Rowe, Paul
Sikkel, Brian Tissot, Liz Walsh, Sedonia Washington, Jerry Wolff and Ulrike Zelck. The Blaustein lab was a great place for ideas to grow: discussions with Andy Blaustein, Grant Hokit, Joe
Kiesicker, Hugh Lefcort, Rick O'Hara, Dede Olson, Jeff Peterson and Susan Walls were always interesting.
Finally, I would like to thank my family for always supporting me in my academic endeavors. Special thanks go to
Jeff, Joanne and Richard Fuller, Nancy Hayward, John Nelson, and my husband, friend and colleague Paul Sikkel. My son Niko has been the joy in my life for the past 17 months and has renewed my belief that learning keeps the child in us alive. CONTRIBUTION OF AUTHORS
Jerry Hefner and Elizabeth Wrosch were involved in the design, sample collection and writing of chapter 2. TABLE OF CONTENTS
Page
CHAPTER 1: GENERAL INTRODUCTION 1
Abstract 2
Introduction 2
Theoretical Studies 4
Empirical Studies (Level of the Individual Host) 7
Reproductive Success 8 Survival 9 Other Measures of Mortality 9 Transmission 15 Immunity 16
Population Level Empirical Studies 18
Conclusions 20
Overview of Thesis 22
Study Sites 23 Study Animal 25
CHAPTER 2: PARASITE LIFE HISTORIES 27
Abstract 28
Introduction 28
Methods 30
Results 33
Daily Oocyst Output 33 Diurnal Periodicity in Oocyst Output 35 Fecal Oocyst Concentration 35 Effect of Inoculurn Period and Level 39 Immunity 39
Discussion 43
Daily Oocyst Output 43 TABLE OF CONTENTS (Continued)
Page
Diurnal Periodicity in Oocyst Output and Fecal Oocyst Concentration 45 Immunity 47
CHAPTER 3: POPULATION DYNAMICS OF E. ARIZONENSIS and E. DELICATA 49
Abstract 50
Introduction 50
Methods 53
Data Collection 53
Statistical Analysis 55
Host Factors 55 Environmental Factors 56
Results 57
Overall Prevalence and Intensity 57 Hypothesis 1: Host. Factors 59
Prediction 1: Occurrence of Double Infections 59 Prediction 2a: Variability Among Sub-Populations of Deer Mice, Age and Sex Differences 59 Prediction 2b: Differences Among Reproductive Classes 64
Hypothesis 2: Environmental Factors 64
Discussion 67
Factors Influencing E. arizonensis 67 Factors Influencing E delicata 68
CHAPTER 4: THE EFFECT OF E. ARIZONENSIS ON DEER MOUSE SURVIVAL 72
Abstract 73
Introduction 73 TABLE OF CONTENTS (Continued)
Page
Methods 76
Observations 76 Experiments 77
Eimeria arizonensis Cultures 77 Field Experiment 78 Enclosure Experiment 80
Statistical Analysis 80
Results 82
Prediction 1: Establishment of Residency 82 Prediction 2: Overwinter Survival 82
Field Observations 82 Enclosure Experiment 84
Prediction 3: Weight Changes 87
Field Observations 87 Field Experiment 87 Enclosure Experiments 87
Discussion 90
CHAPTER 5: VARIABLE LEVELS OF IMMUNITY TO E. ARIZONENSIS 95
Abstract 96
Introduction 97
Methods: Field Observations 101
Results: Field Observations 102
Prediction 1, Complete Immunity 102 Prediction 2, Partial Immunity 104
Methods: Immunity Experiments 104 TABLE OF CONTENTS (Continued)
page
Laboratory Experiment 106 Field Experiment 106 Enclosure Experiment 108 Statistical Analyses 109
Results: Immunity Experiments 111
Effectiveness of inoculations 111 Laboratory Experiment 111 Comparisons Between Laboratory, Field and Enclosure Results 114
Discussion 114
BIBLIOGRAPHY 121 LIST OF FIGURES
Figure Page
1.1 Parameters Affecting Subpopulations of Deer Mice 5
2.1 Oocyst Output of E. arizonensis and E. delicata 34
2.2 Percent Daily Oocyst. Output 36
2.3 Total Oocyst Output Over a 20 Day Infection 41
3.1 The Prevalence of E. arizonensis and E. delicata 58
3.2 The Intensity of E. arizonensis and E. delicata 60
4.1 Over-Winter survival of Free-Living Deer Mice 85
4.2 Over-Winter Survival of Enclosure Deer Mice 86
4.3 Weight Change of Enclosure Deer Mice 89
5.1 Number of Times Individual, Free-Living Deer Mice Were Infected With Naturally Occurring Eimeria arizonensis Each Year 103
5.2 Change in Oocyst Output with Successive Infections.105
5.3 Experimental Infections in Naive Animals 112
5.4 Immunity to Experimental Infections in the Field 115 LIST OF TABLES
Table Page
1.1 Studies Addressing Parasite-Mediated Population Regulation in Vertebrates 10
1.2 Studies by Host and Parasite Taxa 21
2.1 Oocysts Per Fecal Pellet 37
2.2 ANOVA of the Number of Oocysts Excreted 40
2.3 ANOVA of the Reproductive Index 40
2.4 Immunity to E. arizonensis 42
2.5 A Comparison Between E. arizonensis and E. delicata Life History Parameters 48
3.1 Frequency of Double Infections 61
3.2 Percent infected with E. arizonensis in different sex and age classes of deer mice 62
3.3 Percent infected with E. delicata in different sex and age classes of deer mice 63
3.4 Percent infected with E. arizonensis in different reproductive classes of adult deer mice 65
3.5 Percent infected with E. delicata in different reproductive classes of adult deer mice 66
4.1 Protocol for Field Experiment 79
4.2 Predictions and Experimental Paradigms 81
4.3 E. arizonensis and Deer Mouse Recruitment 83
4.4 Weight Change in Free-Living Deer Mice 88
5.1 Protocol for Laboratory and Field Experiments 107
5.2 Sample Sizes in Enclosure Experiments 110
5.3 Infections Among Sham Inoculated Animals 113 Host Parasite Relationship Between Deer Mice (Peromyscus maniculatus) and their Eimerian Parasites (Protozoa)
Chapter 1
General Introduction 2
Abstract
Parasites may regulate the population dynamics of their hosts. In this review, I summarize mathematical models addressing this topic and the parameters of the most basic model are outlined. I then review empirical studies which address model parameters at the level of the individual host. I next summarize studies which directly examine whether parasites regulate host populations (ie., at the population level). All empirical studies included in this review examine macroparasite (Protozoa, helminths and arthropods) vertebrate systems and were conducted under natural or semi-natural conditions. This review illuminates several areas for further study including 1) experimental studies at the population level, 2) experimental studies at the level of the individual host and 3) studies which include a greater spectrum of host and parasite taxa.
Introduction
The relative importance of factors regulating animal populations is one of the most controversial and widely discussed topics in ecology (Hestbeck 1987). Much of the controversy has focused on emigration, competition and predation. For example McMahon & Tash (1988) demonstrated that emigration maintains desert pupfish population density and Buskirk and Smith (1991) showed that intraspecific 3 competition was density dependent and negatively affected salamander populations. The potential for parasites to regulate their host populations has also been established through mathematical models (Anderson & May 1979, 1986, May & Anderson 1979, Aron
& May 1982). Moreover, many laboratory studies have documented negative effects of parasites on individual hosts and a few studies have shown that parasites can regulate host populations under controlled laboratory conditions (see reviews by Gregory & Keymer, 1989, Keymer & Read 1991, Scott
& Dobson 1989, Scott & Lewis 1987). However, many factors differ between the laboratory and the field including food levels, the presence of other parasites, predators and competitors. Thus, while it is clear that parasites can control host populations under laboratory conditions, it is unclear from models and laboratory studies how important parasites are in the presence of other factors such as food limitation and predation. Until recently, emprical studies of host-parasite relationships conducted under field conditions have largely been lacking. While a few long-term field experiments have been conducted at the population level (see below), the majority of recent studies have focused on specific aspects of host-parasite interactions such as parasite-induced mortality or host immune response. In this review, I will briefly outline models of animal population regulation by parasites. I will then summarize studies of parasites in 4 vertebrate hosts conducted under natural or semi-natural (ie., field enclosures) conditions which examine specific parameters of the models at the level of the individual host. Finally, I will discuss studies conducted at the population level under natural and semi-natural conditions.
Theoretical Studies
Until 1979, epidemiological models of host-parasite systems were static, examining one point in time (Anderson &
May 1979) . Anderson & May (1979) and May & Anderson (1979) modelled changes in host and parasite populations over time.
Figure 1.1 (modified from Anderson & May 1979) shows the parameters of the basic model (Anderson & May 1979). Animals susceptible to, infected with and immune to parasites are represented by X, Y and Z, respectively (a = intrinsic host birth rate and b = the intrinsic host death rate. Alpha = parasite induced death rate). Beta, v and y represent the rate at which hosts move between states X, Y and Z by transmission (B), recovery (v) and loss of immunity (y). There are several assumptions of the basic model which have been addressed in subsequent models. For example, X, Y and Z are assumed to have the same birth rate; Dobson (1988) incorporated parasite induced reduction in host reproductive success. Another assumption is that all susceptible hosts(X) are equally susceptible to parasitism; Anderson & May (1985) 5
.___._._.____...------T---- births (a) deaths (b + alpha) deaths (b)
(B)
X Y (susceptible) (infected)
(v)
(y) Z (immune)
deaths (b)
Figure 1.1. Parameters affecting subpopulations of deer mice. Parameters from the basic model of Anderson and May (1979) on parasite-mediated population regulation. B, v and y represent transmission, recovery and loss of immunity, respectively and are the rates at which animals move between subpopulations X, Y and Z. Arrows leading from the circle represent host deaths. Susceptible and immune hosts die at some intrinsic rate (b). Infected hosts have an additional, parasite-induced death rate (alpha). All host subpopulations are assumed to have the same intrinsic birth rate (a) and offspring are assumed to be susceptible (ie., no vertical transmission of parasites or immunity). 6 incorporated genetic heterogeneity in host susceptibility. Anderson & May (1979) also discussed two fundamentally different types of parasites. Microparasites (viruses & bacteria) reproduce within the host, recovered hosts become completely immune and immunity can last the host's lifetime.
Thus, the prevalence (% infected) of microparasites is often used in epidemiology studies (Aron & May 1982).
Macroparasites (helminths and arthropods) generally do not complete a reproductive cycle within a host, immunity is density dependent (ie., immunity increases with increasing exposure) and is generally lost at some rate (Anderson
1985). Thus, because macroparasite infections are not all or-nothing, the intensity of infection (# parasites per host) is thought to be a better measure in epidemiological studies (Anderson 1986). Protozoans reproduce like microparasites but generally elicit density-dependent immunity, thus, they do not fit neatly into either category. Although Anderson and May (1979) originally included protozoans as microparasites, subsequent work has generally treated them as macroparasites (Anderson 1991, Aron & May
1982, Nowell & Higgs 1985).
Several conclusions can be drawn from Anderson and
May's model: 1) there is a certain baseline host population density below which a parasite cannot persist, 2) this level decreases with increasing transmission and 3) increases with increasing pathogenicity. Finally, for a parasite to 7 regulate its host population, parasites must be overdispersed such that only a few hosts carry the majority of parasites, parasite induced mortality must increase with parasite burden and parasite survival/reproduction must be density dependent within individual hosts (Gregory & Keymer
1989) .
Empirical Studies (Level of Individual Host)
I have limited the scope of the studies reviewed to vertebrate-macroparasite systems, because 1) acquired immunity will play a role in these systems and 2) vertebrate-microparasite studies have been summarized elsewhere (Minchella & Scott 1991, Scott 1988). In addition, to avoid anecdotal accounts, I have only included observational studies which examine at least 25 hosts. Finally, a great deal of recent work has addressed the effect of parasites on host behavior, especially on sexual selection. For example, parasites affect behavior of the hosts such that transmission is facilitated or they weaken hosts such that sexual displays are affected. In addition, females may avoid mating with parasitized males. These studies have been reviewed recently (Combes 1991, Dobson
1988, Hart 1990, Moore 1987) and are not included here.
Examination of Anderson & Mays' models suggests that there are 5 basic parameters which could affect parasites' ability to determine host population size. The parameters 8
can be divided into pathogenicity factors (effects on reproduction or survival) and susceptibility factors (rate
of parasite transmission [B], recovery [v] and loss [y] of
immunity. Table 1.1 summarizes studies that address these parameters by host and parasite taxa.
Reproductive Success
Most studies have examined the effect of parasites on correlates of reproductive success such as pregnancy rate
(Boonstra et al. 1980, Theis & Schwab 1992), clutch size
(Clinchy & Barker 1994, Dobson et al. 1992, McPhail &
Peacock 1983, Poiani 1993, Schall 1983), testis size (Boonstra et al. 1980, Timm & Cook 1979), dominance (Schall
& Dearing 1987), or nestling weight (de Lope et al. 1993, de
Lope & Moller 1993, Moore & Bell 1983). A few studies have documented parasite effects on fledging success (Howe 1992,
Hudson 1986, de Lope et al. 1993, de Lope & Moller 1993,
Moller 1993). Four of the studies on reproductive success have included field manipulations of nest mite (de Lope et al. 1993, de Lope & Moller 1993, Moller 1993) or nematode
(Hudson 1986) levels and all 4 of these studies showed
significant negative effects of parasites on fledging success (all in birds). 9
Survival
Several studies have measured the effect of parasites on adult survival using recapture data (Boonstra 1977,
Boonstra et al. 1980, Gulland 1992, Hudson et al. 1992 a, b,
Keith et al. 1986, Lehmann 1992, Miller & Getz 1969, Moss et al. 1990, Samson et al. 1987, Weatherhead & Bennett 1992).
Other studies have used statistical methods comparing expected and observed distribution of parasites in host populations (Adjei et al. 1986, Dobson et al. 1992, Gordon &
Rau 1982; method sensu Croften 1971, Pacala & Dobson 1988).
Temple (1987) showed that a predator (red-tailed hawk) captured highly parasitized prey in prey species which were difficult to capture but not in species that were easy to capture (based on number of strikes to capture). Only 3 of the studies included in this survey manipulated parasite burdens to determine whether there is a causal relationship between parasites and host mortality (Lehmann 1992, Hudson et al. 1992b, Samson et al. 1987). All three studies showed that arthropod or nematode parasites decreased host survival.
Other Measures of Mortality
Several authors have examined mechanisms through which parasites might influence host survival or reproduction including: 1) weight change associated with parasitism
(Bennett et al. 1988, Booth et al. 1993, Boonstra et al. 10
Table 1.1. Studies addressing parasite-mediated population regulation in vertebrates (sensu Anderson & May 1979). Parasites taxa include Protozoa (P), Acanthocephala (Ac), Cestoda (C), Nematoda (N), Trematoda (T) and Arthropoda (Ar). 0 indicates observational studies and E indicates experimental studies. Under headings reproduction,survival and other, (-) indicates a negative effect of the parasite, 0 indicates no effect. Under transmission, parameters measured are indicated by B = behavioral, H = habitat differences, R = reinfestation time, S = seasonal variation, Sury = free-living parasite survival,T = temperature variation, V = vertical transmission. Under immunity, A-I and A-P = age-intensity or age-prevalence, spleen = spleen size, RE = changes in immunity with reproductive effort. Table 1.1 11
Host Parasite Repro Sury er Trans Immunity References Fish
Gasterosteus aculeatus Schistocephalus 0, (3-spine stickleback) solidus (C) McPhail & Peacock 1983 Culaea inconstans Apatemon qracilis 0, Gordon & Rau 1982 (brook stickleback) ( ) Saurida tumbil Callitetrarhynchus 0, Adjei et al. 1986 S. undosquamis qracilis (Ac?) 0,- (lizard fish)
Amphibians Scaphiopus couchii Pseudodiplorchis 0, T 0,Y A-I (spade-foot toad) americanus (T) Tinsley 1990, Tocque 1993, Tocque & Tinsley 1991 Reptiles Sceloporus occidentalis Plasmodium o, (western fence lizard) mexicanum (P) Schall 1983, Schall & Dearing 1987, Schall et al. 1983 Anolis lineatopus Centrorhynchus 0,N A-P/I (anole lizard) spinosus (Ac) Vogel & Bundy 1')87 Mesocoelium 0,N A-P/I danforthi (T) Cyrtosumum 0,N A-P/I sceloperi (N) Thelandros 0,N A-P/I cubensis (N) Anolis spp. various helminths o, o, 0, H Dobson et al. 1992
Birds Various passerines Hematozoa (P) Bennett et al. 1988 Oreosoptes montanus Protocalliphora 0, -/0 Howe 1992 (sage-thrasher) braueri (Ar) Ficedula hypoleuca Ceratophyllus E, B Mappes et al. 1994 (pied flycatcher) qallinae (Ar) Delichon urbica Oeciarus E, de Lope et al. 1993 (house martin) hirundinis (Ar) Sturnus vulqaris Plaqiorhynchus 0, Moore & Bell 1983 (starling) cylindraceus (Ac) Troglodytes aedon Protocalliphora Johnson et al. 1991 (house wren) pavarom (Ar) Table 1.1, cont. 12 Host Parasite Repro Sury Trans Immunity References Agelaius phoeniceus Hematozoa (P) (red-winged blackbird) 0,N A-P Weatherhead & Bennett 1991
Molothrus ater Hematozoa (P) 0,0 (brown-headed cowbirds) 0,0 Weatherhead & Bennett 1992 Parus major Hematozoa (P) (great tit) 0,Y RE Norris et al. 1994
Hirundo rustica Oeciacus E, (swallow) hirundinis (Ar) de Lope & Moller 1993, Moller Ornthonyssus bursa 1993 (Ar) Columba livia Ischnocera sp. (rock dove) (Ar) Booth et al. 1993 Anser caerulesceus various (N,T,P) 0,0 (lesser snowgoose) Clinchy & Barker 1994 Laqopus laqopus Trichostronqylus 0,0 (red grouse) tenius (N) 0,SAT 0,N A -P Moss et al 1990,1993, Shaw & Moss 1989, Wilson 1983, Hudson 0,-/E, 0,-/E, 1986, Hudson et al. 1992 a,b Mammals Microtus montanus Babesia microti 0,Y Spleen Watkins et al. 1991 (montane vole) (P) M. townsendii V.,31fartia vigil 0, (Townsend's vole) (Ar) Boonstra 1977 Cuterebra qrisea 0, 0, 0, (Ar) Boonstra et al. 1980 M. aqrestis Babesia (P) (short-tailed vole) 0,N A-P Healing 1981, Clethrionomys qlareolus Hepatozoa (P) Haukisalmi et al. 1988 (bank vole) 0,S C. q. skomerensis Trypanosoma (P) (island bank vole) 0,Y A -P C. rutilis various helminths (red vole) Apodemus sylvaticus (wood mouse) Eimeria spp. (P) 0,S 0,Y A-I Ball & Lewis 1984 Heliqmosomoides polyqyrus (N) Gregory 1992, Gregory et al. 1992, Quinnell 1992, Montgomery & Montgomery 1988 Peromyscus leucopus Cuterebra sp. (Ar) O, 0, (white-footed mouse) 0,Y Spleen Miller & Getz 1969, Munger & 0,0 Karasov 1989, Payne et al. 1965, Hymenolepis Timm & Cook 1979 citelli (C) Munger & Karasov 1994 Table 1.1, cont. 13
Host Parasite Repro Sury Ot Trans Immunity References P. maniculatus Taenia 0,0 0, S 0,N A-P (deer mouse) taeniaeformis (C) Theis & Schwab 1992 Capillaria hepatica (T) 0, S 0,N A-P Herman 1981 Ixodes dammini (Ar) 0, H Adler et al. 1992 ectoparasites E,R Glicken & Schwab 1980 Rattus norveqicus Toxoplasma gondii 0, V (brown rat) (P) Webster 1994 Siqmodon hispidus blood parasites (cotton rat) 0,varies Lochmiller et al. 1994 Gerbillus andersoni ectoparasites (Ar) E,- (gerbil) E,' Lehmann 1992 Spermophilus elegans Eimeria spp. (Wyoming ground squirrel) 0, H 0,N A-P Stanton et al. 1992 S. townsendii Eimeria spp. (Townsend's ground 0, S 0,N A-P Wilber et al. 1994 squirrel) Tamius striatus total parasite 0, var (eastern chipmunk) burden Temple 1987 Sylvilaqus floridanus (cottontail rabbit) Sciurus carolinensis (gray squirrel) Myocaster coypus Eimeria spp. (nutria) 0, S Ball & Lewis 1984 Lepus americanus Obeliscoides 0, 0, S (snowshoe hare) cuniculi (T) 0,N A-P Keith et al. 1985, 1986 Nematodirus 0,0 trianqularis (N) 0,Y A-P Trichuris leporis 0,0 0, (N) 0,Y A-P Protostronqvlus 0,0 0, boughtoni (N) 0, S 0,Y A-P Taenia pisiformis 0,0 (C) 0,Y A-P Eimeria spp. (P) 0,0 0 0,Y A-P Ovis aries Ostertraqia o, (Soay sheep) circumcincta (N) 0, S 0,Y A-P Gulland 1992, Gulland & Fox 1992 Trichostronqvlus E, Sury 0, S sp. (N) 0,Y A-P Cooperia carticii E,Sury 0, S 0,Y A -P (?) E,sury O. canadensis Protostronqvlus E, 0, S (bighorn sheep) stilesi (N) 0,Y A-I Arnett et al. 1993 Samson et al. P. sp. 1987 0,Y RE Festa-Bianchet 1989 Table 1.1, cont. 14
Host Parasite Repro Su Other Trans Immunity References Odocoileus virqinianus Fascioloides magna (white-tailed deer) (T) Eimeria mccordocki 0,0 Mulvey et al. 1994 (2) 0,Y A-P Ranqifer tarandus Hypoderma tarandi Samuel & Trainer 1971 (reindeer) (Ar) E,Y Sex Alopex laqopus Trichinella sp. Folstad et al. 1989 (arctic fox) (N) 0,0 0,Y A-P Homo sapiens Schistosoma Prestrud et al. 1993 (humans) mansoni (T) S. haemotobium (T) 0,Y A-I Fulford et al. 1992 0,Y A-P Woolhouse et al. 1991 15
1980, Keith et al 1986, Lehmann 1992, Mulvey et al. 1994,
Munger & Karasov 1994, Prestrud et al. 1993, Weatherhead &
Bennett 1992), 2) changes in metabolic rate (Booth et al.
1993, Munger & Karasov 1989, 1994), 3) changes in the relative abundance of plasma proteins (Payne et al. 1965),
4) anemia (Lehmann 1992, Schall et al. 1982, Tocque 1993) and 5) changes in conception date (Mulvey et al. 1994).
Three studies included in this survey were experimental
(Booth et al. 1993, Lehmann 1992, Munger & Karasov 1989) with variable results (see Table 1.1).
In general, one problem with these studies is that it is often unclear whether parasite induced changes are actually associated with mortality and to what degree
(Munger & Karasov 1989). Thus, unless some measure of survivorship is also included, the studies are difficult to interpret.
Transmission
Transmission of parasites from one host to another is especially difficult to measure in the field because one must know how many of the available parasites are actually encountered and subsequently recruit into the host. In some cases, vertical transmission (when a parasite passes to offspring through the placenta) may also occur (Webster
1994). Where the parasite's life-cycle includes free-living stages, it is also relevant to know the survival time of the 16 free-living parasite under various environmental conditions.
Most field studies which touch on the subject of transmission document the association between the prevalence or intensity of infections and temporal changes (Ball &
Lewis 1984, Gregory 1992, Haukisalmi et al. 1988, Moss et al. 1993), differences in habitats (eg., xeric vs. mesic; Adler et al. 1992, Dobson et al. 1992, Stanton et al. 1992) or weather variables (Moss et al. 1993, Tocque & Tinsley
1991). A few studies have examined survival of free-living stages (Levine 1980, Gulland & Fox 1992) and reinfestation after removal (Glicken & Schwab 1980). Hosts may also avoid
(Mappes et al. 1994) or be manipulated by parasites (Moore & Bell 1983). Behavioral mechanisms of parasite avoidance and parasite manipulation of hosts to increase transmission have been reviewed by Combes (1991), Dobson (1988), Hart (1990) and Moore (1987).
Immunity
Vertebrate hosts clearly become immune to a wide variety of parasites under laboratory conditions (Gregory, Keymer & Clarke 1990, Higgs & Nowell 1988, Lillehoj 1988, Wakelin
1978, 1987, Wassom et al. 1986). However, it is less clear how frequently immunity occurs under field conditions. Both genetic (Wakelin 1992) and environmental (McCallum 1990) factors may modify immunocompetence. Several methods have been used to detect immunity including serological measures 17
(Elliott et al. 1994, Lochmiller et al. 1994), increases in spleen size (Timm & Cook 1979, Watkins et al. 1991) and the presence of convex age-intensity/age-prevalence curves
(Arnett et al. 1993, Ball & Lewis 1984, Fulford et al. 1992,
Gulland & Fox 1992, Haukisalmi et al. 1988, Healing 1981,
Hermann 1981, Keith et al. 1986, Quinnell 1992, Prestrud et al. 1993, Samuel & Trainer 1971, Shaw & Moss 1989, Stanton et al. 1992, Tinsley 1990, Vogel & Bundy 1987, Watkins et al. 1991, Weatherhead & Bennett 1991, Wilber et al. 1994).
The first 2 methods only detect previous exposure or general immune readiness, rather than current immune response to a specific pathogen. The last method assumes that exposure increases with age after an initial period of low juvenile exposure. Thus, as animals first become exposed, parasite burden increases; as immunity develops, parasite burden decreases. This method is widely employed (used by 20 of 24 authors included here), but convex age-intensity curves can be caused by other factors (Bundy 1988, Fulford et. al 1992,
Pacala & Dobson 1988). Two studies of immunity included here were experimental. However, there were problems with both.
Folstad et al. (1989) examined differences in intensity among animals with different hormonal levels. Thus, behavioral differences among age or sex classes could explain differences in parasite burdens. Quinnell (1992) documented immunity in enclosure populations of wood mice
(Apodemus sylvaticus). However, except for natural 18 temperature and lighting, enclosures did not mimic natural environments. In addition, challenge infections (a second experimental infection given to the same animal designed to test the immune response to the first infection) were conducted in the laboratory. Thus, harsh environmental conditions which could compromise the immune response probably were not adequately imitated.
Population Level Empirical Studies
Three studies have examined host population regulation by macroparasites in vertebrate hosts at the population level. All examined mammal-nematode systems; 1 used comparative data and the other 2 were experiments conducted in field enclosures.
The nematode Parelaphostronqylus tenius causes neurological damage in moose but not in deer. It has been hypothesized that moose decline in the presence of deer and that this decline is caused by increased transmission of P. tenius from deer to moose with increasing deer population density (Whitlaw & Lankester 1994). Whitlaw and Lankester used reports of deer and moose densities and moose neurological disease on 6 sites over an 80-year period to test this hypothesis. They found that moose and deer densities were negatively correlated but there was no significant correlation between moose declines and reports of P. tenius related disease. They critique the quality of 19
their data set and conclude that "historical data are unlikely to provide further insight" in this system.
Barker et al. (1991) and Gregory (1991) used enclosures
to examine population regulation in the house mouse (Mus
musculus)-Capillaria hepatica system and the wood mouse
(Apodemus sylvaticus)-Heliqmosomoides polygyrus system,
respectively. Barker et al. (1991) found little evidence of
population regulation by C. hepatica; control (n=3) and
treatment (n=3) populations fluctuated synchronously and overall population densities, reproduction and survival did
not differ between control and treatment populations over the 1.5 year study. They suggested that food quality or
intraspecific competition may have regulated mouse population density and masked any effect of C. hepatica.
Gregory (1991) found that control populations (n=3) grew
faster and had significantly greater host densities at the end of the 2-year study than treatment (n=3) populations.
Reproduction was not affected, rather treatment animals did
not survive as well after weaning as control animals. Although these studies were closer to field conditions than
laboratory studies, in many ways the enclosures were closer to the laboratory than the field. Enclosures were small
(Barker et al. = 15 x 15 m; Gregory = 9 x 4 m), food and water were supplied ad libitum and predators and
interspecific competitors were excluded. While Gregory
(1991) demonstrated that H. polyqyrus can regulate wood mouse populations in the absence of other factors, he does 20 not provide information about the relative importance of these parasites in population regulation.
Conclusions
1) Future studies should include a wider variety of host and parasite taxa. The majority of field work has focused on avian and mammalian hosts (Table 1.2). Even within these groups, only a few host orders have been examined. The majority of studies in mammals have been conducted on rodents and ungulates; in birds most hosts have been passerines. The parasites examined have also had a relatively narrow focus in birds: most studies have been on nest mites and hematozoa (Protozoa). In mammals the scope of the parasites studied is substantially broader and includes protozoans, cestodes, trematodes, nematodes and arthropods. The taxonomic scope of experimental studies is particularly limited (Table 1.2).
2) More studies should address all individual parameters of population regulation but especially transmission and immunity. The aspects of population regulation that have been addressed include all model parameters, however, the majority of work has focused on parasite-induced host mortality. Other than documenting seasonal variation in parasites and age-intensity curves, very few studies have examined transmission dynamics or acquired immunity. 21
Table 1.2. Studies by host and parasite taxa. The number of observational (0) and experimental (E) studies addressing various model parameters in host and parasite species. Repro = parasite effects on host reproduction, Sury = parasite effects on host survival, Other = parasite effects on other aspects of host fitness (eg., weight loss), Trans = transmission and Immun = host immune response.
Repro Sury Other Trans Immun Taxonomic group 0 E 0 E 0 E 0 E 0 E
Host Taxa
Fish 1 0 3 0
Amphibians 1 0 1 0 1 0
Reptiles 2 0 1 0 1 0 1 0 1 0
Birds 4 3 3 1 3 1 2 1 2 0
Mammals 3 0 6 2 5 2 10 1 14 2
Parasite Taxa
Protozoa 2 0 2 0 4 0 6 0 8 0
Acanthocephala 1 0 1 0
Cestoda 2 0 1 0 1 1 1 0 2 0
Trematoda 1 0 2 0 3 1 2 0 5 0
Nematoda 2 1 5 3 4 0 6 0 11 1
Arthropoda 3 2 3 1 3 2 1 1 1 1 22
3) Field experiments are critical. Keymer and Read
(1991) called for field experiments to test the causal relationship between parasites and their hosts. Although most of the 52 studies reviewed here did not incorporate field manipulations, those that did were relatively recent
(10 of 12 were published after 1988, 8 after 1991). Thus, experimental evidence about the ability of parasites to regulate host populations is beginning to accumulate. Within the experimental studies, only 3 of 12 examined transmission or immunity. All others examined various aspects of mortality. 4) Long-term experiments at the population level are needed. Two of the 3 studies of parasite-mediated population regulation were inconclusive (Barker et al. 1991, Whitlaw &
Lankester 1994). The third provided evidence that parasites can regulate hosts but did not address the conditions under which this might occur (Gregory 1991). In addition, studies should follow several host generations to adequately assess the impact of parasites on population dynamics.
Overview of Thesis
I examined the host-parasite interactions of deer mice
(Peromyscus maniculatus) and its intestinal protozoans
(Eimeria arizonensis and E. delicata). My dissertation provides new information in several of the areas addressed above: 23
1) Most studies of protozoans have focused on blood parasites. Little is known about other groups in this
Kingdom.
2) I have examined several of the individual parameters
including parasite-induced host mortality, transmission
dynamics and acquired immunity. 3) The second chapter provides information on parasite life- history and was conducted in the laboratory. All other
chapters were conducted in the field.
4) Chapters 4 and 5 include field experiments on survival
and acquired immunity, respectively. These represent the first experimental field work on protozoan parasites. To my knowledge, the immunity experiments are the first to address
whether individual hosts acquire immunity under natural and
semi-natural conditions in any host or parasite taxa.
Study Sites
Observational and experimental studies on free-living animals were conducted in MacDonald Forest, Oregon State University's experimental forest (approx. 15 km NW,
Corvallis, OR). The elevation at the trapping sites ranged
from 500-1500 ft. All areas trapped were chosen to be
similar in the composition of dominant species. Dominant
trees included Douglas fir (Psuedotsuqa menziesii), grand
fir (Abies qrandis), big leaf maple (Acer macrophyllum) and
vine maple (Acer circinatum). Shrubs included Oregon grape 24
(Berberis nervosa) and salal (Gaultheria shallon); other common understory plants included poison oak (Rhus diversiloba), sword fern (Polystichum munitum) and moss.
Many other species of small mammals co-occurred with deer mice. Common species (captures at least 1 in 200 trap nights) included voles (Microtus oregoni and Clethrionomys californicus), Townsend's chipmunk (Tamias townsendii), flying squirrels (Glaucomys sabrinus) shrews (Sorex vaqrans and S. trowbridqii) and shrew-moles (Neurotrichus qibsii).
Species caught infrequently (< 1 in 1000 trap nights)
included red tree voles (Phenacomys lonqicaudus), bushy- tailed woodrats (Neotoma cinerea), Pacific jumping mice
(Zapus trinotatus), Douglas' squirrels (Tamiasciurus douqlasii), coast moles (Scapanus orarius) and ermine
(Mustela erminea). Deer mouse predators known to occur in this area include weasels (M. erminea and M. frenata), skunks
(Mephitis mephitis), raccoons (Procyon lotor), coyotes
(Canis latrans), bobcats (Felis rufus), garter snakes
(Thamnophis sertalis and T. eleqans), gopher snakes
(Pituophis melanoleucus) and spotted, screech, long-eared and great horned owls (Strix occidentalis, Glaucidium qroma, Otus asio, Asio otus and Bubo virqinianus, respectively).
The research site for the enclosure studies was located at the Hyslop Experimental Farm of Oregon State University, approx. 10 km north of Corvallis, OR. Twenty-four 0.2 ha
(0.5 acre) enclosures were constructed at the research site; 25
4 were used in my research. Each enclosure was 45 x 45 m and was constructed of galvanized sheet metal approx. 90 cm high and 90 cm deep to prevent escape of or entry by burrowing animals. Each enclosure was planted with alfalfa in summer
1991. During the course of my study, vegetation was maintained below a stand height of 0.4 m. A 1 m wide strip along the inside fence of each enclosure was kept bare or mowed short to reduce usage and therefore minimize escape rates of small mammals. One hundred trap stations set in a 10 x 10 array at 5 m intervals were established in each enclosure. Each station was marked by a 70 cm high wire flag. One Sherman live trap was placed at each trapping station for a total of 100 traps/enclosure.
Study Animal
The biology of Peromyscus, including deer mice (P. maniculatus) has been reviewed by King (1968) and Kirkland and Layne (1989). Deer mice are new world rodents in the
Family Muridae, Subfamily Sigmodontinae (Carleton 1989). Gestation lasts 24 days and litters range between 2-8 pups
(mean = 4.5; Millar 1989). Laboratory-reared pups from the MacDonald Forest population could be weaned as early as 18 days and both sexes can become reproductively mature as early as 60 days (pers. obs.). Typically, juvenile females remain on their natal home range and juvenile males disperse 26 before they become reproductively mature (Wolff 1989).
Breeding tends to be seasonal with the majority of reproduction occurring in spring and fall, however, a small percentage of animals continue to breed throughout the year in many years. Typically, females born in the spring begin to reproduce in the fall but females born in the fall do not reproduce until the following spring (pers. obs.). Mating systems are generally promiscuous, ie., both sexes mate with several different partners (Wolff 1989).
Deer mice live in a wide variety of habitats including meadows, clear-cuts, riparian stream beds, and in old growth
(MacMillen & Garland 1989). At most densities they have overlapping home ranges of up to 750 m2. However, aggression increases with density and at high densities during the breeding season, they exhibit territoriality (Wolff 1989).
During colder weather, non-reproductive animals may nest in groups of up to 4. They primarily eat seeds and arthropods, but will also eat conspecific young (pers. obs.). Deer mouse populations exhibit annual density cycles with peaks occurring in the spring. Several factors are thought to affect population density including food availability, intra- and interspecific competition and predation (Kaufman & Kaufman 1989). 27
Chapter 2
Parasite Life Histories
C.A. Fuller, J. Hefner and E. Wrosch
In Press, Journal of Parasitology 28
Abstract
We compared the life histories of 2 eimerians (Eimeria arizonensis and E. delicata) which co-occur in the deer mouse (Peromyscus maniculatus). Laboratory-reared deer mice were given 103, 104 (E. delicata), or 103, 104, or 105 (E. arizonensis) oocysts by stomach intubation. Eimeria arizonensis infections lasted longer (11-13 days) than E. delicata infections (9-10 days). Eimeria arizonensis infections also produced more oocysts for each oocyst ingested at both levels compared. Both parasites exhibited periodicity in oocyst output at all levels compared.
However, peaks in E. arizonensis output occurred at approximately 20-24 hr intervals whereas peaks in E. delicata output occurred at 12-16 hr intervals. Finally, deer mice developed immunity to both eimerians after only 1 inoculation at all levels tested. Based on these parameters, we expect E. arizonensis to have a greater reproductive potential than E. delicata in free-living deer mice.
Introduction
Many aspects of host-parasite relationships are receiving increasing attention from ecologists including population regulation of animals by their parasites (Scott and Dobson, 1989; Gregory, 1991; Holt, 1993), host-parasite coevolution (Reduker et al., 1987; Gardner, 1991) and the effect of parasites on host behavior (Schall and Dearing, 29
1987; Dobson, 1988; Brodeur and McNeil, 1989; Moore et al., 1994). To conduct basic ecological research in these areas, it is vital to obtain information about parasite life histories and the interaction between the host environment and the parasite. For example, the age and immune status of a host population may affect a parasite's ability to invade those hosts and thus will influence the population dynamics of both the host and/or parasite (Anderson, 1991). Eimerians (Apicomplexa: Coccidia) are excellent parasites to use in ecological studies in the laboratory and in the field. They are easily maintained and experimental infections are easily produced. Because they are generally parasites of the gastrointestinal tract which shed oocysts in their host's feces, long-term, non-invasive monitoring of infections can be conducted. Finally, infections generally are of short duration (usually < 20 days), so it is possible to follow repeated infections in individual hosts.
Eimerian parasites of domestic animals, e.g., poultry and livestock, have received a great deal of attention (Catchpole, et. al 1976; Rose, 1987; Augustine, 1988). In these systems, much is known about the effect of host age and immune response, inoculum level and periodicity. Much less is known about coccidians which occur exclusively in natural populations. For example, hundreds of rodent coccidians have been described (Levine and Ivens, 1989); however, only a few studies have documented aspects of their 30 life histories (Tilahun and Stockdale, 1981; Higgs and
Nowell, 1988) . Here, we present a comparison of the infection patterns caused by Eimeria arizonensis and E. delicata in the deer mouse, Peromyscus maniculatus. The prepatent (the period between ingestion of infective oocysts and shedding of oocysts) and patent (period during which oocysts are shed) periods, periodicity of oocyst output and the effect of a prior infection were investigated for both eimerians.
Methods
Oocysts of Eimeria arizonensis were isolated from the feces of 2 P. maniculatus live-trapped from MacDonald
Douglas Forest (14 km NW, Corvallis, Benton County OR) in
May 1990. Oocysts of Eimeria delicata were isolated from 8 P. maniculatus from the same population in May-September 1991. Sporulated oocysts of both species were maintained at
4 C in 2.5% (w/v) aqueous potassium dichromate (K2Cr2O7) solution and new oocysts were generated periodically in captive P. maniculatus (see below). Oocysts were no more than 7-wk-old at the time of inoculation.
Ten male and 10 female P. maniculatus collected from MacDonald Douglas Forest in the fall of 1989 were used to start a coccidian-free colony. Animals from the colony were never observed to shed oocysts and all hosts used in this study were 3rd-5th generation laboratory-reared. Deer mice 31 were verified to be coccidian-free by microscopic examination of feces several times before inoculation. All deer mice were housed in hanging wire mesh cages, provided with nesting material and Purina Mouse Breeder Chow and water ad lib. and a pan was suspended from each cage to collect fecal samples. The room was on a 8L:16D light-dark cycle; the temperature was maintained between 24-28 C. These conditions approximate those found May September in western Oregon.
Twenty-five, 9-wk-old P. maniculatus were randomly divided into 5 groups, anesthetized with metafane and inoculated by stomach intubation with either 103, 104, or 105 sporulated E. arizonensis oocysts, or with 103 or 104 sporulated E. delicata oocysts (N = 5 in all groups). Feces were collected from each individual deer mouse for 20 days
PI in 24 + 0.25 hr periods. Samples were stored in 2.5%
(w/v) K2Cr207 solution at 4 C until examined.
Feces were collected in 4 hr sampling periods on days
5-7 PI to determine whether periodicity in oocyst output occurred. These samples were collected in dry pans and weighed before suspension in K2Cr207 solution.
Fecal samples were homogenized and filtered through a
285 um mesh screen, allowed to settle and then resuspended in a total volume of 10 or 50 ml water (for 4 and 24 hr samples, respectively). The oocysts in 2 subsamples were counted using MacMaster's counting chambers using sugar for flotation (Dunn & Keymer, 1986). Samples from each 32 consecutive day were examined until no oocysts were seen for
3 days and every 2-3 days thereafter until day 20 PI. If no oocysts were detected in 4 hr samples, the entire sample was examined using Sheather's sugar flotation method (Ash &
Orihel 1987). On day 21 PI all animals (except 1 animal originally inoculated with 104 E. arizonensis oocysts which died during the challenge inoculation) were reinoculated with the same number and species of oocysts as in the initial inoculation.
In addition, 18 naive control animals were similarly inoculated (N: E. arizonensis, 103=4, 104=5, 105=5; E. delicata, 103=4). Feces from individual animals were collected for 20 days PI. Because these samples were very large, they were filtered through a graded series of brass mesh filters (30 and 60 mesh) and stored in 2.5% (w/v)
K2Cr2O7 solution until examined. The filtrate was suspended in 100-500 ml water and 4-5 subsamples were examined using
MacMaster's chambers. If no oocysts were detected, 2, 5 ml samples were examined by Sheather's sugar flotation method. There were clear differences in the number of oocysts shed by naive treatment animals during the first inoculation and naive control animals during the challenge inoculation.
Thus, inoculation period (first inoculation or challenge inoculation) was included as a factor in all analyses of total oocyst output and reproductive index. The oocyst output of E. arizonensis was compared across treatment levels by 2-Way ANOVA. Levels of the treatment were compared 33 using the Least Significant Difference test. T-tests were used to compare output of E. delicata oocysts between animals inoculated with 103 oocysts (first and challenge inoculations) and between animals inoculated with 104 oocysts (first inoculation only). Where necessary, data were arcsin and log transformed to meet the assumptions of parametric analyses. In all cases, alpha of 0.05 was considered significant.
Results
Daily Oocyst Output
The prepatent period of E. arizonensis was 4 days.
Animals began to shed oocysts day 4 PI with a peak in oocyst output generally on day 5-8 PI (Figure 2.1 A). After 10 days animals inoculated with 104 and 105 oocysts no longer shed oocysts. Oocysts were detected in feces of animals inoculated with 103 oocysts until day 12 PI.
The prepatent period of E. delicata was 6 days and the patent period was 3 days for animals inoculated with 104 and
4 days for animals inoculated with 103 oocysts (Figure 2.1 B). The peak in oocyst output occurred on day 7 PI. In addition, animals shed small (<1000) numbers of oocysts intermittently until day 20 PI (when animals were reinoculated). 34
103 OOCYSTS 04 OOCYSTS 105 OOCYSTS 107 A 106
106
104
103
102
10'
10°
106
105
104 H
103 7
102 H
10'
10°
2 3 4 5 6 7 8 9 10 11 12 13 14 DAYS POST-INOCULATION
Figure 2.1. Oocyst output of E. arizonensis and E. delicata. Total daily oocyst output in deer mouse feces from days 2-14 post-inoculation for Eimeria arizonensis (A) and E. delicata (B). Animals inoculated with E. delicata shed small (< 1000) numbers of oocysts intermittently until reinoculated (not shown). Animals were inoculated with 103, 104 or 105 oocysts on day 0. Values given at each inoculum level are means (+SE) of five animals. 35
Diurnal Periodicity in Oocyst Output
Both E. arizonensis and E. delicata exhibited a clear cycle in oocyst output (Figure 2.2 A,B). For E. arizonensis peaks in oocyst output occurred during the dark phase and were most apparent after day 5 PI. Eimeria delicata showed 4 peaks in 2 days at 12 16 hr intervals. Thus, periodicity occurred in this species, but not a diurnal cycle.
Fecal Oocyst Concentration
Individual deer mouse fecal pellets weighed 0.02 + 0.0013 g. This value was used to estimate the concentration of oocysts per fecal pellet during light and dark periods.
Oocyst concentration of E. arizonensis ranged from 1.5-14.6 x 104 oocysts per fecal pellet and pellets produced during the light period had a 1.3-2 greater concentration of oocysts than pellets produced during the dark period (Table
2.1).
Oocyst concentration per pellet of E. delicata was much lower than for E. arizonensis; however, pellets produced during the light period also contained a higher concentration of oocysts than those produced during the dark period (Table 2.1). 36
103 00CYSTS 104 00CYSTS 105 00CYSTS
A 40
30
20
10
LIGHT DARK LIGHT DARK LIGHT DARK 0
5 6 7 8
DARK LIGHT DARK LIGHT 50-1 B
40-?
30
20 -1
10
0 7 8 9 DAYS POST-INOCULATION
Figure 2.2. Percent daily oocyst output. Percent of total daily oocyst output shed in deer mouse feces during four hr sampling periods for Eimeria arizonensis (A) and E. delicata (B) during days of peak oocyst output. Animals were on a 16L:8D light-dark cycle. Symbols represent the end of each sampling period. 37
Table 2.1. Oocysts per fecal pellet. The number of oocysts per Peromyscus maniculatus fecal pellet (0.02g feces) during the day and night on various days post-inoculation (PI). Values represent means of 5 animals in all cases; It = light period = 16 hrs. dk = dark period = 8 hrs. Table 2.1 Oocysts of Oocysts of Eimeria arizonensis (x 104) Eimeria delicata (x 102)
Inoculum Day 5 PI Day 6 PI Day 7 PI Day 7 PI Day 8 PI level It dk It dk It dk It dk It dk
103 6.3 4.9 14.6 7.3 4 2.6 21.4 2.6 3.8 2.3
104 9.2 6.7 7.3 4.0 - 18.4 4.4 4.9 2.2
105 3.3 1.9 2.3 1.5 - - -- - 39
Effect of Inoculum Period and Level
Both inoculum period and level of E. arizonensis significantly affected total oocyst output of naive animals
(Table 2.2). Total oocyst output increased with inoculum period and decreased with the number of oocysts inoculated
(Fig. 2.3 A). In addition, there was a significant interaction between period and level (Table 2.2), indicating that the effect of period depended on the number of oocysts ingested. The reproductive index (RI = the number of oocysts shed per oocyst ingested) also increased with period and decreased with level (Table 2.3).
For E. delicata there was no significant difference in total oocyst output or RI of naive animals between the first and challenge inoculations at 103 oocysts (t = 0.08, P >
0.05; Figure3B) or in total oocyst output at different inoculum levels (t = 2.287, P > 0.05). However, RI for animals inoculated with 103 oocysts was higher than for animals inoculated with 10' oocysts (Table 2.4; t = 3.611, P
< 0.01).
Immunity
All animals shed fewer E. arizonensis oocysts during challenge infections than when initially inoculated (Table
2.4). Only 1 of 14 animals showed < 90% decrease in oocyst output compared with the output of controls inoculated with the same number of oocysts. This animal shed 34% of the 40
Table 2.2. ANOVA of the number of oocysts excreted. The factor "level" is the number of oocysts inoculated and "period" is whether animals were in the first or second experimental group.
Factor df F P
Level 2 20.4 0.001
Period 2 18.0 0.001
Interaction 4 3.7 0.04
Error 34
Table 2.3. ANOVA of reproductive index (number of oocysts excreted for each oocyst ingested). The factor "level" is the number of oocysts inoculated and "period" is whether animals were in the first or second experimental group.
Factor df F P
Level 2 131.36 0.001
Period 2 58.08 0.001
Interaction 4 2.69 0.05
Error 34 41
30 A c(4) Naive Treatment Animals 25 Naive Control Animals Co0 20 C w VJ > 15 0 0 --I 10 I 0 b
1000 10,000 100,000
20 B d
1000 10,000 INOCULUM LEVEL
Figure 2.3. Total oocyst output over a 20 day infection (+SE). Deer mice were naive (not previously infected) when inoculated with Eimeria arizonensis (A) and E. delicata (B). The sample size was 5 except where shown by parentheses. The same letters indicate grou7s that did not differ significantly at the 0.05 significance level (Least Significant Difference Test). 42 Table 2.4. Immunity to E. arizonensis. The effect of challenge infection on Eimeria arizonensis oocyst output in deer mice.
Mean oocyst output x 106 (+SE) Inoculum of animals at 1st & 2nd % Reduction in level inoculation oocyst output*
Naive control Challenge
103 10.15 (1.84) 0.54 (0.17) 97.5
104 8.38 (2.33) 0.39 (0.23) 97.9
105 3.09 (0.67) 0.39 (0.24) 95.5
* compared to naive control animals ** excluding one outlier (see text) 43 expected value, probably did not become immune and was excluded from further analysis. Table 2.4 shows that the mean percent decrease in oocyst output was at least 90%. Three animals were challenged with E. delicata at each inoculum level. None of these animals shed fecal oocysts for at least 20 days PI.
Discussion
In this study we examined several parameters of the life history of 2 eimerians. Eimeria arizonensis has been reported previously in P. maniculatus, P. eremicus, P. leucopus, Reithrodontomys fulvescans, R. meqalotis and R. montanus (Reduker et al., 1985, Upton et al., 1992); E. delicata has only been reported in P. maniculatus (Levine and Ivens, 1960; Reduker et al., 1985). However, with the exception of prepatent and patentperiods for E. arizonensis, little is known about the life history of these 2 species.
Daily Oocyst Output The prepatent and patent periods reported herein for E. arizonensis are similar to those reported previously. Reduker et. al (1985) found a prepatent period of 3-4 days and a patent period of 9-11 days in deer mice. Upton et. al
(1992) reported a patent period of 4-14 days in the western harvest mouse (Reithrodontomys megalotis) and a patent period of 4-16 days in the pinyon mouse (P. truei). Eimeria 44 delicata had a longer prepatent and shorter patent period than E. arizonensis (Table 2.5). Interestingly, the patent period in deer mice inoculated with 103 oocysts of either species was longer (+2 days for E. arizonensis; +1 day for
E. delicata) than in animals inoculated with higher oocyst levels. This suggests that the patent period depends on initial dose. Oocysts of E. delicata also reappeared intermittently after the majority of oocysts were shed but
E. arizonensis oocysts did not. Although it is possible that animals became reinfected with E. delicata by cage contamination, despite precautions taken (frequent cage changes, wire cages), it seems unlikely. When the patent period and sporulation time (minimum of 5 days; pers. obs.) are taken into account, animals that were reinfected from contaminated cages would not begin to shed oocysts until at least day 16 PI, but E. delicata oocysts reappeared as early as day 12 PI. We found that the total oocyst output (E. arizonensis) and the reproductive index (both E. arizonensis and E. delicata) declined with increasing inoculum. This pattern of negative density dependence has been observed in many other eimerians (Hall, 1935; Brackett and Bliznick, 1952; Tilahun and Stockdale, 1981; Higgs and Nowell, 1988; Lillehoj,
1988). In general, these studies report an optimum inoculum which produces the largest reproductive index. Below this level, the parasites may be too sparse to reproduce efficiently during the sexual phase. Above the optimum 45 level, host cells/nutrients may be limited (see Fernando
[1982] for a review of potential mechanisms). We did not find a lower limit for either species; however, our lowest dose (103) may have already been higher than the optimum.
It is unclear why output of oocysts increased in the second group of naive animals we inoculated. Oocyst viability may have changed slightly in the 3 weeks between inoculations although it seems unlikely that enough oocysts would die to create the differences seen here. Alternately, the deer mice may have undergone sexual maturation during this time (Millar, 1989). Several studies have shown effects of age and reproductive hormones on parasites (Augustine, 1988; Higgs and Nowwell, 1988; Harder et. al, 1992).
Diurnal Periodicity in Oocyst Output and Fecal Oocyst Concentration
Although periodicity has been documented in many helminth parasites (Hawking, 1975), there have only been a few studies of this phenomenon in coccidial parasites (genera Isospora and Eimeria) and, to our knowledge, all of these have been in avian hosts (Boughton, 1988). These studies showed that peak oocyst output occurred shortly before or during the beginning of the dark period. If the light-dark cycle was changed, oocyst output also changed such that peak oocyst output still occurred at the beginning of the dark period. Parasite invasion of host tissue occurred at the same time relative to the light period 46 regardless of when the inoculation occurred (Boughton, 1988). Apparently periodicity is determined by environmental factors and is not intrinsic to the parasite or the host. The cycle of oocyst output in E. arizonensis was similar to those reported in previous studies in that peak oocyst output occurred at night, despite the fact that, in contrast to the birds used in Boughton's studies, deer mice are nocturnal. Interestingly, E. delicata showed roughly 2 peaks in oocyst output per day rather than 1. Boughton (1988) suggested that there are 2 waves of cellular invasion 24 hr apart and each reproductive cycle lasts 48 hr. Thus, peaks occur at 24 hr intervals. Eimeria delicata may be unusual in having accelerated development. More total oocysts were shed at night; however, because less feces was produced during the day, each daytime pellet contained a higher concentration of oocysts than those produced at night (Table 2.1). This may have ramifications for parasite transmission. The oocysts shed at night will be in relatively low concentrations and spread widely throughout the environment. However, there will be a very concentrated oocyst source inside of deer mouse nests. Animals from the population from which our laboratory stock originated have been observed to use latrine sites below bedding in artificial nest boxes and to share nests with up to 3 adults (C.A. Fuller, pers. obs.). Nests may provide a sufficiently warm and moist environment to facilitate oocyst sporulation. Thus, deer mice may be exposed to high 47
concentrations of oocysts in their nests and this may be the primary site of transmission.
Immunity
Both species examined in this study elicited immunity
(>90% reduction in oocyst output) after 1 large, inoculating dose. However, E. arizonensis elicited partial immunity while E. delicata seemed to elicit complete immunity at both
levels tested. Other studies have reported a large decrease
in oocyst output after 1 large exposure (Hall, 1935;
Augustine, 1988; Lillehoj, 1988) or after several smaller,
e.g., 102, doses (Higgs and Nowell, 1988, Fitz-Coy and
Edgar, 1989). Both cell mediated and antibody response are thought to be involved in the immune response to eimerians
(Rose, 1987) .
Although E. arizonensis and E. delicata co-occur in the
deer mouse, they vary widely with respect to most life-cycle parameters examined (Table 2.5). Eimeria arizonensis infections produced a much larger number of oocysts over a slightly longer patent period and did not elicit complete immunity after 1 exposure as did E. delicata infections. In addition, the number of oocysts per fecal pellet was much
(3000-5000 x) greater for E. arizonensis. These findings
suggest that E. arizonensis has a much greater reproductive potential in wild deer mice than E. delicata. 48
Table 2.5. A comparison between Eimeria arizonensis and Eimeria delicata life history parameters.
Parameter Eimeria arizonensis E. delicata
Length in days: Prepatent period 3 6 Patent period 7-9 3-4
Total oocyst output > 106 « 106
Development of immunity partial complete
Periodicity in oocyst output 1 peak/day 2 peaks/day
Oocysts/fecal pellet > 1.5 x 104 < 3000 49
Chapter 3
Population Dynamics of E. arizonensis and E. delicata 50
Abstract
Because many parasites have a free-living stage, their survival is affected by 2 sets of factors: 1) host factors and 2) abiotic factors in the external environment. I examined the population dynamics of 2 deer mouse eimerians
(Protozoa) under field conditions to determine which factors were more important. The prevalence of Eimeria arizonensis was not affected by host sex or reproductive condition and increased with host age. However, E. arizonensis prevalence was strongly correlated with environmental temperature. In contrast, E. delicata had a significantly higher prevalence in juvenile than in adult deer mice, but was not related to any other factor measured. In addition, there was no evidence of competition between the 2 eimerians. These data suggest that host factors were probably less important than abiotic factors in the population dynamics of E. arizonensis and more important than abiotic factors in E. delicata.
Introduction
By definition, parasites have a negative impact on their host. Thus, parasites may be important in host population regulation (Anderson & May 1979, May & Anderson
1979). Parasites may also cause or exacerbate the decline of threatened or endangered host species (Scott 1988).
Understanding the basic population dynamics of parasites can give us insights into which individual hosts will be 51 affected (eg., male or female, old or young) and whether the effect will occur at the population level (Anderson 1986,
Scott 1988). Because parasites generally live in at least 2 environments (within/attached to the host and free-living outside of the host), 2 sets of factors may affect their survival. Within host factors include host immune status and resource competition among parasites (Holmes et al. 1977).
Factors external to the host include abiotic parameters such as temperature, moisture and uv radiation (Levine 1980).
Two different categories of parasites are recognized:
1) microparasites (viruses, bacteria and protozoans) go through multiple reproductive cycles within the host and usually induce complete immunity after only one infection
(Anderson & May 1979); 2) macroparasites (helminths and arthropods) shed eggs or larvae from the host and larvae must infect a new host before reproduction occurs. Acquired immunity is often partial and the degree of immunity may be dose (density) dependent (May & Anderson 1979). In addition, acquired immunity to macroparasites may be lost if enough time passes between infections. Several studies have documented population dynamics of natural macroparasite infections in free-living hosts
(Arnett et al. 1993, Gregory 1992, Gulland & Fox 1992,
Prestrud et al. 1993, Shaw & Moss 1989, Theis & Schwab 1992,
Wilson 1983). Less attention has focused on the population dynamics of microparasite infection (Ball & Lewis 1984,
Healing 1981, Stanton et al. 1992, Weatherhead & Bennett 52
1991, 1992, Wilber et al. 1994). I compared the population dynamics of 2 deer mouse protozoans, Eimeria arizonensis
(Ea) and E. delicata (Ed; Apicomplexa; Eimeriidae). These parasites have only one host, are transmitted by ingestion
of infective oocysts (the form excreted) and generally live
in the hosts' intestinal epithelium. The infections are usually of short duration (< 20 days) and, because reproduction occurs by multiple fission within the host, as many as 106 oocysts can be produced for each oocyst
ingested. Oocysts are shed directly into the lumen of the
intestine and can be detected by fecal examination. Oocysts
are shed in an uninfective form and require approx. 1 week at room temperature (pers. obs.) to become infective
(sporulate). During this time, extreme temperatures and humidity have been shown to affect oocyst survival and
sporulation rate (Graat et al. 1994, Marquardt et al. 1960).
Although eimerians are generally categorized as microparasites because of their mode of reproduction, they are often more similar to macroparasites with respect to acquired immunity (Healing & Nowell 1985, Quinnell & Keymer
1990) .
In this study, I compared the population dynamics of E. arizonensis and E. delicata using naturally-infected free- ranging deer mice. I used these data to determine whether host factors, environmental factors or a combination of the two are important determinants of parasite abundance in this system. If host factors are the primary determinant of 53 parasite population dynamics, several predictions can be made. First, if competitive exclusion among parasites is important, E. arizonensis and E. delicata will be found together less frequently than expected by chance (Reduker &
Duszynski 1985, Lafferty et al. in press). Second, if immunity is important, young animals should be parasitized more frequently than older animals due to exposure
(Prediction 2a). Alternatively, because reproductive hormones often lead to changes in immunocompetence
(Alexander & Stimson 1988, Bundy 1988), there may be differences in parasite prevalence among males and females or reproductive and non-reproductive animals (Prediction
2b). Third, if environmental factors are most important in determining parasite population dynamics, temperature and/or moisture during oocyst sporulation will be correlated with parasite prevalence.
Methods
Data Collection
Two groups of deer mice were captured in Sherman live- traps in MacDonald Forest (15 km NW Corvallis, OR). The first group was part of a mark-recapture study. These animals were individually marked and released at their point of capture. Only the first capture of each animal was used in the current study. The second group of animals was collected by live-trapping at other sites around MacDonald 54
Forest. These sites were not trapped more than once and were at least 500 m from the mark-recapture grid. Both groups were captured between 15 April 17 Nov 1990, 30 March 17
Nov 1991 and 7 March 12 July 1992. Animals from both sources were processed in the same manner and assumed to be comparable with respect to Eimeria infections. Animals were assessed for age (juvenile or adult, based on pelage) and reproductive status (females: reproductive = pregnant and/or lactating, non-reproductive = no visible signs of reproduction; males: reproductive = scrotal testes, non- reproductive = abdominal testes). All feces were collected from traps, weighed and incubated in a 2.5% w/v potassium dichromate (K2Cr2O7) solution at room temperature for one week to allow any coccidia present to sporulate (this is necessary to identify oocysts to species). Fecal samples were macerated and examined for the presence (prevalence) of oocysts by Sheather's sugar flotation (Ash & Orihel 1987).
Oocysts can only be detected in patent infections. However,
the length of the prepatent period (the time from ingestion of infective oocysts to the beginning of oocyst output) was
obtained from laboratory experiments (prepatent to patent ratio: Ea = 3:9 days, Ed = 6:4 days; Chapter 2) and used to estimate the total number of animals infected. In addition, the number of oocysts per gram of feces was estimated as 0,
1-9, 10-99, 100-999, etc., to compare the severity of
infection between Eimeria species. 55
Temperature and rainfall data were obtained from a weather station at Hyslop Farm, approx. 2.5 km east of
MacDonald Forest.
Statistical Analyses
In all cases, alpha of 0.05 was considered statistically significant.
Host Factors
Data were pooled by month because trapping occurred at variable intervals. Due to the high occurrence of 0 prevalence in monthly samples, transformation did little to increase the normality of the data set. Thus, nonparametric statistics were used in analyses comparing infection prevalence among various sex, age and reproductive classes of deer mice. To account for seasonal variability, prevalence data were compared with either Wilcoxon Sign-Rank
Test or Friedman's 2-way ANOVA (Zar 1984). These analyses compared the data set by month either pairwise (Wilcoxon
Test) or by groups of > 2 (Friedman's ANOVA). In both analyses, any month with missing values or tied ranks was excluded. If Friedman's ANOVA was significant, Wilcoxon Test was used as a post-hoc test to determine which groups differed. Where predictions were tested, 1-tailed alpha levels are given; other alpha levels are 2-tailed. 56 To determine whether E. arizonensis and E. delicata
occurred together more or less frequently than expected by
chance, data were divided into 4 time periods (late 1990,
early 1991, late 1991 and early 1992). The probability of having both infections was calculated as the product of the
proportion of animals infected with either eimerian:
PEaxPEd.
The occurrence of observed and expected double infections
within each time period were compared by X2 analysis.
Environmental Factors
In deermice, the prepatent period of E. arizonensis and
E. delicata is 3 and 6 days respectively. Oocysts are shed for up to 9 (Ea) or 4 (Ed) days and sporulation takes 4-7 days under laboratory conditions (Chapter 2). If environmental factors such as rainfall or temperature affect prevalence of infection, they should have the greatest impact during oocyst sporulation (see above) because this is the primary time when oocysts are exposed to the external environment. This period occurs 7-13 days before the actual observed infection. Thus, I used mean high and low temperature and total rainfall for the 2-week period prior to the capture period in analyses.
I used stepwise multiple regression to compare the relative strengths of environmental factors (Wilkinson 1990). All variables were examined for normality and 57 transformed as necessary prior to analysis. The residual errors from the final model were examined graphically to insure they met the assumptions of linear regression.
Results
Overall Prevalence and Intensity
432 deer mice were captured in 21 months of the study.
Of these, all were assessed for the presence of E. arizonensis and 348 were assessed for the presence of E. delicata (E. delicata was incorporated into the study in
August 1990). Over the course of the study, the prevalence of E. arizonensis ranged from 0-80% and of E. delicata from
0-53% (Figure 3.1 A). Peaks in E. arizonensis prevalence occurred in spring (June, April and May in 1990, 1991 and
1992) and spring/summer for E. delicata (July and May, 1991 and 1992).
There were no significant differences in monthly prevalence of patent infections between E. arizonensis and E. delicata infections (Wilcoxon Sign-Rank Test, z=0.3341,
N=16 months). However, when the estimated number of animals with prepatent infections (see above) was added to animals shedding oocysts, significantly fewer animals were infected with E. arizonensis than with E. delicata (z=2.43, P=0.015,
N=16; Figure 3.1 B).
Oocyst output per gram feces was quantified in 94 E. arizonensis and 94 E. delicata infections. Infected deer 58
1.0 A 1990 1991 1992
B
0.0 Ma Ju Se Ma Ma Ju Se No Ap Ju
Figure 3.1. The prevalence of E. arizonensis and E. delicata. The proportion of deer mice shedding oocysts (ie, with patent infections) of Eimeria arizonensis (solid lines) and E. delicata (dotted lines) (A) and the estimated proportion of animals with prepatent and patent infections (see text) (B). The proportion of animals with patent and prepatent infections combined was greater for E. delicata but there were no differences between species for the number of animals with patent infections alone. See tables 3.2 and 3.3 for sample sizes. 59 mice shed significantly greater numbers of E. arizonensis than E. delicata oocysts (Figure 3.2; X2=36.15, df=4, P <
0.001) .
Hypothesis 1: Host Factors
Prediction 1: Occurrence of Double Infections
Eimeria arizonensis and E. delicata were not found together significantly more or less frequently than expected by chance in any of the 4 time periods examined (Table 3.1).
Prediction 2a: Variability Among Sub-Populations of Deer Mice, Age and Sex Differences
Friedman's analysis indicated that there were significant differences in prevalence among different classes of animals with naturally occurring E. arizonensis infections (Fr=8.61, df=3, P < 0.02; Table 3.2). Further analysis showed that juveniles were infected less frequently than adults, the opposite of predicted (z=2.86, N=14, P <
0.005). In addition, males were infected more frequently than females; however, this difference was not significant
(z=1.76, N=17, P=0.08). There were also significant differences in prevalence among E. delicata infections (Fr=6.69, df=3, P < 0.04; Table
3.3). As predicted, juveniles were infected more frequently than adults (z=2.05, N=11, P < 0.02). There were no 60
40 E. arizonensis E. delicata
30
10
7 0 1 2 3 4 5 6 Oocysts Per Gram Feces
Figure 3.2. The intensity of E. arizonensis and E. delicata. The number of animals with various levels of oocysts in their feces (1 = 1-9 oocysts/gram, 2 = 10 99, etc.). The data were combined for all infected animals in which oocysts/gram was assessed across the 3 years of the study. Animals infected with Eimeria arizonensis shed significantly more oocysts than those infected with E. delicata. 61
Table 3.1. Frequency of double infections. The number of deermice infected with separate Eimeria arizonensis and E. delicata infections, infected with both or neither. E. arizonensis and E. delicata did not appear together more or less frequently than expected by chance in any of the 4 time periods. This suggests that facilitation or cross-immunity are unlikely. Numbers in parentheses indicate the number of months included in each time period. Df = 1 in all analyses.
Time period Ea Ed Both Neither X2
1990-late (3) 8 13 3 56 0.29
1991-early (4) 31 16 17 41 0.24
1991-late (5) 5 24 5 50 0.75
1992-early (5) 18 20 2 39 2.31 62
Table 3.2. Percent infected (sample size) with Eimeria arizonensis in different sex and age classes of deer mice.
Adults Juveniles Males Females Males Females
1990 April 12 (16) 18 (11) 0 (4) 25 (4) May 20 (5) 0 (1) 20 (5) 0 (2) June 33 (6) 100 (1) 25 (4) 0 (5) July 0 (9) 12 (8) 0 (3) (0) Aug 11 (9) 12 (8) 0 (7) 0 (9) Sept 38 (13) 18 (11) 12 (8) 8 (12) Oct 0 (1) 0 (2) (0) (0)
1991 March 67 (6) 75 (4) (0) (0) April 83 (6) 75 (4) (0) (0) May 50 (22) 42 (24) (0) (0) June 38 (21) 25 (12) 33 (3) 0 (3) July 33 (6) 0 (3) 29 (7) 0 (1) Aug 0 (9) 25 (8) 0 (2) 11 (9) Sept 0 (6) 0 (3) 0 (3) 0 (1) Oct 20 (5) 0 (2) 0 (5) 20 (5) Nov 0 (2) (0) 20 (5) 0 (2)
1992 March 20 (5) 25 (8) (0) 0 (1) April 75 (4) 50 (2) 0 (1) 40 (5) May 33 (3) (0) (0) (0) June 40 (15) 6 (18) 25 (4) 0 (1) July 50 (4) 0 (2) 0 (4) 0 (2) 63
Table 3.3. Percent infected (sample size) with Eimeria delicata in different sex and age classes of deer mice.
Adults Juveniles Males Females Males Females
1990 Aug 11 (9) 0 (8) 0 (7) 22 (9) Sept 31 (13) 0 (11) 37 (8) 50 (12) Oct 0 (1) 0 (2) (0) (0)
1991 March 33 (6) 50 (4) (0) (0) April 33 (6) 25 (4) (0) (0) May 27 (22) 29 (24) (0) (0) June 38 (21) 8 (12) 67 (3) 67 (3) July 33 (6) 67 (3) 71 (7) 0 (1) Aug 44 (9) 25 (8) 50 (2) 56 (9) Sept 67 (6) 0 (3) 0 (3) 0 (1) Oct 20 (5) 0 (2) 60 (5) 0 (5) Nov 0 (2) (0) 0 (5) 0 (2)
1992 March 40 (5) 50 (8) (0) 100 (1) April 0 (4) 0 (2) 0 (1) 20 (5) May 33 (3) (0) (0) (0) June 27 (15) 22 (18) 50 (4) 0 (1) July 25 (4) 0 (2) 50 (4) 0 (2) 64 significant differences in prevalence between males and females (z=1.29, N=14) .
Prediction 2b: Differences Among Reproductive Classes
There were no significant differences in prevalence of naturally-occurring E. arizonensis infections among reproductive and nonreproductive males (Table 3.4; Z=0,
N=11). However, reproductive females were infected significantly more frequently than nonreproductive females
(Table 3.4; z=2.07, N=9, P < 0.04).
There were no significant differences in the prevalence of naturally-occurring E. delicata infections among reproductive and nonreproductive males or females (Table
3.5; z=0.153, N=10 and z=0.398, N=7 for males and females, respectively).
Hypothesis 2: Environmental Factors
The prevalence of E. arizonensis was negatively correlated with mean high temperature (r = -0.725, N = 24, P < 0.001), mean low temperature (r = -0.583, P < 0.005). The correlation between prevalence and total rainfall was not significant (r = 0.382, P < 0.10). In stepwise regression analysis, only mean high temperature explained a significant amount of variance (prevEa = 103.4 + -1.067temp, F = 24.37,
P < 0.001, R2 = 0.526) . 65
Table 3.4. Percent infected (sample size) with Eimeria arizonensis in different reproductive classes of adult deer mice. R = reproductively active, NR = not reproductively active.
Males Females
R NR R NR
1990 April 33 (3) 0 (2) 50 (2) 0 (3) May 33 (3) (0) (0) (0) June 20 (5) 100 (1) 100 (1) (0) July 0 (7) (0) 14 (7) (0) Aug 17 (6) 0 (2) 14 (7) 0 (1) Sept 0 (2) 50 (8) 25 (4) 14 (7) Oct (0) (0) (0) 0 (1)
1991 March 33 (3) 100 (3) 100 (1) 67 (3) April 100 (1) 75 (4) (0) 67 (3) May 42 (17) 75 (4) 43 (7) 47 (15) June 33 (15) 50 (6) 43 (7) 0 (5) July 40 (5) 0 (1) 0 (2) 0 (1) Aug 0 (6) 0 (3) 29 (7) 0 (1) Sept 0 (3) 0 (3) 0 (1) 0 (2) Oct (0) 0 (4) (0) 0 (2) Nov (0) 0 (1) (0) (0)
1992 March 33 (3) 0 (2) 20 (5) 33 (3) April 75 (4) (0) (0) 50 (2) May 33 (3) (0) (0) (0) June 47 (15) (0) 9 (11) 0 (5) July 100 (2) 0 (1) (0) 0 (2) 66
Table 3.5. Percent infected (sample size) with Eimeria delicata in different reproductive classes of adult deer mice. R = reproductively active, NR = not reproductively active.
Males Females
R NR R NR
1990 Aug 17 (6) 0 (2) 0 (7) 0 (1) Sept 0 (2) 37 (8) 0 (4) 0 (7) Oct (0) (0) (0) 0 (1)
1991 March 33 (3) 33 (3) 0 (1) 67 (3) April 0 (1) 25 (4) (0) 33 (3) May 35 (17) 0 (4) 29 (7) 27 (15) June 20 (15) 83 (6) 0 (7) 20 (5) July 40 (5) 0 (1) 100 (2) 0 (1) Aug 50 (6) 33 (3) 29 (7) 0 (1) Sept 100 (3) 33 (3) 0 (1) 0 (2) Oct (0) 25 (4) (0) 0 (2) Nov (0) 0 (1) (0) (0)
1992 March 0 (3) 100 (2) 60 (5) 33 (3) April 0 (4) (0) (0) 0 (2) May 33 (3) (0) (0) (0) June 27 (15) (0) 27 (11) 20 (5) July 50 (2) 0 (1) (0) 0 (2) 67
The prevalence of E. delicata was not significantly correlated with any environmental factor (rhightemp 0 . 1 36, rlowtemp = 0.014, rrain = 0.301, N = 20) and no factor explained a significant proportion of the variance in regression analysis. The prevalence of E. arizonensis and E. delicata were not significantly correlated (r = 0.157, N = 20, P > 0.05).
Discussion
Factors Influencing E. arizonensis
There was no evidence of competitive exclusion between
E. arizonensis and E. delicata (Prediction 1). There was also little evidence that deer mice became immune to E. arizonensis as measured by a decrease in prevalence with host age (Prediction 2a). In fact, adults were significantly more likely to be infected than juveniles. This is consistent with other studies. Although we showed that deer mice can become immune under laboratory conditions (Chapter 2), experiments conducted under natural and semi-natural conditions suggest that these animals do not appear to exhibit the same degree of immunocompetence as in the laboratory. GThis pattern is presumably because free-living animals are exposed to more immunosuppressing factors
(Chapter 5). There were differences in E. arizonensis prevalence among sex and reproductive classes of deer mice. Males were 68 slightly more likely to be infected than females and reproductively active adult females were more likely to be infected than nonreproductive adult females. Although the sex ratio varied over the course of the study, it was not significantly correlated to E. arizonensis prevalence (R =
0.041, N = 20). In addition, only 62 of 432 animals (14.3%) were reproductive females. Thus, neither host sex or reproductive class seemed likely to affect E. arizonensis population dynamics (Prediction 2b).
On the other hand, mean high environmental temperature during the period that oocysts were on the substrate, was strongly correlated with E. arizonensis prevalence. This pattern suggests that environmental factors were most important in determining E. arizonensis population dynamics
(Prediction 3) .
Factors Influencing E. delicata
The opposite seems to be true for E. delicata. Juvenile deer mice were far more likely to be infected than adults, suggesting that immunity to E. delicata is common (Prediction 2a). Evidence from the laboratory also suggests that E. delicata is more immunogenic than E. arizonensis; laboratory reared animals developed complete immunity after only one exposure (Chapter 2). Unfortunately, field experiments with E. delicata were not conducted because of logistical constraints. 69
No differences in E. delicata prevalence were seen among sex or reproductive classes of deer mice, suggesting that reproductive hormones are not important in immunocompetence to this species of parasite (Prediction
2b). In addition, E. delicata prevalence was not significantly correlated with any of the environmental factors measured (Prediction 3). Thus, it appears that the population dynamics of E. delicata, in contrast to E. arizonensis, were primarily controlled by host immunity.
Reduker & Duszynski (1985; after I calculated X2 from data given) and Stanton et al. (1990) also found little evidence of competitive exclusion among eimerians. There are several mechanisms through which co-occurring eimerians can avoid competition. Eimerians inhabiting the same host may partition the cells they use. Kheysin (1972) found that species of co-occurring eimerians often utilize different areas of the digestive tract even in the absence of other species. Alternatively, even if > 1 species of parasite is in the same place at the same time, rapid host cell turnover and/or low parasite abundance may prevent competition. In the present study, E. arizonensis and E. delicata may not co-occur in the same host individual; they may avoid competition simply because one primarily occurs in adults and the other in juveniles.
Results are more variable for blood microparasites.
Turner & Cox (1985) also found little evidence of species interactions in rodents, Schall & Bromwich (1994) found 70 positive associations between 2 species of lizard malaria and Forbes et al. (1994) found positive, negative and random associations among blue grouse hematozoans. Thus, further studies among a wider variety of hosts may uncover evidence of non-random associations among eimerians. Several studies have found evidence of acquired
immunity to microparasites among free-living birds and mammals, either by examining age-prevalence relationships
(Ball & Lewis 1984, Healing 1981 in 1 of 4 Hematozoans, Samuel & Trainer 1971, Weatherhead & Bennett 1991 males
only) or differences in prevalence among sex/reproductive
groups (Ball & Lewis 1984, Keith et al. 1986, Weatherhead &
Bennett 1991 in 1 of 4 hematozoa). However, other studies have found no evidence of immunity (age-prevalence: Healing
1981 in 3 of 4 Hematozoa, Keith et al. 1986, Stanton et al. 1992, Wilber et al. 1994; sex/reproductive status: Ball
& Lewis 1984). Thus, while acquired immunity may be important in some host-parasite systems, it seems to be
unimportant in others. In addition, the proportion of animals in a population exhibiting acquired immunity to a single parasite species may vary over time due to host or parasite genetic factors (Wakelin 1992) or immunosuppression
(Chapter 5). Most studies examining the effect of climate on
parasites have found that parasites were more abundant in
mesic than in xeric habitats (Dobson et al. 1992, Ford et
al. 1990, Shults et al. 1990). This is consistent with 71 results for E. arizonensis but not E. delicata in the present study. Interestingly, none of the earlier studies examined the effect of weather within a habitat and all pool several species of parasite. Thus, it is possible that differences between parasite species and within habitats were obscured. In my system, both parasites were exposed to identical environmental conditions during sporulation; however, only
E. arizonensis seems to be primarily controlled by this factor. This pattern suggests that acquired immunity when present, supersedes environmental factors in determining parasite population dynamics, at least within a single habitat. However, even in the presence of acquired immunity, habitat differences may be more important.
Deer mice rapidly develop immunity to E. delicata, thus it seems unlikely that this parasite will influence host population dynamics unless it affects juvenile survivorship.
In contrast, immunity does not play as important a role in
E. arizonensis population dynamics. Deer mice become reinfected frequently and E. arizonensis does affect host survival in the field (Chapter 4). Eimeria arizonensis and
E. delicata are closely related and have very similar life histories, but apparently their populations are limited by different mechanisms. Thus, it is clear that each parasite must be examined individually to determine its effect on its host. 72
Chapter 4
The Effect of E. arizonensis on Deer Mouse Survival
C.A. Fuller and A.R. Blaustein 73
Abstract
By definition, parasites negatively affect their hosts.
However, few experimental studies have documented this effect under natural or semi-natural conditions. We studied the affect of Eimeria arizonensis on deer mouse recruitment, over winter survival and weight. In field observations, E. arizonensis was negatively related with recruitment success in females and over-winter survival in males. Experimental manipulation of E. arizonensis in large field enclosures also showed that it negatively affects male over-winter survival. There was no relationship between deer mouse weight and infections, thus, the mechanism through which E. arizonensis affects survival remains unclear.
Introduction
A major goal of ecology is to understand the factors which affect animal populations. Predation, emigration and competition have been the focus of most of the empirical research on this topic (see review in Hestbeck 1987). However, Anderson and May demonstrated mathematically that parasites could regulate the population density of their hosts (Anderson & May 1979, May & Anderson 1979). Since the publication of their seminal papers, a few studies have directly examined the potential for parasites to regulate host populations in the laboratory (Scott & Anderson 1984, 74
Scott 1987) and in field enclosures (Barker et al. 1991,
Gregory 1991). Although studies of host population regulation by parasites should ideally be conducted at the population
level, this is often impractical in natural free-living animals. Thus, most field work has focused on the effect of
parasites on the individual host (but see Washburn et al.
1991). If a parasite regulates its host's population density, it must negatively affect the survival and/or
reproduction of the host in a density dependent manner.
There is now ample correlative evidence from the field that
parasites in many taxa (e.g., protozoans, nematodes, trematodes and arthropods) have a negative impact on reproduction and/or survival in a variety of hosts,
including mollusks, insects, fish, reptiles, birds and mammals (Arnqvist & Maki 1990, Burgett et al. 1990, Festa-
Bianchett 1989, Goater et al. 1989, Johnson et al. 1991, Lim
& Green 1990, McPhail & Peacock 1983, Schall 1983). Ideally,
controlled field experiments should be conducted to examine
associations found in field observations. However, experimental manipulation of parasite levels under field conditions has been accomplished in only a few host-parasite
systems (Hudson 1986, Hudson et al. 1992a,b, Lehman 1993).
Scott & Dobson (1989) suggested that parasites which
are directly transmitted (i.e., complete their life cycle in
only one host) and have known pathology would be the most
likely to have a detectable effect on their host's 75 population density. Many eimerians (intestinal protozoans in the class Coccidia) meet both of these criteria. They are transmitted directly through ingestion of contaminated feces and many are known to be highly pathogenic (see Long 1982).
We studied the effect of Eimeria arizonensis on deer mouse
(Peromyscus maniculatus) survival using both field observations and field and enclosure experiments. Eimeria arizonensis infections last 10-12 days and up to 107 oocysts are shed in the feces during this period (Chapter 2). Because infections are of relatively short duration, are easy to detect using non-invasive methods and individual hosts become infected repeatedly, this system is ideal for examining the dynamics of infections and the effect of infections on individual hosts. Deer mice are excellent subjects because they readily enter live-traps. In addition, they can be recaptured and sampled repeatedly in the field.
We tested predictions of the hypothesis that E. arizonensis has a negative impact on deer mouse survival and thus may limit deer mouse populations. Although actual survival is difficult to measure under field conditions, length of residency in a population has frequently been used as a measure of survivorship (e.g. Boonstra et al. 1980, Lehmann 1992). First, we predicted that animals which are infected when they first enter a population are less likely to become residents than animals which are uninfected when they enter a population. Second, animals that do not reappear in spring (and presumably die over the course of 76 the winter) were infected more frequently the previous year than animals that do reappear in spring. Third, E. arizonensis infections are associated with weight loss in deer mice.
Methods
Observations
A 10 x 12 trapping grid was established in MacDonald
Forest (15 km NW Corvallis, OR) in March 1990. Sherman live- traps were spaced 10 m apart in 1990. Traps were spaced 15 m apart in 1991 and 1992 to increase sample size without increasing trapping effort. Traps were set for 2-3 consecutive nights at 2-3 week intervals from 15 April 17
Nov, 1990, 30 March 17 Nov, 1991 and 7 March 12 July, 1992. All deer mice were either ear-tagged or toe-clipped for individual identification at first capture. All animals were released each morning at their point of capture.
Animals were sexed, weighed and adult females were assessed for reproductive condition. A second group of animals was trapped in MacDonald forest using the same methods as above. However, after collection these animals were killed for use in another study. Data from these animals were combined with the first capture of each mark-recapture animal in analyses concerning the effect of E. arizonensis on deer mouse weight. 77
Feces were collected from traps, weighed and incubated in K2Cr2O7 solution (2.5% w/v) at room temperature for one week. This procedure facilitates the development of oocysts into the infective or sporulated stage and is necessary for identification to species. Soiled traps were washed before they were returned to the field.
Fecal samples were macerated and examined by Sheather's sugar flotation (Ash & Orihel 1987). The number of oocysts per gram in fecal samples was estimated by either counting all oocysts present or averaging counts from 10 representative microscope fields and extrapolating to the entire sample. Thus, counts were estimated as levels 0-6 where 0 = no oocysts detected, 1 = < 10 oocysts/gram, 2 =
10-99 oocysts/gram, 3 = 100-999 oocysts/gram, etc. For analyses of over-winter survival, we distinguished new infections from continued infections. If an animal which was shedding < 1000 oocysts/gram had been shedding oocysts during the previous trapping session, the second infection was assumed to be a continuation of the first (Chapter 2).
Experiments
Eimeria arizonensis Cultures
The E. arizonensis oocysts used in these experiments were obtained from the feces of two deer mice from the
MacDonald Forest population in Spring 1990. The oocysts were stored at 4°C and fresh oocysts were generated periodically 78
using laboratory born, Eimeria free deer mice which also originated from the MacDonald forest population. The oocysts
used in the experiments were generated 6 days 4 months before the initial inoculations, sporulated at room
temperature and refrigerated until used.
Field Experiment
The animals from the mark-recapture population were
used after the observational study had ended. In addition, a
second mark-recapture grid was established approx. 1 km from
the first. The protocol for the field experiment is
summarized in Table 4.1. Animals were lightly anaesthetized and inoculated by stomach intubation with 20,000 sporulated oocysts 2 times at 3-week intervals (Oct. and Nov. 1992) or
once (Nov. 1992). Animals were released within 1 hr of
inoculations and were free-living throughout the experiment. Traps were set overnight on days 5 or 6 post-inoculation
([PI] these are the days of peak oocyst output for E.
arizonensis, Chapter 2) and animals were released the following morning. Feces were collected out of traps,
weighed to the nearest 0.01 g and examined by MacMaster's technique (Dunn & Keymer 1986) to verify success of
infections. 79
Table 4.1. Protocol for field experiment.
Inoculum (N) Event Days PI*
Treatment Control
Inoculation 0 2 x 104 (9) water (11)
Feces collection 5 or 6
Challenge 20 or 21 2 x 104 (9) 2 x 104 (12)
Feces collection 25 or 26 (5-6 PC)**
* PI = post-inoculation ** PC = post-challenge 80
Enclosure Experiment
Free-living deer mice were live-trapped and placed in
4, 50 m x 50 m (2 treatment and 2 control enclosures) field enclosures in the fall of 1993. There were 100 live-traps
(10 x 10 grid with 5 m spacing) in each enclosure and each enclosure was surrounded by a 1 m high fence above ground and a 0.5 m fence below ground. Thus, animals were exposed to aerial but not terrestrial predators and could not disperse from the enclosures. The dominant vegetation was alfalfa; deer mice were not given water or additional food but were provided with nest boxes (20/enclosure). In this experiment animals in treatment enclosures were inoculated with 5 x 105 oocysts and animals in control enclosures were inoculated with water 1 3 times at 5-week intervals. The majority of animals were first inoculated in Sept. 1993, given a second infection in Oct. 1993 and a third infection in Nov. 1993 at 5 week intervals. A second group of animals was first inoculated in Oct. 1993 and given a second infection in Nov. Animals were trapped as in field studies and fecal samples were collected directly from the animals. Fecal samples were processed as in the field experiment.
Statistical Analyses
Table 4.2 shows the predictions, protocols which address them and statistical methods for each. In the analysis of prediction 1 (recruitment), we used a three-way 81
Table 4.2. Predictions and experimental protocols. Predictions 1 & 2 are from the hypothesis that Eimeria arizonensis negatively affects deer mouse survival by negatively affecting recruitment (21) and over-winter survival (P2). P3 predicts that deer mice will lose weight during infections. a = 3-way X2 contingency analysis, b = Mann-Whitney U test, c= Paired t-test, d= Student's t-test and e = ANOVA.
Predictions Study
1 2 3
Field Observations Xe Xb Xc
Field Experiment Xd
Enclosure Experiment Xe Xd 82
Chi-square analysis (factors = sex, infection status and year) to factor out the effect of year on infection rate.
Because there was no evidence of variability within treatments, we pooled data within treatments for analyses of weight change. In tests of all predictions, 1-tailed probabilities are given unless stated otherwise and alpha of
0.05 was considered significant.
Results
Prediction 1: Establishment of Residency
Deer mice were considered to be residents if they were present in at least four trapping sessions (a minimum of 6 weeks). Males that became residents in the population were infected during their first captures as frequently as males that did not become residents (Table 4.3; X2=1.841, df=2, N=79, P>0.2). In contrast, females that were infected when they first appeared, were significantly less likely to become residents than females that were not infected (Table
4.3; X2=3.566, df=2, N=63, P<0.05) .
Prediction 2: Over-Winter Survival
Field Observations
Only residents (see above) which were present in one of the last two trapping sessions of each year were included in this analysis. Males that remained in the population after 83
Table 4.3. E. arizonensis and deer mouse recruitment. Does the frequency of infection predict the likelihood of establishing residency? Values given are percent (sample size) of animals infected in at least one of first two captures. * Females which were infected with E. arizonensis in at least one of their first two captures were significantly less likely to become residents than females which were uninfected during their first two captures (3-way X2 contingency analysis, P < 0.05).
Became residents? Males Females* Yes No Yes No
Year:
1990 32 (22) 10 (19) 19 (16) 40 (10)
1991 69 (13) 60 (10) 21 (14) 50 (8)
1992 73 (11) 75 (4) 54 (11) 75 (4)
Total 52 (46) 33 (33) 29 (41) 50 (22) 84 the winter had been infected significantly less frequently the previous year than males that did not remain in the population. This pattern was clear in the spring of both 1991 and 1992 (Figure 4.1 A; 1990-91: Mann-Whitney U=36.5,
N=14, P<0.05; 1991-92: Mann-Whitney U=12, N=7, P>0.02).
Females that remained in the population after the winter of
1990 were not infected less frequently than females that did not return (Figure 4.1 B; Mann-Whitney U=17, N=12, P>0.2).
There were not enough resident females present in the population during the last two trapping sessions of 1991 to conduct an appropriate statistical analysis (N=4). However, the mean proportion of infections is presented in Figure 4.1
B for comparison.
Enclosure Experiments
As in field observations, males in treatment populations were significantly less likely to survive the winter than males in the control populations (ANOVA: F = 40.65, P < 0.025). There were no significant differences between control and treatment survival for females (F =
0.004; Figure 4.2). 85 0.6
0.5
0.4
0.3
0.2
0.1 H
0.0 0.6 b
0.5
0.4
0.3
0.2
0.1
0.0 MALES FEMALES
Figure 4.1. Over-winter survival of free-living deer mice. Proportion (SE) of captures free-living deer mice were infected with naturally-occurring Eimeria arizonensis in 1990 (a) or 1991 (b) vs whether they survived the following winter (open bars represent animals which survived, shaded bars represent animals which did not survive). Males which did not survive the winter had been infected significantly more frequently (* P < 0.05) than males which survived in both years. There were no significant differences for females in either year. Numbers in columns are sample sizes. 86
I CONTROL POPULATIONS TREATMENT POPULATIONS
100
80 To E
(/) 60
4a.
szlj 40 O
2 a) 20 0_
MALES FEMALES
Figure 4.2. Over-winter survival of enclosure deer mice. Percent of population which survived the winter in control and treatment enclosure populations. Control deer mice were sham inoculated and treatment deer mice were inoculated with 5 x 105 sporulated oocysts 1-3 times in the fall. Treatment males were significantly less likely to survive the winter than control males (* P < 0.05) but there were no significant differences in female survival. 87
Prediction 3: Weight Changes
Field Observations
There were no significant differences between infected and uninfected adults. However, infected juveniles were significantly heavier than uninfected juveniles (Table 4.4).
Field Experiment
Only males were used in this experiment. There were no significant differences in weight change from inoculation to
Day 5/6 PI between control and treatment animals in Oct (t =
-0.49, df = 20) or in Nov (t = -1.50, df = 21) 1992 (Table
4.4) .
Enclosure Experiments
There were no significant differences in weight change for adult males or for non-reproductive adult females in the enclosure experiment whether weights were analyzed separately or pooled by month (ie., ignoring the number of times animals had been inoculated; Figure 4.3 A,B). Other age/sex classes were not present in sufficient numbers to warrant analysis. 88 Table 4.4. Weight change in free-living deer mice. Weights (g) of infected and uninfected deer mice in field observations and experiments: mean + SE. Numbers in parentheses are # of months (observations) and # of animals
(experiments) . * P < 0.025, Mann-Whitney U Test. Observations: Data were pooled by month within each group and the mean weight of infected and uninfected animals was compared by paired t-tests for months in which both infected and uninfected animals were present. This was done to correct for seasonal weight changes. Experiments: Weight change from Day 0 Day 5 post inoculation (PI) of free-living deer mice inoculated with water (control) or 2 x 105 sporulated Eimeria arizonensis oocysts (treatment) .
Infected Uninfected
Observations
Juvenile* (9) 11.8+0.5 12.8+0.6
Adult male (15) 17.0+0.3 16.9+0.4
Adult female:
Pregnant (6) 20.7+0.7 18.3+1.4
Lactating (5) 17.8+0.6 16.5+0.7
Non-reproductive (6) 14.9+0.6 15.1+0.7
Total females (17) 17.8+0.7 16.6+0.6
Experiments (Adult males only)
October 1992 -0.15+0.4 (10) -0.37+0.2 (12)
November 1992 -2.18+0.4 (19) -0.75+0.6 (4) 89
1 Control Animals IIIN Treatment Animals
I 1 1 Sept, 0 Oct, 0 Oct, 1 Nov, 1 Nov, 2
Sept, 0 Oct, 1 Nov, 3 Month and Number of Previous Inoculations
Figure 4.3. Weight change of enclosure deer mice. Weight change was measured from from Day 0 Day 5 post-inoculation (PI) in animals inoculated with water (control) or 5 x 105 sporulated Eimeria arizonensis oocysts (treatment) 1-3 times. There were no significant differences in weight change for either males (a) or females (b) in any month and regardless of the number of times inoculated. Numbers at the base of columns are sample sizes. Bars are standard errors. 90
Discussion
We tested three predictions of the hypothesis that E. arizonensis has a negative impact on deer mouse survival. We found that, as predicted, the frequency of infection was negatively associated with two measures of survivorship, establishment of residency (Prediction 1) and overwinter survival (Prediction 2). However, males and females were not affected in the same way. Females infected with naturally- occurring E. arizonensis when they first appeared in the population were less likely to become residents than females that were uninfected, but there was no similar trend for males. On the other hand, once a male became a resident, the frequency with which he was infected over the remainder of the year was negatively associated with overwinter survival both in naturally-occurring and experimental infections. In this case, no similar trend was seen for females. In addition, the fact that the differences in over-winter survival are significant in all years/studies despite relatively small samples, suggests that the effect is especially strong. Although it is common for males and females of a host species to have different rates of infection (Alexander &
Stimson 1988, Bundy 1988), it is unclear why E. arizonensis would affect different aspects of deer mouse life history.
In the study of recruitment, females which did not persist were infected as frequently as males, whether the males 91 persisted or not. Females which did persist were infected less frequently than other animals. This pattern suggests that female immigrants may be more affected by infections than male immigrants. In contrast, in observations of overwinter survival, males which did not reappear in spring were infected more frequently than all other animals.
Although males and females in the enclosure experiment were inoculated with equal frequency, it is not known how often they became reinfected between Nov. (the last experimental inoculation) and March (collection). Thus, the probability of surviving the winter may be a function of the number of times animals were infected. Several studies have found negative associations between host survivorship and parasites in naturally- occurring infections (Arnqvist & Maki 1990, Boonstra et al.
1980, Burgett et al. 1990, Hudson et al. 1992a, Gulland &
Fox 1992, Lehmann 1992, Ross et al. 1989 and Royce & Rossignol 1990). However, other observational studies did not find any correlation between parasitism and host survivorship (Goater et al. 1989, Johnson et al. 1991, Moss et al. 1990). Keith et al. (1986) found evidence that only one of six parasites examined affected snowshoe hare (Lepus americanus) survival. Only a few studies have used controlled field manipulations to directly test the effect of parasites on host survival. In some cases, experiments have confirmed trends seen in observations (Hudson et al.
1992a, Lehman 1992, Trout et al. 1992); in at least one 92 case, results of experiments have contradicted observations
(Samson 1987). Because E. arizonensis does not cause obvious pathology in deer mice, we also examined weight loss as an indication of a subtle mechanism through which E. arizonensis might influence deer mouse survival (Prediction 3). Results of naturally-occurring and experimental infections did not support the prediction that E. arizonensis infections cause weight loss in deer mice. Our results indicated no significant weight loss associated with infections in adults. However, naturally-infected, free-ranging juveniles weighed significantly more than uninfected juveniles, the opposite of prediction 3. Although juveniles may have gained more weight while infected, it may also be that heavier juveniles were older. Older juveniles may be more active outside the nest, increasing their encounter rate with E. arizonensis; they may also have less maternal immunity, leaving them more susceptible to infections. Unfortunately, we were unable to include juveniles in either experiment. A few other authors have examined weight loss as a possible mechanism for parasite induced reduction in survival and/or reproductive success in birds and mammals
(Booth et al. 1993, Howe 1992, Keith et al. 1986, Lehmann
1992, Moore & Bell 1983). Of these, only Booth et al.
(1993), studying rock doves (Columba livia) and Keith et al.
(1986), studying snowshoe hares (L. americanus) found significant negative associations between weight change and 93 parasitism. Other authors have examined changes in metabolic rate (Booth et al. 1993, Munger & Karasov 1989), organ pathology (Gellar & Christian 1982, Watkins et al. 1991) and blood abnormalities (Wiger 1977) with variable results.
None of the factors discussed above addresses the direct mechanism of host mortality. Among other things, parasitized animals may succumb to starvation, cannibalism and predation (Hudson et al. 1992b, Temple 1987, Wiger
1977). In our study, it is unclear which of these factors was important. In winter deer mice typically weighed 20-25% less than at other times of the year (pers obs.) indicating possible nutritional stress. However, some of this weight loss is probably due to seasonal changes in reproductive condition (Millar 1989, pers obs). Predators were observed both in the field (short-tailed weasels, Mustela erminea) and near enclosures (falcons and hawks). It seems likely that parasitized deer mice were less able to avoid predation than unparasitized animals due to poorer general condition
(Wiger 1977, Scott 1988). In addition, adverse weather has been shown to compound the negative affect of parasites
(Lope et al. 1993). It is becoming increasingly clear that a wide variety of parasites (including arthropods, helminths and protozoans) negatively affect survival of free-living hosts.
In addition, several observational and experimental studies have documented negative affects of parasites on host reproduction (Durnin 1993, Lafferty 1993, Moller 1993, 94 Schall 1983). In some cases, the effects appear to be density dependent (Scott 1988). Thus, there is clear empirical evidence that parasites in natural populations have the potential to affect host abundance. Field experiments at the level of the host population are needed to determine whether the effects on reproduction and survival act in a compensatory manner or to regulate host density. 95
Chapter 5
Variable Levels of Immunity to E. arizonensis 96 Abstract
Acquired immunity to parasites may affect both host and parasite population dynamics. Although immunity has been studied experimentally in laboratory-reared hosts, less attention has focused on free-living animals. I examined acquired immunity of free-living deer mice (Peromyscus maniculatus) to naturally-occurring and experimental infections of Eimeria arizonensis (Protozoa: Coccidia). In a mark-recapture study, I found evidence of complete immunity to natural infections in only 1 of 3 years and evidence of partial immunity in 2 of 3 years. I subsequently examined immunity to experimental infections by giving laboratory- reared, free-living and enclosure populations of deer mice 2 or 3 consecutive E. arizonensis infections. Greater than 90%
(14 of 15) of laboratory-reared animals developed immunity after only 1 exposure, confirming that E. arizonensis is immunogenic. However, animals living under natural/semi natural conditions developed significantly less immunity in 2 of 3 experiments. These observational and experimental results suggest that immunocompetence of free-living deer mice to E. arizonensis may be quite variable. These results suggest that care should be taken when generalizing from laboratory to field situations in studies of acquired immunity. 97
Introduction
The ecology of host-parasite relationships differs from that of other symbiotic organisms in that hosts may have
varying levels of immunity to their parasites (Quinnell &
Keymer 1990). In vertebrate hosts, acquired immunity may
play a large role in determining parasite and (indirectly) host population dynamics (Anderson 1991). Two different
types of parasites are recognized: 1) microparasites
(viruses and bacteria) go through multiple reproductive
cycles within the host and usually induce complete immunity
after only one infection (Anderson & May 1979); 2) macroparasites (helminths and arthropods) shed eggs or
larvae from the host and larvae must infect a new host before reproduction occurs; acquired immunity is often
partial and the degree of immunity may be dose (density)
dependent (May & Anderson 1979). In addition, acquired
immunity to macroparasites may be lost if enough time passes
between infections. Although protozoans are generally categorized as microparasites because of their mode of reproduction, they are often more similar to macroparasites with respect to acquired immunity (Healing & Nowell 1985,
Quinnell & Keymer 1990). Acquired immunity to macroparasites and protozoans has
been demonstrated experimentally in many host-parasite
systems in the laboratory (Gregory, Keymer & Clarke 1990,
Higgs & Nowell 1988, Lillehoj 1988, Wakelin 1978, 1987, 98 Wassom et al. 1986) and under semi-natural conditions
(Quinnell 1992). However, it is unclear whether these
results can be generalized to field situations. Laboratory
and semi-natural conditions may differ greatly from field
conditions with respect to temperature, resource
availability and social stress. All of these factors may
reduce the ability of the host to display an immune
response. In addition, genetic variability of host
laboratory stocks may be lower than their free-living
counterparts. Thus, while hosts may be capable of developing
immunity under ideal laboratory conditions, immunocompetence
may be much more variable in the field.
There are several ways to detect acquired immunity in field populations. Convex age-intensity curves can indicate
the presence of acquired immunity among naturally infected,
free-living hosts (Anderson & May 1985). In a convex age-
intensity curve, parasite intensity or burden is low in
young hosts because of low exposure, increases in young
adults as exposure increases and drops again as animals acquire immunity. In protozoans, age-prevalence data (proportion of animals infected) are often used instead of age-intensity data (Aron & May 1982). A few studies have
found evidence of the presence of acquired immunity using these methods (Fulford et al. 1992, Gregory, Montgomery &
Montgomery 1992, Woolhouse et al. 1991). However, the same patterns of infection can also be explained by other processes such as parasite-induced mortality or behavioral 99 differences among hosts of different ages leading to differential susceptibility among age classes (Bundy 1988,
Fulford et. al 1992, Pacala & Dobson 1988).
Another way to detect acquired immunity in naturally infected hosts is to follow individual hosts over time using mark-recapture techniques. If an animal becomes immune it should not become reinfected after losing its initial infection (complete immunity), or subsequent infections should have a lower intensity (partial immunity). However, behavioral differences could also lead to observed differences in parasitism when following individual hosts.
Ideally, both age-intensity/prevalence and mark-recapture studies should be accompanied by field experiments which directly test whether free-living animals exhibit immunity. Eimerians (Apicomplexa: Coccidia) are excellent parasites to use in observational and experimental field studies of acquired immunity because large numbers (often >
106) of oocysts are shed in the feces and infections are of short duration (< 20 days) (Long 1982), thus multiple infections can be followed non-invasively by examination of feces. In addition, experimental infections of known doses can be given. Laboratory studies indicate that even small numbers of eimerians are often enough to induce complete immunity after as few as four exposures and immunity can be detected as a drop in fecal oocyst output (Higgs & Nowell
1988, Lillehoj 1988, Rose 1987). 100
I examined acquired immunity to natural and experimental Eimeria arizonensis infections in deer mice
(Peromyscus maniculatus). I hypothesized that free-living deer mice become immune to natural infections. If animals develop complete immunity, they should not become infected repeatedly (Prediction 1). If animals become partially immune, there should be a decrease in oocyst output from one infection to the next (Prediction 2). Observations in a mark-recapture population in three years suggested that the proportion of hosts developing complete and partial immunity was highly variable among years. Several alternative hypotheses could account for this variability: 1) animals may be immunocompromised under some conditions and observed immunity was due to other factors (see above); 2) E. arizonensis may not be immunogenic; 3) animals were immune but I was unable to detect immunity with the methods I employed. I conducted a series of laboratory and field experiments to differentiate between the following predictions from the hypotheses: 1) if variability observed among naturally infected animals is real, there should be variability among different experimental conditions; 2) if E. arizonensis is not immunogenic, none of the experimentally infected animals should become immune and 3) if I failed to detect immunity to natural infections, animals should develop immunity under all experimental conditions. 101 Methods: Field Observations
A 10 x 12 trapping grid was established in MacDonald Forest (15 km NW Corvallis, OR) in March 1990. Sherman live- traps were spaced 10 m apart in 1990 and 15 m apart thereafter. Traps were set for 2-3 consecutive nights at 2-3 week intervals from 15 April 17 Nov, 1990, 30 March 17
Nov, 1991 and 7 March 12 July, 1992. All deer mice were either ear-tagged or toe-clipped for individual identification at first capture. Animals were released each morning at their point of capture.
Feces were collected from traps, weighed and incubated in K2Cr207 solution (2.5% w/v) at room temperature for one week (this procedure facilitates the development of oocysts into the infective stage and is necessary for identification to species). Soiled traps were washed before they were returned to the field.
Fecal samples were macerated and examined by Sheather's sugar flotation (Ash & Orihel 1987). The number of oocysts per gram in fecal samples was estimated by either counting all oocysts present or averaging counts from 10 representative microscope fields and extrapolating to the entire sample. Thus, counts were estimated as levels 0-6 where 0 = no oocysts detected, 1 = < 10 oocysts/gram, 2 =
10-99 oocysts/gram, 3 = 100-999 oocysts/gram, etc. An attempt was made to distinguish new infections from continued infections: if an animal which was shedding < 1000 102 oocysts/gram had been shedding oocysts during the previous trapping session, the second infection was assumed to be a continuation of the first. I previously determined that E. arizonensis oocysts are first shed in deer mouse feces on day 4 post-inoculation and are shed for 6-8 days, Chapter 2.
Results: Field Observations
Prediction 1, Complete Immunity: animals should not become infected repeatedly
Only animals which were captured > 5 times (ie, at least 11 weeks in the population) were used in this analysis so that they would have adequate opportunities to become reinfected. Of the 198 deer mice captured over the course of the study, 42 individual deer mice were caught > 5 times in
3 years and 8 were caught > 5 times in two years. Thus, the total sample size for this analysis was 50 animal-years.
Figure 5.1 shows that 43-100% of deer mice became infected repeatedly in all years. However, the modal number of times animals became infected in 1990 was 1, indicating that complete immunity may have occurred. In 1991 and 1992, the modal number of times animals became infected was 3 and 4, respectively. Only 3 of 29 animals became infected < 2 times in these two years combined. 103
1990
1991
1992
0 1 3 4 5 6 7
Number of Times Infected
Figure 5.1. Number of times individual, free-living deer mice were infected with naturally occurring Eimeria arizonensis each year. Animals were live-trapped, marked and released. Animals included in this analysis were in the population at least 11 weeks. 104
Prediction 2, Partial Immunity
All animals which were infected more than once were
used in this analysis. Figure 5.2 shows the median change in
oocyst output from 1 infection to the next (0 = no change, positive numbers indicate an increase in oocyst output and
negative numbers indicate a decrease). Animals appear to have developed partial immunity to E. arizonensis after only
one infection in 1990 and after 2 infections in 1991. There
was no evidence of partial immunity in 1992.
Methods: Immunity Experiments
In Chapter 2, we demonstrated that laboratory reared
animals can develop immunity to E. arizonensis. The results
of this experiment are summarized here as a standard to
determine whether the other groups developed immunity (lab reared animals will be referred to as the first group). The
second group of animals tested consisted of free-living deer
mice and the third group consisted of animals in large (0.25
ha) field enclosures. The E. arizonensis oocysts used in these experiments
were obtained from the feces of two deer mice from the MacDonald Forest population in Spring 1990. The oocysts were
stored at 4°C and fresh oocysts were generated periodically
using laboratory born, Eimeria free deer mice which also
originated from the MacDonald forest population. The oocysts 105
Infec 1-2 /// \\X\ Infec 2-3 AL A Infec 3-4
1.0
0.5
0.0 C
a) 0.5 1 C C U _E
0 1 .0 t
a) 75
1990 1991 1992
Figure 5.2. Change in oocyst output with successive infections. Median change in oocyst output from one infection to the next in naturally infected, free-living deer mice. If animals developed partial immunity, there should be a decrease (negative change) in oocyst output with each successive infection. If no immunity develops, there will be no change or an increase in oocyst output (positive change). 106 used in the experiments were generated 6 days 4 months before the initial inoculations.
Laboratory Experiment
Laboratory animals originated from MacDonald forest and had been in captivity 3-5 generations. Table 5.1 shows the protocol for the laboratory experiment. Deer mice were lightly anaesthetized with halothane and inoculated with one of three levels (103, 104, 105) of E. arizonensis by stomach intubation. Animals were housed individually in cages with wire-mesh bottoms. Feces were collected for 20 days post inoculation (PI) in pans suspended under the cages. Animals were inoculated a second time (given a challenge infection) on day 21 PI with the same number of oocysts as in the first inoculation. Feces were again collected for 20 days post- challenge (PC). Fecal samples were examined by McMaster's technique (Dunn & Keymer 1986) to determine total number of oocysts shed over the course of the infection.
Field Experiment
The animals from the mark-recapture population were used after the observational study had ended. In addition, to increase sample size, a second mark-recapture grid was established approx. 1 km from the first. The protocol for the field experiment is shown in Table 5.1. Animals were lightly anaesthetized and inoculated by stomach intubation 107
Table 5.1. Protocol for laboratory and field experiments.
Number of oocysts Event Days PI*
Treatment Control
Laboratory (N = 5 in all cases except @ [N = 4])
Inoculation 0 103, 109, 105 Feces collection 20 Days PI
Challenge 21 103, 104, 105 103, 104, 105 Feces collection 20 Days PC"
Field (sample size in parentheses)
Inoculation 0 2 x 104 (9) water (11) Feces collection 5 or 6
Challenge 20 or 21 2 x 104 (9) 2 x 104 (12) Feces collection 25 or 26 (5-6 PC) * PI = post-inoculation ** PC = post-challenge 108
with either 20,000 oocysts (treatment and naive control) or
water (control). Animals were free-living throughout the experiment. Traps were set overnight on days 5 or 6 PI/PC
(these are the days of peak oocyst output for E.
arizonensis, Chapter 2) and animals were released the
following morning. Feces were collected out of traps,
weighed to the nearest 0.01 g and processed as in the laboratory experiment. For this experiment and the enclosure
experiment, the total number of oocysts shed per gram of
feces was calculated.
Enclosure Experiment
Free-living deer mice were live-trapped and placed in
4, 50 m x 50 m (2 treatment and 2 control enclosures) field
enclosures in the fall of 1993. There were 100 live-traps
(10 x 10 grid with 5 m spacing) in each enclosure and each
was surrounded by a 1 m high fence above ground and a 0.5 m
fence below ground. Thus, animals were exposed to aerial but not terrestrial predators and could not disperse out of the enclosures. The dominant vegetation was alfalfa; deer mice were not given water or additional food but were provided with nest boxes (20/enclosure). All treatment animals were given 50,000 oocysts at each inoculation and all controls were given water.
The majority of animals were first inoculated in Sept.
1993, given a challenge infection in Oct. 1993 and a second 109 challenge infection in Nov. 1993 at 5 week intervals.
However, a second group of animals was first inoculated in Oct. 1993 and given a challenge infection in Nov. Thus, the second group provides a replicate for the first group. Table
5.2 shows the sample sizes and timing of the inoculations for this experiment. Because sample sizes in each enclosure were small, data were pooled within treatments. Animals were trapped and fecal samples were collected as in the field experiment.
Statistical Analyses
The degree of change in oocyst output for each treatment animal was calculated as the percent change in output compared to the mean oocyst output of naive animals inoculated at the same time and with the same number of oocysts (eg., oocyst output of an enclosure animal challenged with 50,000 oocysts in Nov 1993 was compared to the mean output of naive animals inoculated with 50,000 oocysts for the first time in Nov 1993). This approach was necessary to have a common basis for comparing the results of the three experiments. The percent change in oocyst output was then compared between the laboratory experiment and the field/enclosure experiments by Mann-Whitney U test.
The proportion of animals that developed immunity (see below) was compared with Fischer's exact test. Alpha levels were corrected for the number of comparisons made by the 110
Table 5.2. Sample sizes in enclosure experiments. All treatment animals were inoculated with 50,000 oocysts, all control animals were inoculated with water and inoculations were 5 weeks apart. * 1 = initial inoculation, 2 = challenge inoculation, 3 = 2nd challenge. ** Because so few naive animals remained in treatment enclosures in Nov., 8 of 9 naive treatment animals first inoculated in Nov. were laboratory reared.
Number of Times Inoculated*
Month of 1st Treatment Control Inoculation
1 2 3 1 2 3
Sept 32 15 12 23 16 11
Oct 10 6 10 7
Nov 9 0 111
Bonferroni method (Zar 1984). Alpha oh 0.05 was considered statistically significant in all cases.
Results: Immunity Experiments
Effectiveness of Inoculations
Animals inoculated for the first time (naive animals) in all experiments shed large numbers of oocysts in all cases (Figures 5.3 A,B). Very few sham-inoculated field/enclosure animals shed oocysts and of those that did, only one shed at levels close to treatment animals (Table
5.3). Thus, field/enclosure inoculations were successful and it seems highly unlikely that treatment animals shed oocysts during the experiment as the result of natural infections.
Laboratory Experiment
All 15 animals showed a large decrease in oocyst output compared with that of controls. However, 14/15 deer mice showed a > 90% (Mean = 97.1%) decrease. The 15th animal showed only a 66 % decrease. Thus, for purposes of comparison, I considered animals showing a > 90% decrease in oocyst output from one infection to the next to be immune. 112
(1) 30 A 0 10' oocysts
25 10 oocysts tit:2 10 oocysts
20
15
10
0 0 Treatment Control
C VA Sept NN Oct ZEI Nov
Cl) 0) U 0) LL
O
0
(J) (f)
L2.) a Fld Exp Encl Exp
Figure 5.3. Experimental infections in naive animals. The number of Eimeria arizonensis oocysts shed by naive (previously uninfected) animals at their first experimental inoculation (+SE). A) results from the laboratory experiment; treatment deer mice were first inoculated three weeks prior to the first inoculation of control deer mice. B) results from field and enclosure experiments. Animals in the field experiment were given 2 x 104 oocysts; enclosure animals were given 5 x 104 oocysts. The field and enclosure experiments were conducted in 1992 and 1993, respectively. 113
Table 5.3. Infections among sham inoculated animals. Prevalence and intensity of infections among sham inoculated animals in field and enclosure experiments.
Prevalence: Intensity: Experiment/ number of animals oocysts/g feces Month shedding oocysts (total)
Field Exp 4 (12) < 100 (3 animals) > 106 (1 animal)
Enclosure Exp
Sept 2 (23) < 10
Oct 2 (26) < 10
Nov 1 (18) < 1000 114
Comparisons Between Laboratory, Field and Enclosure Results
Significantly fewer animals in the field population and in the first replicate of the enclosure population became immune than in the lab-reared population (Figure 5.4 A).
However, the difference between the laboratory reared animals and the enclosure animals was no longer significant after the second challenge infection was given to enclosure animals. Thus, these animals developed the same degree of immunity as their lab-reared counterparts, but only after 2 exposures. In addition, the second group of enclosure animals (initially inoculated in Oct 1993, 1 month after the initial inoculation of the first group) all developed immunity after only one inoculation. The results are similar if actual decrease in oocyst output is compared (Figure 5.4
B) .
Discussion
In naturally infected, free-living deer mice, there was evidence of complete immunity in 1 of 3 years and evidence of partial immunity in 2 of 3 years. In 1 year in which partial immunity was detected, it was not apparent until animals had been infected at least twice. Thus, it appears that immunity to naturally occurring infections was highly variable between years. My observational results could be explained by several hypotheses: 1) animals may be immunocompromised under some conditions; 2) E. arizonensis 115
100 - A
80
60
40
20
0 100 B
80 -1 -T.
60
40
20
Q) 0 ..,c>i,N ..,c;'(t, ,..,Ckce` ., (<\<<>/ 6c,`' <<:-or \` \c3 ' c;(
Figure 5.4. Immunity to experimental infections in the field. Percent of deer mice which became immune to Eimeria arizonensis (A) and percent decrease (+SE) in oocyst output (B) after 1 or 2 (enclosure, experiment only) challenge infections. Experiments were compared using results from the laboratory as a standard; (*) indicates values significantly different than laboratory results (P < 0.05). 116 may not be immunogenic; or 3) I was unable to detect immunity with the methods I employed. I conducted a series of laboratory and field experiments to differentiate between the hypotheses. The experimental results were consistent with hypothesis 1 but did not support hypotheses 2 and 3; the laboratory reared animals developed immunity to E. arizonensis after only one inoculation (not consistent with
H2). Additionally, in 2 of 3 field/enclosure experiments, animals did not develop the same degree of immunity as in the laboratory (not consistent with H3).
There are several reasons why free-living animals may frequently be immunocompromised compared with laboratory- reared animals. These can be grouped into two categories: 1) genetic (Wakelin & Blackwell 1988, Wassom et al. 1988) and
2) environmental factors (McCallum 1990). It is possible but unlikely that my laboratory population had a higher proportion of genetically "resistant" individuals. The laboratory colony originated from the same populations as the animals used in the field experiments, had only been in captivity 3-5 generations and was maintained as an outbred colony. It seems more likely that the differences seen between laboratory and field populations were due to factors in the host's environment. McCallum (1990) suggests that, in general, transient factors may play a larger role than genetics in determining predisposition to parasite infections in free-living animals. 117
At least one environmental factor would predispose free-living animals to be more likely to develop immunity: exposure to natural E. arizonensis infections prior to experimental inoculation. Despite this, in 2 of 3 experiments, field caught animals exhibited some decrease in immunocompetence with respect to their laboratory-reared counterparts. Many environmental factors would predispose free-living animals to be less likely to develop immunity than laboratory-reared animals including protein deficiency, (McMurray 1984, Slater 1988, Slater & Keymer 1986), very low levels of lipids and some trace elements (Chandra & Amorin
1992, Chandra & Dayton 1982), low night temperature (Noble
1966) and the presence of other concurrent parasitic infections (Cox 1987, Rose et al. 1994). Stress (e.g. due to behavioral interactions among conspecifics) can also lead to impaired immune function especially if it is of short duration. Chronic stress may actually enhance immune response if animals can adapt through behavioral or other means. If animals cannot adapt, chronic stress can also impair the immune response (reviewed in Griffin 1989). Although laboratory animals may be under chronic stress, the environmental conditions were very stable in my experiments. Animals were fed and given access to water ad libitum, housed individually and the room temperature was maintained at 24-27°C. They were also exposed to few other parasites (colony animals were observed to have fleas and a nematode). Thus, laboratory animals may have been able to 118 adapt to chronic stress. In contrast, field conditions were much more variable and animals may not have been able to adapt to stress. In addition, the field/enclosure experiments were conducted in the fall and conditions were harsh. Free-living animals tended to be lighter (11-18 g vs
17.5-24 g in the laboratory), indicating the possibility of nutritional deficiencies, temperature was highly variable and population densities were within the normal range for deer mice (pers obs). Finally, long-term monitoring of deer mice indicated that animals were frequently infected with at least eight other intestinal tract parasites (2 eimerians, 2 isosporans [family Eimeriidae], three pinworms [Nematoda] and a trematode). These factors may explain their apparent lowered ability to wage an immune response.
Although laboratory studies of acquired immunity are quite common, few studies have attempted to determine whether immunity seen in the laboratory also occurs in the field. While the results of some field studies of age- intensity and age-prevalence curves corroborated laboratory studies (Fulford et al. 1992, Gregory et al. 1992, Woolhouse et al. 1991), others have not (Behnke & Wakelin 1973,
Gregory 1992). In part, this may be due to the difficulty in aging field-caught animals (Gregory 1992). A few studies have documented variability in hematological parameters associated with acquired immunity in wild rodents
(Dobrowolska et al. 1974, Lochmiller et al. 1994). However, none of these studies have been experimental. Quinnell 119
(1992) found immunity in an enclosed population of wood mice
(Apodemus sylvaticus) experimentally infected with the nematode Heliqmosomoides polyqvrus. However, the enclosure used in this study was small (7 x 6 m), crowded (1 mouse/2 m2), there was no vegetation, predators were excluded, animals were given food and water ad libitum and they were given an anthelminthic at the beginning of the experiment.
Thus, while Quinnell (1992) came closer to a natural situation than a laboratory because animals were exposed to natural temperature variation, it may not have come close enough for animals to exhibit natural variation in immunocompetence. My field experiment represents the first experimental study of acquired immunity conducted on completely free-living animals. In addition, the enclosures used in the enclosure experiment mimicked a natural situation closely. Although the sample sizes of deer mice in my field/enclosure experiments were low and both were conducted in the fall, field observations of naturally- occurring infections suggest that some degree of
immunocompromise may be common in this system.
Most free-living vertebrates experience more environmental variability than their captive counterparts. Although laboratory studies often provide a much needed
starting point for studies of immunity in free-living
animals, the discrepancy between my laboratory and field
results suggests that care should be taken when generalizing to field situations. The role of acquired immunity to 120 parasites may have ramifications for the population dynamics of both parasites and hosts. Thus, further study on how environmental factors interact to mediate immunity in free- living animals is clearly needed. 121
Bibliography
Adjei, E.L., Barnes, A. and Lester, R.J.G. 1986. A method for estimating possible parasite-related host mortality, illustrated using data from Callitetrarhynchus qracilis (Cestoda: trypanorhyncha) in lizardfish (Saurida spp.). Para. 92: 227-243.
Adler, G.H., Telford, S.R., Wilson, M.L and Spielman, A. 1992. Vegetation structure influences the worm burden of immature Ixodes dammini on its main host, Peromyscus leucopus. Para. 105: 105-111.
Alexander, J. and Stimson, W.H. 1988. Sex hormones and the course of parasitic infection. Para. Today 4: 198-193.
Anderson, R.M. 1986. The population dynamics and epidemiology of intestinal nematode infections. Trans. Roy. Soc. Trop. Med. Hygiene 80:686-696.
Anderson, R.M. 1991. Populations and infectious diseases: ecology or epidemiology. J.Anim. Ecol. 60: 1-50.
Anderson, R.M. and May, R.M. 1979. Population Biology of infectious diseases: Part I. Nature 280: 361-367.
Anderson, R.M. and May, R.M. 1985. Herd immunity to helminth infection and implications for parasite control. Nature 315: 493-496. Arnett, E.B, Irby, L.R. and Cook, J.G. 1993. Sex- and age- specific lungworm infection in rocky mountain bighorn sheep during winter. J. Wildl. Dis. 29: 90-93. Arnqvist, G. and Maki, M. 1990. Infection rates and pathogenicity of trypanosomatid gut parasites in the water strider Gerris odontogaster. Oecologia 84: 194 198.
Aron, J.L. and May, R.M. 1982. The population dynamics of malaria. Population Dynamics of Infectious Diseases: Theory and Applications (Ed. by R.M. Anderson), pp. 139-179. Chapman & Hall, London. 368 p.
Ash, L.R. and Orihel, T.C. 1987. Parasites: A guide to laboratory procedures and identification. ASCP Press, Chicago. 328 pp. 122
Augustine, P. C. 1988. Eimeria arizonensis and E. meleagrimitis: effect of poult age on susceptibility to infection and development of immunity. Avian Dis. 32: 798-802.
Ball, S.J. and Lewis, D.C. 1984. Eimeria (Protozoa: Coccidia) in wild populations of some British rodents. J. Zool. Lond. 202:373-381.
Barker, S.C., Singleton, G.R. and Spratt, D.M. 1991. Can the nematode Capillaria hepatica regulate abundance in wild house mice? Results of enclosure experiments in southeastern Australia. Para. 103: 439-449.
Behnke, J.M. & Wakelin, D. 1973. The survival of Trichuris muris in wild populations of its natural host. Para. 67: 157-164.
Bennett, G.H, Caines, J.R. and Bishop, M.A. 1988. Influence of blood parasites in the body mass of passeriform birds. J. Wildl. Dis. 24: 339-343.
Boonstra, R. 1977. Effect of the parasite Wolfartia vigil on Microtus townsendii populations. Can. J. Zool. 55: 1057-1060.
Boonstra, R., Krebs, C.J., and Beacham, T.D. 1980. Impact of botfly parasitism on Microtus townsendii populations. Can. J. Zool. 58: 1683-1692.
Booth, D.T, Clayton, D.H. and Block, B.A. 1993. Experimental demonstration of the costs of parasitism in free- ranging hosts. Proc. Roy. Soc. London 253: 125-129.
Boughton, D. C. 1988. Circaduodial rhythms in avian Coccidia. Trans. Amer. Micr. Soc. 107: 329-344.
Brackett, S., and A. Bliznick. 1952. The reproductive potential of five species of coccidia of the chicken as demonstrated by oocyst production. J. Para. 38: 133 139.
Brodeur, J., and J. N. McNeil. 1989. Seasonal microhabitat selection by an endoparasitoid through adaptive modification of host behavior. Science 244: 226-228.
Bundy, D.A.P. 1988. Gender-dependent patterns of infection and disease. Para. Today 4: 186-189.
Burgett, D.M., Rossignol, P.A. and Kitprasert, C. 1990. A model of dispersion and regulation of brood mite (Tropilaelaps clarae) parasitism on the giant honey bee (Apis dorsata). Can. J. Zool. 68: 1423-1427. 123
Buskirk, J. van and Smith, D.C. 1991. Density-dependent population regulation in a salamander. Ecol. 72: 1747 1756.
Carleton, M.D. 1989. Systematics and Evolution. In: Advances in the study of Peromyscus (Rodentia), G.L. Kirkland & J.N. Layne (eds.). Texas Tech University Press, Lubbock, Texas, 7-141.
Catchpole, J., C. C. Norton, and L. P. Joyner. 1976. Experiments with defined multispecific coccidial infections in lambs. Para. 72: 137-147.
Chandra, R.K. and Amorin, S.A.D. 1992. Lipids and immunoregulation. Nutr. Res. 12: S137-S145.
Chandra, R.K. and Dayton, D.H. 1982. Trace element regulation of immunity and infection. Nutr. Res. 2: 721-733.
Clinchy, M. and Barker, I.K. 1994. Effects of parasitic infection on clutch size of lesser snow geese from a northern breeding colony. Can. J. Zool. 72: 541-544.
Combes, C. 1991. Ethological aspects of parasite transmission. Amer. Nat. 138: 866-880.
Cox, F.E.G. 1987. Interactions between the parasitic protozoa of small mammals. Mammal Rev. 17: 143-147.
Crofton, H.D. 1971. A quantitative approach to parasitism. Para. 62: 179-193.
Dobrowolska, A., Rewkiewicz-Dziarska, A., Szarska, I. and Gill, J. 1974. Seasonal changes in haematological parameters, level of serum proteins and glycoproteins, activity of the thyroid gland, suprarenals and kidneys in the common vole (Microtus arvalis). J. Inter. Cycle Res. 5: 347-354.
Dobson, A. P. 1988. The population biology of parasite- induced changes in host behavior. The Q. Rev. Biol. 63: 139-165. Dobson, A.P., Pacala, S.V., Roughgarden, J.D., Carper, E.R. and Harris, E.A. 1992. The parasites of Anolis lizards in the northern Lesser Antilles. Oecologia 91: 110-117.
Dunn, A., and A. Keymer. 1986. Factors affecting the reliability of the McMaster technique. J. Helminthol. 60: 260-262. 124
Durnin, J.V.G.A. 1993. Energy requirements in human pregnancy, in human nutrition and parasitic infection. Para. 107S: 169-175.
Elliot, L.F., Boyce, W.M., Clark, R.K. and Jessup, D.A. 1994. Geographic analysis of pathogen exposure in bighorn sheep (Ovis canadensis). J. Wildl. Dis. 30: 315-318. Fernando, M. A. 1982. Pathology and pathogenicity. In Biology of the Coccidia, P.L. Long (ed.). University Park Press, Baltimore, Maryland, p. 228-318.
Festa-Bianchett, M. 1989. Individual differences, parasites and the costs of reproduction for bighorn ewes (Ovis canadensis). J. Anim. Ecol. 58: 785-795.
Fitz-Coy, S. H., and S. A. Edgar. 1989. Eimeria mitis: immunogenicity and cross immunity of two isolates. Avian Dis. 33: 236-237.
Folstad, I., Nilssen, A.C., Halvorsen, 0. and Andersen, J. 1989. Why do male reindeer (Rangifer tarandus) have higher abundance of second and third instar larvae of Hypoderma tarandi than females? Oikos 55: 87-92.
Forbes, M., Weatherhead, P.J. and Bennett, G.F. 1994. Blood parasites of blue grouse: variation in prevalence and patterns of interspecific association. Oecologia 97: 520- 525. Ford, P.L., Duszynski, D.W. and McAllister, C.T. 1990. Coccidia (Apicomplexa) from heteromyid rodents in the southwestern United States, Baja California and northern Mexico with three new species from Chaetodipidus hispidus. J. Para. 76: 325-331. Fulford, A.J.C., Butterworth, A.E., Sturrock, R.F. and Ouma, J.H. 1992. On the use of age-intensity data to detect immunity to parasitic infections, with special reference to Schistosoma mansoni in Kenya. Para. 105: 219-227.
Gardner, S. L. 1991. Phyletic coevolution between subterranean rodents of the Genus Ctenomys (Rodentia, Hystricognathi) and nematodes of the Genus Paraspididera (Heterakoidea, Aspidoderidae) in the neotropics temporal and evolutionary implications. Zool. J. Linn. Soc. 102: 169-201.
Gellar, M.D. and Christian, J.J. 1982. Population dynamics, adrenocortical function and pathology in Microtus pennsylvanicus. J. Mammal. 63: 85-95. 125
Glicken, A. and Schwab, R.G. 1980. Modes of ectoparasite reinfestations of deer mice (Peromyscus maniculatus). J. Wildl. Dis. 16: 577-586. Goater, T.M., Shostak, A.W., Williams, J.A. and Esch, G.W. 1989. A mark-recapture study of trematode parasitism in overwintered Helisoma anceps (Pulmonata), with special reference to Halipequs occidualis (Hemiuridae). J. Para. 75: 553-560. Gordon, D.M. and Rau, M.E. 1982. Possible evidence for mortality induced by the parasite Apatemon qracilis in a population of brook sticklebacks (Culaea inconstans). Para. 84: 41-47. Graat, E.A.M., Henken, A.M., Ploeger, H.W., Noordhuizen, J.P.T.M. and Vertommen, M.H. 1994. Rate and course of sporulation of oocysts of Eimeria acervulina under different environmental conditions. Para. 108: 497-502.
Gregory, R.D. 1991. Parasite epidemiology and host population growth: Heliqmosomoides polyqyrus (Nematoda) in enclosed wood mouse populations. J. Anim. Ecol. 60: 805-821. Gregory, R.D. 1992. On the interpretation of host-parasite ecology: Heliqmosomoides polyqyrus (Nematoda) in wild wood mouse (Apodemus sylvaticus) populations. J. Zool. Lond. 226: 109-121. Gregory, R.D. and Keymer, A.E. 1989. The ecology of host- parasite interactions. Sci. Prog., Oxf. 73: 67-80.
Gregory, R.D, Keymer, A.E. and Clarke, J.R. 1990. Genetics, sex and exposure: the ecology of Heliqmosomoides polyqyrus (Nematoda) in the wood mouse. J. Anim. Ecol. 59: 363-378. Gregory, R.D., Montgomery, S.S.J., and Montgomery, W.I. 1992. Population biology of Heliqmosomoides polyqyrus in the wood mouse. J. Anim. Ecol. 61: 749-757.
Griffin, J.F.T. 1989. Stress and immunity: a unifying concept. Veter. Immun. Immunopath. 20: 263-312.
Gulland, F.M.D. 1992. The role of nematode parasites in Soay sheep (Ovis aries) mortality during a population crash. Para. 105: 493-503.
Gulland, F.M.D. and Fox, M. 1992. Epidemiology of nematode infections of Soay sheep (Ovis aries) on St. Kilda. Para. 105: 481-492. 126
Hall, P. R. 1935. The relation of the size of the infective dose to number of oocysts eliminated, duration of infection, and immunity in Eimeria miyairii infections in the white rat. Iowa State J. Sci. 9: 115-124. Harder, A., F. Wunderlich and P. Marinovski. 1992. Effects of testosterone on Heterakis spumosa infections in mice. Para. 105: 335-342.
Hart, B.L. 1990. Behavioral adaptations to pathogens and parasites: five strategies. Neuro. Biobeh. Rev. 14: 273-294. Haukisalmi. V., Henttonen, H. and Tenora, F. 1988. Population dynamics of common and rare helminths in cyclic vole populations. J. Anim. Ecol. 57: 807-825.
Hawking, F. 1975. Circadian and other rhythms of parasites. Adv. in Para. 13: 123-182. Healing, T.D. 1981. Infections with blood parasites in the small British rodents Apodemus sylvaticus, Clethrionomys glareolus and Microtus agrestis. Para. 83: 179-189. Healing, T.D. and Nowell, F. 1985. Diseases of parasites of woodland rodent populations. Symp. Zool. Soc. Lond. 55: 193-218. Herman, T.B. 1981. Capillaria hepatica (Nematoda) in insular populations of the deer mouse Peromyscus maniculatus: cannibalism or competition for carcasses. Can. J. Zool. 59: 776-784. Hestbeck, J.B. 1987. Multiple regulation state in populations of small mammals: a state-transition model. Am. Nat. 129: 520-532. Higgs, S., and F. Nowell. 1988. Laboratory studies with clones of Eimeria hunqaryensis, a parasite of the wood mouse Apodemus sylvaticus. Para. 97: 213-220.
Holmes, J.C., Hobbs, R.P. and Leong, T.S. 1977. Populations in perspective: community organization and regulation of parasite populations. In: Regulation of parasite populations (G.W. Esch, ed.). Academic Press, NY. 253 PP Holt, R. D. 1993. Infectious diseases of wildlife, in theory and in practice. TREE 8: 423-425. 127
Howe, F.P. 1992. Effects of Protocalliphora braueri (Diptera: Calliphoridae) parasitism and inclement weather on nestling sage thrashers. J. Wildl. Dis. 28: 141-143. Hudson, P.J. 1986. The effect of a parasitic nematode on the breeding production of red grouse. J. Anim. Ecol. 55: 85-92. Hudson, P.J., Dobson, A.P. and Newborn, D. 1992a. Do parasites make prey vulnerable to predation? Red grouse and parasites. J. Anim. Ecol. 61: 681-692.
Hudson, P.J., Newborn, D. and Dobson, A.P. 1992b. Regulation and stability of a free-living host-parasite system: Trichostrongylus tenius in the red grouse. I. Monitoring and parasite reduction experiments. J. Anim. Ecol. 61: 477-486.
Johnson, L.S., Eastman, M.D. and Kermott, L.H. 1991. Effect of ectoparasitism by larvae of the blowfly Protocalliphora parorum (Diptera: Calliphoridae) on nestling house wrens, Troglodytes aedon. Can. J. Zool. 69: 1441-1446. Kaufman, D.W. and Kaufman, G.A. 1989. Population Biology. In Advances in the study of Peromyscus (Rodentia), G. L. Kirkland, Jr., and J. N. Layne (eds.). Texas Tech University Press, Lubbock, Texas, p. 169-232.
Keith, I.M., Keith, L.B. and Cary, J.R. 1986. Parasitism in a declining population of snowshoe hares. J. Wildl. Dis. 22: 349-363.
Keith, L.B., Cary, J.R, Yuill, T.M. and Keith, I.M. 1985. Prevalence of helminths in a cyclic snowshoe hare population. J. Wildl. Dis. 21: 233-253. Keymer, A.E. and Dobson, A.P. 1987. The ecology of helminths in populations of small mammals. Mammal. Rev. 17: 105 116.
Keymer, A.E. and Read, A.F. 1991. Behavioral ecology: the impact of parsitism. In: Parasite-host associations: coexistence or conflict? (C.A. Toft, A. Aeschlimann and L. Bolis, eds.). Oxford Univ Press, Oxford.
Kheysin, Y.M. 1972. Life cycles of coccidia of domestic animals. Univ. Park Press, Baltimore, MD.
King, J.A. 1968. Biology of Peromyscus (Rodentia). Spec. Publ., Amer. Soc. Mammal., 2: 1-593. 128
Kirkland, G.L. and Layne, J.N. 1989. Advances in the study of Peromyscus (Rodentia). Texas Tech University Press, Lubbock, Texas.
Lafferty, K.D. 1993. Effects of parasitic castration on growth, reproduction and population dynamics of the marine snail Cerithidea californica. Marine Ecol. Prog. 96: 229-237.
Lafferty, K.D., Sammond, D.T. and Kuris, A.M. in press. Analysis of larval trematode communities. Ecol. Lehmann, T. 1992. Ectoparasite impacts on Gerbillus andersoni allenbyi under natural conditions. Para. 104: 479-488.
Lehmann, T. 1993. Ectoparasites: direct impact on host fitness. Para. Today 9: 8-13.
Levine, N.D. 1980. Weather and the ecology of bursate nematodes. Int. J. Biometeorol. 24: 341-346.
Levine, N. D. and Ivens, V. 1960. Eimeria and Tyzzeria (Protozoa, Eimeriidae) from deer mice (Peromyscus spp.) in Illinois. J. of Para. 46: 207-212.
Levine, N.D. and Ivens. V. 1989. The coccidian parasites of rodents. CRC Press, Inc., Boca Raton Florida, 227 p.
Lillehoj, H. S. 1988. Influence of inoculation dose, inoculation schedule, chicken age, and host genetics on disease susceptibility and development of resistance to Eimeria tenella infection. Avian Dis. 32: 437-444.
Lim, S.S.L. & Green, R.G. 1991. The relationship between parasite load, crawling behavior and growth rate of Macoma balthica (L.) (Mollusca, Pelecypoda) from Hudson Bay, Canada. Can. J. Zool. 68: 2202-2208.
Lochmiller, R.L., Vestey, M.R. & McMurray, S.T. 1994. Temporal variation in humoral and cell-mediated immune response in a Sigmodon hispidus population. Ecology 75: 236-245.
Long, P.L. 1982. The biology of the Coccidia. Univ. Park Press, Baltimore. 502 pp.
Lope, F. de, Gonzalez, G. and Moller, A.P. 1993. Increased detrimental effects of ectoparasites on their bird hosts during adverse environmental conditions. Oecologia 95:234-240. 129
Lope, F. de and Moller, A.P. 1993. Effects of ectoparasites on reproduction of their swallow hosts: a cost of being multi-brooded. Oikos 67: 557-562.
MacMillen, R.E. and Garland, T. 1989. Adaptive Physiology. In Advances in the study of Peromyscus (Rodentia), G. L. Kirkland, Jr., and J. N. Layne (eds.). Texas Tech University Press, Lubbock, Texas, p. 169-232.
Mappes, T., Mappes, J. and Kotiaho, J. 1994. Ectoparasites, nest site choice and breeding success in the pied flycatcher. Oecologia 98: 147-149.
Marquardt, W.C., Senger, C.M. and Seghetti, L. 1960. The effect of physical and chemical agents on the sporulation of Eimeria zurnii (Protozoa: Coccidia). J. Protozool. 7: 186-189. May, R.M. and Anderson, R.M 1979. Population biology of infectious diseases: part II. Nature 280: 455-461.
McCallum, H.I. 1990. Covariance in parasite burdens: the effect of predisposition to infection. Para. 100: 153 159.
McMahon, T.E. and Tash, J.C. 1988. Experimental analysis of the role of emigration in population regulation of desert pupfish. Ecol. 69: 1871-1883.
McMurray, D.N. 1984. Cell-mediated immunity in nutritional deficiency. Progress in Food and Nutrition Science 8: 193-228. McPhail, J.D. and Peacock, S.D. 1983. Some effects of the cestode (Schistocephalus solidus) on reproduction in the three-spine stickleback (Gasterosteus aculeatus): evolutionary aspects of a host-parasite interaction. Can. J. Zool. 61: 901-908.
Millar, J. S. 1989. Reproduction and development. In Advances in the study of Peromyscus (Rodentia), G. L. Kirkland, Jr., and J. N. Layne (eds.). Texas Tech University Press, Lubbock, Texas, p. 169-232. Miller, D.H and Getz, L.L. 1969. Botfly infections in a population of Peromyscus leucopus. J. Mammal. 50: 277 283.
Minchella, D.J. and Scott, M.E. 1991. Parasitism: a cryptic determinant of animal community structure. TREE 6: 250 254. 130
Moller, A.P. 1993. Ectoparasites enhance the cost of reproduction in their hosts. J. Anim. Ecol. 62: 309 322. Montgomery, S.S.J. and Montgomery, W.I. 1988. Cyclic and non-cyclic dynamics in populations of the helminth parasites of wood mice, Apodemus sylvaticus. J. Helminth. 62: 78-90.
Moore, J. 1987. Some roles of parasitic helminths in trophic interactions: A view from North America. Rev. Chilena His. Nat. 60: 159-179.
Moore, J. and Bell, D.H. 1983. Pathology of Plaqiorhynchus cylindraceus in the starling, Sturnus vulqaris. J. Para. 69: 387-390. Moore, J., Freehling, M. and N. J. Gotelli. 1994. Altered behavior in two species of blattid cockroaches infected with Moniliformes moniliformes (Acanthocephala). J. Para. 80: 220-223.
Moss, R. Trenholm, I.B., Watson, A. and Parr. R. 1990. Parasitism, predation and survival of hen red grouse Laqopus in spring. J. Anim. Ecol. 59: 631-642. Moss, R. Watson, A. Trenholm, I.B. and Parr, R. 1993. Caecal threadworms Trichostronqylus tenius in red grouse Laqopus laqopus: effects of weather and host density upon estimated worm burdens. Para. 107: 199-209. Mulvey, M., Aho, J.M. and Rhodes, 0.E., Jr. 1994. Parasitism and white-tailed deer: timing and components of female reproduction. Oikos 70: 177-182.
Munger, J.C. and Karasov, W.H. 1989. Sublethal parasites and host energy budgets: tapeworm infection n white-footed mice. Ecology 70: 904-921.
Munger, J.C. and Karasov, W.H. 1994. Costs of loot fly infection in white-footed mice: energy and mass flow. Can. J. Zool. 72: 166-173. Noble, G.A. 1966. Stress and parasitism. III. Reduced night temperature and the effect on pinworms of ground squirrels. Exp. Para. 18: 61-62.
Norris, K. Anwar, M. and Read, A. 1994. Reproductive effort influences the prevalence of haematozoan parasites in great tits. J. Anim. Ecol. 63: 601-610. 131
Pacala, S.W. and Dobson, A.P. 1988. The relation between the number of parasites/host and host age: population dynamic causes and maximum likelihood estimation. Para. 96: 197-210. Payne, J.A., Dunaway, P.B., Martin, G.D and Story, J.D. 1965. Effects of Cuterebra anqustifrons on plasma proteins of Peromyscus leucopus. J. Para. 51: 1004 1008. Poiani, A. 1993. Small clutch size as a posible adaptation against ectoparasitism: a comparastive analysis. Oikos: 455-462. Prestrud, P., Stuve, G. and Holt, G. 1993. The prevalence of Trichinella sp. in Arctic foxes (Alopex laqopus) in Svalbard. J. Wildl. Dis. 29: 337-340.
Quinnell, R.J. 1992. The population dynamics of Heligmosomoides polyqyrus in an enclosure population of wood mice. J. Anim. Ecol. 61: 669-679. Quinnell, R.J. and Keymer, A.E. 1990. Acquired immunity and epidemiology. In: Parasites: immunity and pathology (J.M. Behnke, ed.). Taylor and Francis Press, London.
Reduker, D.W. and Duszynski, D.W. 1985. Eimeria ladronensis n. sp. E. albiqula (Apicomplexa: Eimeriidae) from the woodrat, Neotoma albiqula (Rodentia: Cricetidae). J. Protozool. 32: 548-550. Reduker, D.W., Duszynski, D.W. and T.L. Yates. 1987. Evolutionary relationships among Eimeria spp. (Apicomplexa) infecting cricetid rodents. Can. J. of Zool. 65: 722-735.
Reduker, D.W., Hertel, L. and D.W. Duszynski. 1985. Eimeria species (Apicomplexa: Eimeriidae) infecting Peromyscus rodents in the southwestern United States and northern Mexico with description of a new species. J. Para. 71: 604-613. Rose, M.E. 1987. Immunity to Eimeria infections. Veter. Immun. Immunopath. 17: 333-343.
Rose, M.E. and Wakelin, D. & Hesketh, P. 1994. Interactions with Eimeria spp. and Trichinella spiralis in inbred mice. Para. 108: 69-75.
Ross, J. Tittensor, A.M., Fox, A.P. and Sanders, M.F. 1989. Myxomatosis in farmland rabbit populations in England and Wales. Epidem. Infec. 103: 333-357. 132
Royce, L.A. and Rossignol, P.A. 1990. Honey bee mortality due to tracheal mite parasitism. Para. 100: 147-151.
Samson, J., Holmes, J.C., Jorgenson, J.T. and Wishart, W.D. 1987. Experimental infections of free-ranging rocky mountain bighorn sheep with lungworms (Protostronqvlus spp.; Nematoda: Protostrongylidae). J. Wildl. Dis. 23: 396-403. Samuel, W.M. and Trainer, D.O. 1971. Some ecological factors influencing Eimeria mccordocki in white-tailed deer Odocoileus virqinianus. J. Protozool. 18: 306-308.
Schall, J.J. 1983. Lizard malaria: cost to vertebrate host's reproductive success. Para. 87: 1-6.
Schall, J.J., Bennett, A.F. and Putnam, R.W. 1982. Lizards infected with malaria: physiological and behavioral consequences. Science 217: 1057-1059.
Schall, J.J. and Bromwich, C.R. 1994. Interspecific interactions tested: two species of malarial parasite in a west African lizard. Oecologia 97: 326-332.
Schall, J. J., and Dearing, M.D. 1987. Malarial parasitism and male competition for mates in the western fence lizard, Sceloporus occidentalis. Oecologia 73: 389-392.
Scott, M.E. 1987. Regulation of mouse colony abundance by Heliqmosomoides polvqvrus. Para. 95: 111-124.
Scott, M.E. 1988. The impact of infection and disease on animals populations: implications for conservation biology. Conserv. Biol. 2: 40-56.
Scott, M.E. and Anderson, R.M. 1984. The population dynamics of Gyrodactylus bullatarudis (Monogenea) within laboratory populations of the fish host Poecilia reticulata. Para. 89: 159-194. Scott, M.E., and Dobson, A. 1989. The role of parasites in regulating host abundance. Parasitology Today 5: 176 183.
Scott, M.E. and Lewis, J.W. 1987. Population dynamics of helminths in wild and laboratory rodents. Mammal. Rev. 17: 95-103.
Shaw, J.L. and Moss, R. 1989. The role of parasite fecundity and longevity in the success of Trichostronqvlus tenius in low density red grouse populations. Para. 99: 253 258. 133
Shults, L.M., Seville, R.S., Stanton, N.L. and Menkens, G.E. 1990. Eimeria sp. (Apicomplexa: Eimeriidae) from Wyoming ground squirrels (Spermophilus elegans) and white-tailed prairie dogs (Cynomys leucurus) in Wyoming. Great Bas. Nat. 50: 327-331.
Slater, A.F.G. 1988. The influence of dietary protein on the experimental epidemiology of Heliqmosomoides polyqyrus (Nematoda) in the laboratory mouse. Proc. Roy. Soc. Lond., B, 234: 239-254.
Slater, A. and Keymer, A. 1986. Heliqmosomoides polyqyrus: the influence of dietary protein on the dynamics of repeated infection. Proc. Roy. Soc. Lond. B, 229: 69 83. Stanton, N.L., Shults, L.M., Parker, M. and Seville, R.S. 1992. Coccidian assemblages in the Wyoming ground squirrel, Spermophilus elegans. J. Parasitol. 78: 323 328.
Temple, S.A. 1987. Do predators always capture substandard individuals disproportionately from prey populations? Ecol. 68: 669-674. Theis, J.H. and Schwab, R.G. 1992. Seasonal prevalence of Taenia taeniaeformis: relationship to age, sex, reproduction and abundance of an intermediate host (Peromyscus maniculatus). J. Wildl. Dis. 28: 42-50.
Tilahun, G., and P. G. H. Stockdale. 1981. Oocyst production of four species of murine coccidia. Can. J. of Zool. 59: 1796-1800.
Timm, R.M. and Cook, E.F. 1979. The effect of bot fly larvae on reproduction in white-footed mice, Peromyscus leucopus. Amer. Mid. Nat. 101: 211-217. Tinsley, R.C. 1990. The influence of parasite infection on mating success in spedefoot toads, Scaphiopus couchii. Amer. Zool. 30: 313-324. Tocque, K. 1993. The relationship between parasite burden and host resources in the desert toad (Scaphiopus couchii), under natural environmental conditions. J. Anim. Ecol. 62: 683-693.
Tocque, K. and Tinsley, R.C. 1991. The influence of desert temperature cycles on the reproductive biology of Pseudodiplorchis americanus (Monogenea). Para. 103: 111-120. 134
Trout, R.C., Ross, J., Tittensor, A.M. and Fox, A.P. 1992. The effect on British wild rabbit population (Oryctolaqus cuniculus) of manipulating myxomatosis. J. Anim. Ecol. 29: 679-686.
Turner, C.M.R. and Cox, F.E.G. 1985. Interspecific interactions between blood parasites in a wild rodent community. Ann. Trop. Med. Para. 79: 463-465.
Upton, S. J., C. T. McAllister, D. B. Brillhart, D. W. Duszynski, and C. D. Wash. 1992. Cross-transmission studies with Eimeria arizonensis-like oocysts (Apicomplexa) in new world rodents of the Genus Baiomys, Neotoma, Onychomys, Peromyscus and Reithrodontomys (Muridae). Journal of Parasitology 78: 406-413. Vogel, P. and Bundy, D.A.P. 1987. Helminth parasites of Jamaican anoles (Reptilia: Iguanidae): variation in prevalence and intensity with host age and sex in a population of Anolis lineatopus. Para. 94: 399-404.
Wakelin, D. 1978. Immunity to intestinal parasites. Nature 273: 617-620. Wakelin, D. 1987. Parasite survival and variability in host immune responsiveness. Mammal Rev. 17: 135-141.
Wakelin, D. 1992. Immunogenetic and evolutionary influences on the host-parasite relationship. Dev. Comp. Immun. 16: 345-353.
Wakelin, D. and Blackwell, J. 1988. The Genetics of Resistance to Bacterial and Parasitic Infection, Taylor & Francis, London. Washburn, J.0., Mercer, D.R. and Anderson, J.R. 1992. Regulatory role of parasites: impact on host population shifts with resource availability. Science 253: 185 188. Wassom. D.L., Dick, T.A., Arnason, N., Strickland, D. and Grundman, A.W. 1986. Host genetics: a key factor in regulating the distribution of parasites in natural populations. Journal of Para. 72: 334-337.
Watkins, R.A., Moshier, S.E., O'Dell, W.D. & Pinter, A.J. 1991. Splenomegaly and reticulocytosis cause by Babesia microti infections in natural populations of the montane vole, Microtus montanus. J. Protozool. 38: 573 576. 135
Weatherhead, P.J. and Bennett, G.F. 1991. Ecology of red- winged blackbird parasitism by haematozoa. Can. J. Zool. 69: 2352-2359. Weatherhead, P.J. and Bennett, G.F. 1992. Ecology of parasitism of brown-headed cowbirds by haematozoa. Can. J. Zool. 70: 1-7. Webster, J.P. 1994. Prevalence and transmission of Toxoplasma qondii in wild brown rats, Rattus norveqicus. Para. 108: 407-411. Whitlaw, H.A. and Lankester, M.W. 1994. A retrospective evaluation of the effects of parelaphostrongylosis on moose populations. Can. J. Zool. 72: 1-7.
Wiger, R. 1977. Some pathological effects of endoparasites on rodents with special reference to the population ecology of microtines. Oikos 29: 598-606.
Wilber, P.G., Hanelt, B., Van Horne, B. and Duszynski, D.W. 1994. Two new species and temporal changes in the prevalence of eimerians in a free-living population of Townsend's ground squirrels (Spermophilus townsendii) in Idaho. J. Parasit. 80: 251-259.
Wilkinson, L. 1990. SYSTAT: the system for statistics. SYSTAT, Inc., Evanston, IL.
Wilson, G.R. 1983. The prevalence of caecal threadworms (Trichostronqylus tenius) in red grouse (Laqopus). Oecologia 58: 265-268. Wolff, J.O. 1989. Social Behavior. In Advances in the study of Peromyscus (Rodentia), G. L. Kirkland, Jr., and J. N. Layne (eds.). Texas Tech University Press, Lubbock, Texas, p. 169-232. Woolhouse, M.E.J., Taylor, P., Matanhire, D. and Chandiwana, S.K. 1991. Acquired immunity and epidemiology of Schistosoma haematobium. Nature 351: 757-759.
Zar, J.H. 1984. Biostatistical analysis. Prentice-Hall, New Jersey.