Ecology                                                                                                                  July, 2000
ISLAND AND TAXON EFFECTS IN.PARASITISM AND RESISTANCE OF LESSER ANTILLEAN BIRDS.
VICTOR APANIUS [1,3]
NANCY YORINKS [1,4]
ELDREDGE BERMINGBAM [2]
ROBERT E. RICKLEFS [1,2,5]
Abstract. 
Patterns of parasitism in insular avian communities can provide insight into processes that maintain host-parasite associations. On one
hand, replicable relationships within evolutionarily independent communities of the same host and parasite taxa would indicate that
these interactions are stable over time. On the other hand, unique ecological conditions on each island, sporadic colonizations and
extinctions, plus new genetic variation would lead to island-specific host--parasite relationships. We examined the distribution of three
parasitic taxa among avian host species on three islands in the Lesser Antilles: St. Lucia, Martinique, and Dominica. Blood protozoa of
the genus Haemoproteus were found in 34% of host individuals examined. A significant species effect, but no significant island effect,
was observed, suggesting an ecologically stable and replicable host-parasite association. Cysts of the tissue-dwelling protozoan genus
Sarcocystis were observed in 4% (9% on Dominica) of host individuals and were significantly associated with ground foraging.
Epithelial lesions characteristic of avian papilloma virus were recorded in 4% of host individuals on Martinique only. The pattern of
infection with papilloma virus or Sarcocystis (significant island effect) indicated that host species on a particular island are linked in
transmission webs of these parasites. Such island-specific associations suggest a role for either history or unique local ecology in
host-parasite associations.
There was a statistically significant interaction between island and species effects in the prevalence of Haemoproteus. This may stem
from the independent evolution of host--parasite interactions in the different island populations. We were able to assess the extent of
genetic divergence of the host species by analysis of mitochondrial ATPase 6,8 sequences. There was little genetic divergence between
island populations of the host species. Therefore, the variation in Haemoproteus prevalence is not likely to be related to genetic
differentiation of the host populations.
Birds infected with Haemoproteus exhibited elevated leukocyte levels indicative of immunological control of the parasite. After
statistically controlling for the intensity of Haemoproteus infection and host species, leukocyte levels varied significantly among islands
on which the host resided. This is consistent with the idea that insular avian communities are linked by transmission webs of parasites
having broad host specificity.
Key words: blood parasites; Haemoproteus; hematozoans; host-parasite interactions; immune response; Lesser Antilles; leukocyte
counts; papilloma virus; Sarcocystis.
INTRODUCTION
The dynamics and organization of communities of parasites
having many vertebrate hosts has proven to be a challenging
problem for ecologists and evolutionary biologists. Theoretical
models have considered the population dynamics of such
interactions (May and Anderson 1979, Hudson et al. 1985,
Crawley 1992), including the role of the immune response
(Norman et al. 1994). Although empirical investigations have
lagged, in part because of the complexity of natural communities
(e.g., Aho and Bush 1993), temporal and spatial variation in
parasite prevalence and infection by pathogens suggest complex,
spatially heterogeneous dynamics in host-parasite systems on
ecological scales, evolutionary scales, or both (van Riper et al.
1986, Prins and Weyerhaeuser 1987, Atkinson et al. 1988,
Atkinson and Van Riper 1991, Yezerinac and Weatherhead 1995).
Empirical studies of temporal and spatial heterogeneity in
parasite--host interactions are most informative when
comparisons are made among discrete populations. Longitudinal
studies of geographically circumscribed host populations can
provide insight into the ecological and evolutionary dynamics of
multiple host--parasite relationships (Forbes et al. 1994, Bennett
et al. 1995, Merila et al. 1995), but inferences may be obscured by
temporal or spatial variation in the data.
The roles of ecological and evolutionary factors in the outcome
of host-parasite interactions can be teased apart statistically by
examining a set of species over several discrete localities, such as
isolated islands. In this case, island effects represent unique
attributes of individual islands, which may include the array of
habitats on each island, composition of the host community, and
presence of suitable vectors and reservoirs for parasites. To the
extent that an island effect may result from community
composition, it may also, in part, reflect stochasticity of
colonization and extinction independent of an island's ecology.
Host-specific effects may arise because of unique genetic or
ecological attributes of each type of host. From the standpoint of
host-parasite evolution, the most revealing source of variation is
the host X island interaction, which indicates the extent of unique
outcomes of host-parasite relationship in a particular host across
a set of ecologically defined locations. Such statistic al
interactions could result from island-specific genetic factors in
Ecology                                                                                                                  July, 2000
the host resulting from founder events and mutation. Host-
parasite coevolution, i.e., reciprocal genetic changes in host and
parasite populations, would also have to be considered as a
possibility.
Several studies have addressed the distribution of parasites across
island populations, but these have lacked suitable sampling to
detect host X island interactions. An analysis of the gut helminth
parasites of Anolis lizards in the Lesser Antilles related variation
in parasitism to the distribution of hosts among habitats within
islands rather than to ecological differences between islands
associated with area or altitude (Dobson and Pacala 1992).
Repeatable associations between parasites and hosts thus appear
to follow upon environmental factors. However, disjunct
distributions of many parasites suggested that historical factors,
such as colonization and extinction, may have influenced the
particular assemblage of parasite species within each host
population (Dobson and Pacala 1992). Because Anolis comprises
only one or two morphologically distinct species on most islands,
it is difficult to resolve whether the observed patterns represent
the evolutionary outcome of host-parasite interactions or purely
sto chastic processes.
Surveys of island avifaunas in the Indian Ocean likewise indicate
considerable heterogeneity in the distribution of blood parasites.
Haemoproteus spp. and Plasmodium spp. infect passerine birds
on Aldabra (Lowery 1974), but haemoproteids are rare on the
Mascarene Islands (Pierce et al. 1977) and absent in the Malagasy
Republic (Bennett and Blancou 1974). Leucocytozoon spp.
infections have not been observed in Aldabra, but are frequent in
the Mascarene Islands and Malagasy Republic. Presence or
absence of a blood parasite from a particular area may depend on
the presence or absence of competent insect vectors. Strong
island effects and turnover of the avifaunas between islands
preclude the estimation of island X host interactions in these
data.
In this study, we present data on haematozoan parasite
prevalence and white blood cell counts for six species of
passerine bird on three adjacent islands in the Lesser Antilles.
The host species were the Bananaquit (Coereba flaveola),
Antillean Elaenia (Elaenia martinica), Lesser Antillean Bullfinch
(Loxigilla noctis), Streaked Saltator (Saltator albicollis),
Black-faced Grassquit (Tiaris bicolor), and the Black-whiskered
Vireo (Vireo altiloqaus). Data were collected during a single, brief
period in 1991. Our sampling design allowed us to estimate island
and host effects as well as host X island interactions. Significant
host effects in the prevalence of the protozoan parasite
Haemoproteus spp. demonstrated that the sampling program was
large enough to discern statistically significant host-parasite
combinations. Moreover, blood leukocyte counts, which indicate
an individual host's response to its parasite load (Fallis et al. 1951,
Desser et al. 1968), show that both deterministic and historical
factors shap e host- parasite associations. Finally, we suggest a
framework for gauging the relative importance of these factors
based on island biogeographic analysis.
STUDY AREA AND METHODS
This study was conducted on St. Lucia, Martinique, and
Dominica, Lesser Antilles, between 20 July and 10 August 1991.
These islands are similar in size (616- 1100 [km.sup.2]) and reach
elevations (960-1450 m) that create a variety of habitats ranging
from montane forest to dry lowland scrub (Beard 1949, Lack
1976). We sampled both wet and dry habitats in different parts of
each island to obtain a representative sample of each island's
avifauna. To some extent, the sample sizes (9- 37 individuals per
island population) reflect the relative abundances of the host
species on each island. Birds were captured with mist nets at the
following locations (and dates): on St. Lucia, Edmond Forest
Preserve (20- 22 July), and Anse La Sorciere (23-25 July); on
Dommica, Springfield Plantation-Mt. Joy Estate (27-29 July),
Syndicate Estate (30-31 July), Ponte Casse (1 August), and
Glanvillia Quarry (2 August); on Martinique, Arboretum (6-7
August), Point Rouge (8 August), Le Francoise (9 August), and
Grand Fond (10 August).
Blood samples were taken by jugular venipuncture and smears
were prepared and fixed in absolute methanol in the field. Blood
smears were stained in Giemsa in order to visually identify
leukocytes and parasites. Hemoparasite and leukocyte levels were
estimated in five randomly chosen fields at 400X magnification
(Russo et al. 1986) and are expressed as the number of parasites
or leukocytes per 13000 erythrocytes. Hemoparasite and
leukocyte identifications follow Campbell (1988) and Hawkey and
Dennett (1989). All microscopy was performed by the same
observer (N. Yorinks). Prevalence refers to the percentage of
hosts infected and parasitemic intensity, or parasitemia, refers to
the number of parasites per 13 000 erythrocytes.
All birds underwent physical examination of the superficial
surfaces, especially featherless areas (Harrison and Ritchie 1994),
to detect the distinctive skin lesions attributable to papilloma
virus infection (Gerlach 1986) and the characteristic
intramuscular "rice-grain" macrocysts of the protozoan parasite
Sarcocystis spp. (Box et al. 1984).
Pectoral muscle biopsies and blood samples were collected
nondestructively from all captured birds (Seutin et al. 1993) and
were preserved as described in Seutin et al. (1991). All samples
were collected and transported under the appropriate permits and
licenses. DNA extractions in the laboratory of E. Bermingham
followed the protocol of Seutin et al. (1993). We amplified a 1074
bp segment of mtDNA that spanned the full tRNALys, ATPase
8, and ATPase 6 genes by polymerase chain reaction (PCR) with
the primer pair CO2GQL and CO3HMH. (Primer sequences and
amplification conditions are available upon request from E.
Bermingham.) Amplification products were cleaned
electrophoretically and purified by adsorption onto silica in the
presence of high salt buffer (Geneclean procedure, Bio101, Vista,
California, USA), and were sequenced in the L direction using
DyeDeoxy Terminator Cycle Sequencing (Applied Biosystems
Division of Perkin Elmer, Foster, California, USA) with the
primers CO2GQL, LYSL, PKL, PKLD, and TPL. Nu cleotide
sequences were determined by electrophoresing purified
amplified product in an Applied Biosystems 373A DNA
sequencer.
Ecology                                                                                                                  July, 2000
Estimates of interpopulation nucleotide divergence were based
on the overlapping coding region (841 base pairs) of the ATPase
6 and ATPase 8 genes. Pairwise divergences of silent nucleotide
substitutions between individual sequences were calculated
according to Takahata and Nei (1985) and Takahata (1989).
We used SAS/STAT version 6.08 (SAS Institute 1990) to
estimate correlation coefficients within island populations
(PROC NESTED) and to perform ANOVAs (PROC
GENMOD) of parasitemic intensity and leukocyte levels. Host
species and island effects in contingency tables of parasitic
infection were estimated using PROC CATMOD. The
contingency table analysis accounts for the nature of the
prevalence data (frequencies) and the unbalanced sample sizes.
These will be explained in more detail, along with other statistical
tests.
RESULTS
Blood protozoa
Gametocytes of Haemoproteus spp. (the only blood parasite
observed) appeared in blood smears of four of the six avian host
species scored on each island (Fig. 1). The pooled prevalence rate
was 34%. Five of the six species were consistently infected or
uninfected across the three islands included in this study. A
categorical model relating Haemoproteus infection to host
species and island revealed a significant species effect, but no
island effect, and a significant host X island interaction (Table 1).
Of the four species infected with haemoproteids, the probability
of infection varied significantly among islands in Coereba flaveola
(log-likelihood [[chi].sup.2] = 13.0, df = 2, P = 0.0015), Loxigilla
noctis ([[chi].sup.2] = 10.2, df = 2, P = 0.006), and possibly Vireo
altiloquus ([[chi].sup.2] = 7.1, df = 2, P = 0.03). Prevalence of
Haemoproteus in Tiaris bicolor did not vary among islands
([[chi].sup.2] = 0.4, df = 2, P = 0.84).
For Loxigilla noctis, the prevalence of haemoproteid infection
did not differ significantly between sites on the same island
(habitat effect: [[chi].sup.2] = 0.6, df = 1, P = 0.45), although
prevalence did differ significantly between islands (island effect:
[[chi].sup.2] = 11.4, df = 2, P = 0.003; Table 2). On St. Lucia,
sufficient numbers of three species were captured to assess
differences between sampling sites. The analysis revealed that
prevalence did not differ significantly between sites ([[chi].sup.2]
= 1.2, df = 1, P = 0.26), despite significant differences among
host species ([[chi].sup.2] = 22.0, df = 2, P [less than] 0.0001;
Table 2).
There were no discernible differences of intensity of
haemoproteid infection among infected individuals of the four
species known to be infected (Poisson regression, PROC
GENMOD: log-likelihood [[chi].sup.2] = 3.2, df = 3, P [greater
than] 0.15). For Loxigilla noctis (shown in Fig. 2), the sample size
was sufficient to test whether intensity differed between islands,
but significant differences could not be detected (log-likelihood
[[chi].sup.2] = 1.5, df = 2, P [greater than] 0.15).
Genetic distances between island populations
Populations of the six species included in this study are relatively
undifferentiated over the islands of St. Lucia, Martinique, and
Dominica (Table 3), although a strong genetic disjunction has
been observed in the Bananaquit Coereba flaveola between the
islands of St. Lucia and St. Vincent (Seutin et al. 1994). Average
pairwise genetic distances between haplotypes on different
islands (percentage of nucleotide substitution) varied from a low
of 0.06% for Saltator albicollis between Martinique and St. Lucia
to a high of 0.49% for Loxigilla noctis between the same two
islands. Average pairwise nucleotide divergence within islands for
several populations for which we obtained sequences for three or
more individuals was similar to between-population distances:
Coereba flaveola (Dominica, 0.08%, n = 3; St. Lucia, 0.10%, n =
9); Loxigilla noctis (Martinique, 0.56%, n = 3; St. Lucia, 0.23%, n
= 5). Thus, these island populations of these species appear to be
differentiated weakly, if at all.
Tissue parasites
Cysts of Sarcocystis sp. were observed in skeletal muscle of 8 of
27 potential host species sampled on Dominica, 2 of 20 species
on Martinique, and none of 21 species on St. Lucia (Table 4). A
categorical model of the presence or absence of Sarcocystis in
hosts on the three islands revealed a weak island effect
([[chi].sup.2] = 5.4, df = 2, P = 0.067). Of seven host species
with eight or more individuals sampled on each island,
Sarcocystis was found in three (Saltator albicollis, Tiaris bicolor,
Loxigilla noctis) on one island (Dominica), whereas only one
infected individual of Loxigilla noctis was found on another
island (Martinique). Because of the small sample of species, this
distribution is not significantly heterogeneous ([[chi].sup.2] = 2.6,
df = 2, P = 0.27). An additional five species were infected with
Sarcocystis on Dominica, whereas only one other species
(Myiarchus oberi) was found to be infected on another island
(Martinique).
Skin lesions indicative of papilloma virus infection were found in
only seven island populations, six of these on the island of
Martinique (Table 4). In spite of the small sample size and an
overall prevalence of only 1.5%, this pattern was heterogenous
enough to produce a significant island effect ([[chi].sup.2] = 7.2,
df 2, P 0.027).
Leukocyte levels
We counted four types of leukocytes: heterophils (or
neutrophils), eosinophils, monocytes, and lymphocytes.
Lymphocytes are undifferentiated agranulocytes, which are
apparently involved in production of antibody. Monocytes also
lack granules and are primarily phagocytotic, as are the granulated
heterophils. The function of the granulated eosinophils is less
well understood, but may be related to extracellular defense
against large parasites, such as helminths (Roitt et al. 1989). Of
the four types of leucocytes, the abundances of only heterophils
and eosinophils were significantly correlated within island
populations (r = 0.21, n = 352, P [less than] 0.01; Table 5),
indicating relative independence in the levels of these cell types in
the blood. All four types of leukocytes summed give the
leukocyte, or white blood cell (WBC), count. This summary
variable was normally distributed following log transformation.
Ecology                                                                                                                  July, 2000
Here, we present results only for WBC counts because additional
analyses of individual leukocyte typ es (not shown) did not
provide additional or conflicting information.
Among individuals infected with Haemoproteus, leukocyte levels
were positively correlated with parasitemic level (r = 0.39, P [less
than] 0.01, n = 82; Table 5), due primarily to a correlation
between heterophil level and parasitemia (r = 0.30, P [less than]
0.01, n = 82). WBC levels were elevated in individual birds
infected with haemoproteids, compared to noninfected
individuals, after species and island effects were removed (Table
6). WBC levels were also significantly correlated with the intensity
of infection (parasitemia), over and above the heterogeneity of
parasitemia with respect to individual species and islands (Table
7). WBC levels were not related to Sarcocystis and papilloma
virus infection, nor to the interaction of these tissue parasites
with haemoproteid infection (analyses not shown).
WBC levels differed significantly between avian host species and
islands (Tables 6 and 7, Fig. 3). The interaction between these
two effects was of marginal statistical significance (P = 0.06), and
became nonsignificant (P [greater than] 0.15) when haemoproteid
intensity was included in the model. Thus, it appears that a
significant component of the variance in WBC levels can be
attributed to blood parasite (Haemoproteus) infection, but that
additional significant components can be attributed to both
island and host species. This is consistent with the significant
host effects observed in haemoproteid infections and significant
island effects observed in Sarcosystis spp. and papilloma virus
infections; these effects might apply to other parasite organisms
not monitored in this study.
DISCUSSION
This analysis has revealed strong species effects and species X
island interactions in Haemoproteus infections, and significant
relationships of leukocyte counts to Haemoproteus infection and
parasitemia. Furthermore, there were significant species and
island effects over and above the response of leucocytes to
Haemoproteus. These results suggest species-specific variation in
the interactions of hosts with their pathogens, as well as
independence of individual island populations with respect to
their interactions with pathogens.
Differences in hematozoan prevalence among
Lesser Antillean, Jamaican, and mainland (Panama) populations
The comparability of our sample from the Lesser Antilles with
those of other studies depends, in part, on the temporal stability
and seasonal consistency of hematozoan prevalences. In a sample
of almost 1600 resident birds on the island of Jamaica, Bennett et
al. (1980) found that prevalence varied seasonally, from a low
during the fall and winter to a high during the spring and
summer, which encompasses the period of our study. In addition,
many individuals were recaptured over periods of three months
or more. Of individuals infected at the time of their initial
capture, about half (9/20) lost traces of the infection on blood
smears; a much lower percentage of apparently uninfected birds
(9/283 = 3.2%) either gained infections or their parasitemias
increased above the threshold of detection. Because individuals
seem to lose infections at a high frequency, and because the
percentage of infected individuals exhibited significant seasonal
variation, it is important that different locations be compared at
the s ame season of the year. Our study was conducted at a time
(toward the end of the breeding season) when prevalences of
blood parasites should have been high.
We did not attempt to identify parasites to the species level. In
the case of Haemoproteus, new species are described when
parasites exhibit distinctive morphology in a new host species
(Bennett et al. 1994). At present, it is uncertain whether
morphological variation reflects genetic differences between
parasites on different hosts, or host-induced phenotypic
differentiation in a single population of parasites (Atkinson 1991,
Burry-Caines and Bennett 1992). Resolution of this issue awaits
genetic analyses of Haemoproteus on different hosts.
Haemoproteid infections are common in neotropical passerines
found within the Caribbean region. Coereba flaveola showed low
prevalence ([less than]5%) of Haemoproteus on Jamaica (Bennett
et al. 1980) and in Panama (Galindo and Sousa 1966, Sousa and
Herman 1982). This is consistent with our observations from
Martinique (0 of 9 infected), but contrasts with the higher
prevalence in this species found on Dominica (14/22 = 64%).
The low prevalence of infection on Jamaica further emphasizes
the between-island heterogeneity that we observed in this species.
The absence of visible haemoproteid infection in Elaenia
martinica in this study is consistent with the low prevalence of
Haemoproteus in other species of Elaenia in continental localities
(3-10%; White et al. 1978, Sousa and Herman 1982) and on
Jamaica (10%; Bennett et al. 1980). In view of the heterogeneity
observed within some species, including the genetically
undifferentiated Coereba flaveola on Dominica and Martinique
(Seutin et al. 1994), one would not necessarily expect such
consistency among different species within a genus, even though
hematozoan prevalence does exhibit family-level heterogeneity
(Ricklefs 1992).
The genus Loxigilla is endemic to the West Indies; hence,
comparison of hematozoan prevalence in L. noctis can only be
made with other species in the same subfamily (Emberizidae:
Emberizinae). Haemoproteid infection in emberizines on Jamaica
was quite low (3%; Bennett et al. 1980) and averaged 10% in
Panamanian (Sousa and Herman 1982) and South American
populations (White et al. 1978). Within the Emberizinae,
infection varies greatly, although prevalences [greater than]30%
are unusual for neotropical finches (White et al. 1978). Thus, the
high prevalence of Haemoproteus spp. infection in L. noctis and
the variation between islands is noteworthy. Among other
emberizines, the prevalence of Haemoproteus in Tiaris (7/42 =
17%) would appear to be typical; the absence of Haemoproteus
(0/49) from Saltator (Emberizinae: Cardinalinae) lies in striking
contrast.
The prevalence of haemoproteid infection in Vireo altiloquus is
not unusual compared to other neotropical species of the genus
Vireo. On Jamaica, for example, 44% of Vireo altiloquus were
Ecology                                                                                                                  July, 2000
similarly infected (Bennett et al. 1980). For vireonids in general,
prevalence was 23% for Panama (Sousa and Herman 1982) and
14% for the neotropics (White et al. 1978); prevalences of all
hematozoans together in tropical Vireonidae were 34% (White et
al. 1978).
Other blood parasites
There was no evidence of Leucocytozoon spp. infection in the
Lesser Antillean avifauna. This parasite was also absent from a
survey of nonmigratory Jamaican birds, yet small numbers of
wintering temperate migrants were found to harbor
Leucocytozoon spp. (Bennett et al. 1980). On Cuba, a survey of
45 birds comprising 20 species failed to detect Leucocytozoon
infections (Zajicek and Mauri Mendez 1969). Furthermore,
Leucocytozoon is infrequent in the blood of tropical birds in
general (White et al. 1978). It is not known whether this is related
to the distribution and abundance of their principal vectors,
simulid flies, in the tropics.
Infections with Plasmodium spp. were not observed in the Lesser
Antilles, although plasmodial infections have been reported for
Coereba flaveola from Jamaica (6/139 = 4%; Bennett et al. 1980)
and Panama (1/24 = 4%; Sousa and Herman 1982). Plasmodium
infections have also been reported in Tiaris bicolor from Jamaica
(1/46 = 2%; Bennett et al. 1980) and in Saltator albicollis from
Panama (1/54 = 2%; Sousa and Herman 1982). Emberizids
typically have a low prevalence of malaria: 5% in the continental
neotropics (White et al. 1978) and 2% on Jamaica (Bennett et al.
1980). Plasmodium infection was observed on Cuba,
but as a rare ([less than]1%) infection of domestic chickens
(Zajicek and Mauri Mendez 1969). We also did not observe
trypanosomes or microfilariae in peripheral blood smears. Blood
smears can detect infection by these hematozoa, but are thought
to have low diagnostic sensitivity (Bennett 1962). Despite this
fact, trypansome infections were found in Goereba flaveola (1/24
= 4%), Tiaris bicolor (2/46 = 4%), and Vireo altiloquus (1/55 =
2%) from Jamaica (Bennett et al. 1980). In Panama, trypanosome
infections have been recorded for Coereba flaveola (1/24 = 4%;
Sousa and Herman 1982). Microfilarial infections have been
reported for Coereba flaveola from Jamaica (1/139 = 1%;
Bennett et al. 1980) and for Coereba flaveola (1/24 = 4%) and
Saltator albicollis (1/54 = 2%) from Panama (Sousa and Herman
1982). It is possible that Plasmodium, Trypanosoma, and
microfilaria infections were overlooked in our populations, but it
is doubtful that these species would have provided additional
biogeographic information.
Effect of host species
The prevalence of Haemoproteus varied significantly among host
species; it was absent from Elaenia martinica and Saltator
albicollis. The strong species effect suggests either that each
species is infected by a genetically differentiated population of the
parasite, or that exposure to vectors of the parasites or resistance
to the disease itself varies among hosts. Haemoproteus is
transmitted by sandflies, or biting midges, of the genus
Culicoides (Diptera: Ceratopogonidae; Harwood and James 1979,
Kettle 1982), and by louse-flies (Diptera: Hippoboscidae; Kettle
1982), both of which occur in the Lesser Antilles (Bequaert 1954,
Wirth 1974). The two species lacking haemoproteids in our
sample were captured in the same areas and forest strata
(understory) as species that were heavily infected. There were also
no obvious differences in feeding ecology of infected and
noninfected species. Thus, the distribution of haemoproteids
among host species suggests that variation in prevalence is due to
unique attributes of each host-parasite interaction.
The observation that particular bird species are consistently
infected with Haemoproteus across islands implies that suitable
vectors are not limiting. This leads to the prediction that
hippoboscid louse flies are the most important vectors, because
these ectoparasites would be transported by their hosts during
colonization events. In addition, hippobocids are more likely to
be restricted to particular host species than are free-living
ceratopogonid midges. Unfortunately, little is known of the host
specificity of the hippoboscid and ceratopogonid vectors of
Haemoproteus or of the parasites themselves.
Sarcocystis is known to have a wide avian host range, but
transmission between host species evidently requires contact with
the ground (Box et al. 1984). The distribution of Sarcocystis
among potential hosts is consistent with transmission through
soil. On Dominica, the only island with substantial occurrence,
Sarcocystis was found in 6 of 11 entirely or partly ground-feeding
species of bird, but in only 2 of 16 species that rarely or never
forage on the ground ([[chi].sup.2] = 5.5, df = 1, P [less than]
0.025).
Papilloma virus spreads by direct transmission through the air
(Gerlach 1986), so one might assume that infection would be
more prevalent among abundant species than among less
common ones. Except for Troglodytes aedon on Dominica,
papilloma virus infections were almost completely restricted to
Martinique. Of the six species represented by more than 12
individuals in our sample, four carried papilloma infections,
whereas only two of 14 species with 12 or fewer individuals were
infected. However, prevalences were low (generally [less
than]10%), so our ability to detect infections in small samples
was limited. Infected species exhibited no obvious ecological
attributes in common. Together with the strong island effect
evident for papilloma virus, virtually restricted to Martinique, this
suggests between-island differences in the overall community
ecology of the total host--parasite complex. Presumably, this
effect must be linked to aspects of the transmission of the
disease, including the possibility of introd uction from poultry,
pet birds, or contaminated material transported to the islands.
Island effects
Prevalence of Haemoproteus sp. did not vary significantly among
islands (Table 1). Sarcocystis was found almost exclusively on
Dominica, and papilloma virus infections on Martinique. Birds
on St. Lucia were apparently free of both diseases. Considering
the distribution of these diseases over bird species, regardless of
their abundance in our sample, these distributions were
significantly heterogeneous in the case of papilloma virus (P
0.027) and marginally so in the case of Sarcocystis (P = 0.067).
Ecology                                                                                                                  July, 2000
The island effect for Sarcocystis may reflect the presence on
Dominica of a suitable mammalian reservoir for the disease (Box
et al. 1984), but this idea has not been assessed directly. There is
no such ready explanation for the distribution of papilloma virus,
but its localization may reflect unique genotypes of the pathogen
on Martinique, coupled with broad infectivity across potential
hosts.
Species X island interactions
Although no island effect was evident in the prevalence of
Haemoproteus, several species exhibited significant, but
independent, variation in prevalence among islands (the species X
island effect in Table 1). In the absence of information on host
specificity of Haemoproteus, the simplest explanation for these
statistical interactions invokes genetic variation in resistance to
Haemoproteus, or in host-specific patterns of infectivity by a
shared pathogen, or both. Based on mtDNA haplotypes, none of
the six species in this sample exhibits significant genetic
divergence between islands. Other species do show significant
divergence indicative of independent evolution of populations
over the same set of Lesser Antillean islands (E. Bermingham
and R.E. Ricklefs, unpublished observations). It is possible that
there is no longer effective gene flow between island populations
of the species in this sample, but that there has not been enough
time for these populations to accumulate unique mtDNA
variation. Genetic fact ors related to disease resistance, such as
genes of the major histocompatibility complex, may diverge
under strong selection much more rapidly than mtDNA. It is also
possible that generalized pathogens may have evolved different
spectra of infectivities as a result of their relationships with other
hosts on a particular island.
Mitochondrial DNA sequences are generally thought to diverge
at a rate of 2% per [10.sup.6] years, representing a nucleotide
substitution rate of [10.sup.-8]/yr (Shields and Wilson 1987,
Bermingham and Lessios 1993, Knowlton et al. 1993, Tarr and
Fleischer 1993; see Klicka and Zink 1997: note 11). Genetic
distances between haplotypes in different island populations
range between [sim]0.1% and 0.5% in our sample, which may not
significantly exceed distances between haplotypes within island
populations. Even if between-island distances represent
significant genetic divergence (the three populations of Loxigilla
noctis, e.g., have no haplotypes in common in our small sample),
estimated time since evolutionary isolation would be a maximum
of 50 000-250 000 years, and almost certainly much less.
Therefore, any genetic effects that may have produced
differences in Haemoproteus prevalence among island
populations of the same species must have arisen over periods of
tens of thousands of years (generations), which is consistent with
selective enhancement of available genetic variation, but less so
with the production of new genetic variation by mutation.
Immune system activity
The significant difference between species in WBC levels is
consistent with the strong species effect in prevalence of
Haemoproteus (Bennett and Hawkey 1988, Hawkey and Bennett
1988). The positive relationship between WBC levels and the
intensity of haemoproteid infection recalls hematological changes
that accompany infection with other avian hematozoa, such as
Leucocytozoon spp. (Fallis et al. 1951, 1956, Desser et al. 1968).
Even after factoring out the presence/absence or intensity of
Haemoproteus infection, species effects on levels of blood
leukocytes remained significant (Tables 6 and 7). This suggests
either that other parasites not monitored in this study also vary
significantly among species, or that the general response of
leukocyte counts to infection varies in a species-specific manner.
It was surprising, however, that within the entire sample of host
species, island effects on white blood cell counts were significant
even after statistically controlling for host species and blood
parasite intensity. Apparently, unrelated hosts that are members
of the same community are uniformly responding to
yet-unidentified parasites that may have broad host specificity but
restricted distribution over islands. Leukocyte data thus indicate
that unrelated species inhabiting the same community (island) are
linked, possibly through parasites of broad host specificity. These
agents might be arthropod-transmitted viruses, water
contaminated with enteric bacteria, or helminths whose infective
eggs and larvae are found in soil and invertebrate hosts.
Immunological characterization of community-wide host-parasite
interactions
In addition to providing information on the distribution and
prevalence of particular parasites in specific hosts in great detail,
comparison of immunological parameters among hosts and
location may provide a simple, indirect characterization of the
organization of local host-parasite associations and their
consistency between localities. Analysis of variance in leukocyte
levels can identify effects of species, location, and season, as well
as quantify the intrinsic variation in response to parasitism within
a local population at a particular time.
Surveys of blood parasites in bird populations suggest a high
degree of geographic heterogeneity in the prevalence of any
particular infectious agent (Forbes et al. 1994, Bennett et al. 1995,
Meril[ddot{a}] et al. 1995, Yezerinac and Weatherhead 1995).
This variation makes it difficult to assess the overall level of
infection in a population as a whole from studies of individual
parasites. Our hematological results show that leukocyte levels
are sensitive to a prominent blood parasite, but also appear to
respond to other infectious agents not surveyed. Thus, leukocyte
levels are likely to provide a general index to the prevalence of
infections within a population.
It is commonly believed that bird populations in a natural
community are linked by pools of circulating parasites, but as yet
this has been poorly documented. The island effect on leukocyte
abundance strongly implies the existence of local infectious
agents widely shared among different hosts. Such patterns could
be explored further by examining host immunologic responses
and immunogenetic variation. Circulating levels of cytokines and
immunoglobulins can provide corroborative information on the
generalized activity of the immune system. The distribution of
major histocompatibility complex genes in the host populations
would provide relevant immunogenetic information (Apanius et
al. 1997). These indicators would not only integrate information
Ecology                                                                                                                  July, 2000
from a variety of parasitic exposures, but also provide a better
mechanistic understanding of these ecological interactions.
ACKNOWLEDGMENTS
We thank Gilles Seutin, William Schew, and Doug Wechsler for
assistance in the field, Gilles Seutin and Jeffrey Hunt for
assistance in the laboratory, and Anders P. Moller for comments
on the manuscript.
(1.) Department of Biology, Leidy Laboratories, Universiiy of
Pennsylvania, Philadelphia, Pennsylvania 19104-6018 USA
(2.) Smithsonian Tropical Research Institute, Unit 0948, APO
AA 34002-0948 USA, and Apto. Postal 2072, Balboa, Republicaf
de Panama
(3.) Present address: Department of Biological Sciences, Florida
International University, University Park, Miami, Florida 33199
USA. E-mail: apanius@fiu.edu
(4.) Present address: School of Veterinary Medicine, 2015 Linden
Drive West, University of Wisconsin, Madison, Wisconsin 53706
USA.
(5.) Present address: Department of Biology, University of
Missouri, 8001 Natural Bridge Road, St. Louis, Missouri 63
121-4499 USA.
LITERATURE CITED
Aho, J. M., and A. O. Bush. 1993. Community richness in
parasites of some freshwater fishes from North America. Pages
185-193 in R. E. Ricklefs and D. Schiuter, editors. Species
diversity in ecological communities: historical and geographical
perspectives. University of Chicago Press, Chicago, Illinois, USA.
Apanius, V., D. Penn, P R. Slev, L. R. Ruff, and W. K. Potts.
1997. The nature of selection on the major histocompatibility
complex. Critical Reviews in Immunology 17:179-224.
Atkinson, C. T. 1991. Vectors, epizootiology, and pathogenicity
of avian species of Haemoproteus (Haemosporina:
Haemoproteidae). Bulletin of the Society of Vector Ecology
16:109-126.
Atkinson, C. T., D. J. Forrester, and E. C. Greiner. 1988.
Epizootiology of Haemoproteus meleagridis (Protozoa:
Haemosporina) in Florida: seasonal transmission and vector
abundance. Journal of Medical Entomology 25:45-51.
Atkinson, C. T., and C. Van Riper, III. 1991. Pathogenecity and
epizootiology of avian haematozoa: Plasmodium,
Leucocytozoon, and Haemoproteus. Pages 19-48 in J. E. Loye
and M. Zuk, editors. Bird--parasite interactions: ecology,
evolution, and behaviour. Oxford University Press, Oxford, UK.
Beard, J. S. 1949. The natural vegetation of the Windward and
Leeward Islands. Clarendon Press, Oxford, UK.
Bennett, G. F. 1962. The hematocrit centrifuge for laboratory
diagnosis of hematozoa. Canadian Journal of Zoology 40:
124-125.
Bennett, G. F, and J. Blancou. 1974. A note on the blood
parasites of some birds from the Republic of Madagascar. Journal
of Wildlife Diseases 10:239-240.
Bennett, G. F, M. A. Peirce, and R. A. Earl[acute{e}]. 1994. An
annotated checklist of the valid avian species of Haemoproteus,
Leucocytozoon (Apicomplexa: Haemosporida) and Hepotozoon
(Apicomplexa: Haemogregarinidae). Systematic Parasitology
29:61-73.
Bennett, G. F, D. Squires-Parsons, P Siikam[ddot{a}]ki, E.
Huhta, K. Allander, and L. Hillstr[ddot{o}]m. 1995. A
comparison of blood parasites of three Scandinavian populations
of the Pied Flycatcher (Ficedula hypoleuca). Journal of Avian
Biology 26:33-38.
Bennett, G. F, H. Witt, and E. M. White. 1980. Blood parasites of
some Jamaican birds. Journal of Wildlife Diseases 16:29-38.
Bennett, P M., and C. M. Hawkey. 1988. Comparative
hematology: phylogenetic and ecological aspects in mammals and
birds. Pages 33-48 in D. J. Blackmore, editor. Animal clinical
biochemistry. Cambridge University Press, Cambridge, UK.
Bequaert, J. C. 1954. The Hippoboscidae or louse-flies (Diptera)
of mammals and birds. Part II. Taxonomy, evolution and
revision of American genera and species. Entomologica
Americana, 34 (new series): 1-232.
Bermingham, E., and H. Lessios. 1993. Rate variation of protein
and mtDNA evolution as revealed by sea urchins separated by
the Isthmus of Panama. Proceedings of the National Academy of
Sciences (USA) 89:2734-2738.
Box, E. D., J. L. Meier, and J. H. Smith. 1984. Description of
Sarcocystis falculata Stiles 1893, a parasite of birds and opossums.
Journal of Protozoology 31:521-524.
Burry-Caines, J. R., and G. F. Bennett. 1992. The
haemoproteidae (Apicomplexa:Haemosporina) of the avian
families Fringillidae and Emberizidae s.l. Canadian Journal of
Zoology 70:1149-1160.
Campbell, T. W. 1988. Avian hematology and cytology. Iowa
State University Press, Ames, Iowa, USA.
Crawley, M., editor. 1992. Natural enemies: the population
biology of predators, parasites and diseases. Blackwell, Oxford,
UK.
Desser, S. S., A. M. Fallis, and P. C. C. Garnham. 1968. Relapses
in ducks chronically infected with Leucocytozoon simondi and
Parahaemoproteus nettionis. Canadian Journal of Zoology
46:281-285.
Ecology                                                                                                                  July, 2000
Dobson, A. P, and S. W. Pacala. 1992. The parasites of Anolis
lizards of the northern Lesser Antilles. Oecologia 92:118-125.
Fallis, A. M., R. C. Anderson, and G. F. Bennett. 1956. Further
observations on the transmission and development of
Leucocytozoon simondi. Canadian Journal of Zoology 34:
289-404.
Fallis, A. M., D. M. Davies, and M. A. Vickers. 1951. Life history
of Leucocytozoon simondi Mathis and Leger in natural and
experimental infections and blood changes in the avian host.
Canadian Journal of Zoology 29:305-328.
Forbes, M., P. J. Weatherhead, and G. F. Bennett. 1994. Blood
parasites of blue grouse--variation in prevalence and patterns of
interspecific association. Oecologia 97:520-525.
Galindo, P, and O. Sousa. 1966. Blood parasites of birds from
Almirante, Panama with ecological notes on the hosts. Revista
Biologia Tropica 14:27-46.
Gerlach, H. 1986. Viral diseases. Pages 408-433 in G. J. Harrison
and L. R. Harrison, editors. Clinical avian medicine and surgery.
W. B. Saunders, Philadelphia, Pennsylvania, USA.
Harrison, G. J., and B. W. Ritchie. 1994. Making distinctions in
the physical examination. Pages 144-175 in B. W. Ritchie, G. J.
Harrison, and L. R. Harrison, editors. Avian medicine: Principles
and application. Wingers Publishing, Lake Worth, Florida, USA.
Harwood, R. F, and M. T. James. 1979. Entomology in Human
and Animal Health. Seventh edition. Macmillan, New York, New
York, USA.
Hawkey, C. M., and P. M. Bennett. 1988. Comparative
haematology: clinical aspects in mammals and birds. Pages 49-62
in D. J. Blackmore, editor. Animal clinical biochemistry.
Cambridge University Press, Cambridge, UK.
Hawkey, C. M., and T. B. Dennett. 1989. Color atlas of
comparative veterinary hematology: normal and abnormal blood
cells in mammals, birds and reptiles. Iowa State University Press,
Ames, Iowa, USA.
Hudson, P. J., A. P Dobson, and D. Newborn. 1985. Cyclic and
non-cyclic populations of red grouse: a role for parasitism? Pages
77-89 in D. Rollinson and R. M. Anderson, editors. Ecology and
genetics of host-parasite interactions. Academic Press, London,
UK.
Kettle, D. 5. 1982. Medical and veterinary entomology. John
Wiley, New York, New York, USA.
Klicka, 3., and R. M. Zink. 1997. The importance of recent ice
ages in speciation: a failed paradigm. Science 277:1666-1669.
Knowlton, N., L. A. Weigt, L. A. Solorzano, D. E. K. Mills, and
F. Bermingham. 1993. Divergence in proteins, mitochondrial
DNA and reproductive compatibility across the Isthmus of
Panama. Science 260:1629-1632.
Lack, D. 1976. Island biology illustrated by the land birds of
Jamaica. University of California Press, Berkeley, California, USA.
Lowery, R. 5. 1974. Blood parasites of vertebrates of Aldabra.
Philosophical Transactions of the Royal Society of London B
260:577-580.
May, R. M., and R. M. Anderson. 1979. Population biology of
infectious diseases, part 2. Nature 280:455-46 1.
Meril[ddot{a}], J., M. Bj[ddot{o}]rklund, and G. F. Bennett.
1995. Geographic and individual variation in haematozoan
infections in the greenfinch, Corduelis chloris. Canadian Journal
of Zoology 73: 1798-1804.
Norman, R., M. Begon, and R. G. Bowers. 1994. The population
dynamics of microparasites and vertebrates hosts- the importance
of immunity and recovery. Theoretical Population Biology
46:96-119.
Pierce, M. A., A. S. Cheke, and R. A. Cheke. 1977. A survey of
blood parasites of birds in the Mascarene Islands, Indian Ocean.
Ibis 119:451-461.
Prins, H. H. T., and F. J. Weyerhaeuser. 1987. Epidemics in
populations of wild ruminants: anthrax and impala, rinderpest
and buffalo in Lake Manyara National Park, Tanzania. Oikos
49:28-38.
Ricklefs, R. E. 1992. Embryonic development period and the
prevalence of avian blood parasites. Proceedings of the National
Academy of Sciences (USA) 89:4722-4725.
Roitt, I. M., J. Brostoff, and D. K. Male. 1989. Immunology.
Second edition. Lippincott, Philadelphia, Pennsylvania, USA.
Russo, E. A., L. McEntee, L. Applegate, and J. S. Baker. 1986.
Comparison of two methods for determination of white blood
cell counts in macaws. Journal of the American Veterinary
Medical Association 189:1013-1016.
SAS Institute. 1990. SAS user's guide: statistics. Version 6 edition.
SAS Institute, Cary, North Carolina, USA.
Seutin, G., J. Brawn, R. E. Ricklefs, and E. Bermingham. 1993.
Genetic divergence among populations of a tropical passerine,
the Streaked Saltator (Saltator albicollis). Auk 110:117-126.
Seutin, G., N. K. Klein, R. E. Ricklefs, and E. Bermingham.
1994. Historical biogeography of the Bananaquit (Coereba
flaveola) in the Caribbean region: a mitochondrial DNA
assessment. Evolution 48:1041-1061.
Seutin, 0., B. N. White, and P. T. Boag. 1991. Preservation of
avian blood and tissue samples for DNA analyses. Canadian
Ecology                                                                                                                  July, 2000
Journal of Zoology 68:82-90.
Shields, G. F., and A. C. Wilson. 1987. Calibration of
mitochondrial DNA evolution in geese. Journal of Molecular
Evolution 24:212-217.
Sousa, 0. E., and C. M. Herman. 1982. Blood parasites of birds
from Chiriqui and Panama Provinces in the Republic of Panama.
Journal of Wildlife Diseases 18:205-221.
Takahata, N. 1989. Gene genealogy in three related populations:
consistency probability between gene and population trees.
Genetics 122:957-966.
Takahata, N., and M. Nei. 1985. Gene genealogy and variance of
interpopulational nucleotide differences. Genetics 110: 325-344.
Tarr, C. L., and R. C. Fleiseher. 1993. Mitochondrial-DNA
variation and evolutionary relationships in the Amakihi complex.
Auk 110:825-831.
van Riper, C., S. G. van Riper, M. L. Goff, and M. Laird. 1986.
The epizootiology and ecological significance of malaria in
Hawaiian land birds. Ecological Monographs 56: 327-344.
White, E. M., E. C. Greiner, 0. F. Bennett, and C. M. Herman.
1978. Distribution of the hematozoa of Neotropical birds.
Revista Biologia Tropica 26:43-102.
Wirth, W. W. 1974. The West Indian sandfiies of the genus
Culicoides (Diptera: Ceratopogonidae). USDA Agricultural
Research Service, Washington, D.C., USA.
Yezerinac, S. M., and P. J. Weatherhead. 1995. Plumage
coloration, differential attraction of vectors and haematozoa
infections in birds. Journal of Animal Ecology 64:528-537.
Zajicek, D., and M. Mauri Mendez. 1969.
Hemopar[acute{A}]sitos de algunos animales de Cuba. Poeyana,
Series A, 66:1-10. 
COPYRIGHT 2000 Ecological Society of America
in association with The Gale Group and LookSmart.
COPYRIGHT 2000 Gale Group