ORIGINAL PAPER Effects of barred owl (Strix varia) range expansion on Haemoproteus parasite assemblage dynamics and transmission in barred and northern spotted owls (Strix occidentalis caurina) Krista E. Lewicki • Kathryn P. Huyvaert • Antoinette J. Piaggio • Lowell V. Diller • Alan B. Franklin Received: 4 March 2014 / Accepted: 9 December 2014 / Published online: 20 December 2014 Ó Springer International Publishing Switzerland (outside the USA) 2014 Abstract The range of the barred owl (Strix varia) has expanded westward over the past century and now entirely overlaps the range of the federally threatened northern spotted owl (S. occidentalis caurina) in the Pacific Northwest. We compared Haemoproteus blood parasite assemblages among northern spotted owls in their native range and barred owls in both their native and invasive ranges to evaluate predictions of five hypotheses about parasites and biological invasions: (1) Enemy Release, where hosts benefit from a loss of parasites in their invasive range, (2) Enemy of My Enemy, where invasive hosts introduce parasites to naı ¨ve native hosts, (3) Parasite Spillback, where invasive hosts act as a new reservoir to native parasites, (4) Increased Susceptibility, where native hosts introduce parasites to naı ¨ve invasive hosts, and (5) Dilution Effect, where invasive species act as poor hosts to native parasites and decrease the density of potential hosts in their invasive range. We used haplotype network analyses to identify one haplotype common to both owl species throughout North Amer- ica, three more haplotypes that appeared to be isolated to the barred owl’s historic range, and a fifth haplotype that was only found in California. Based on infection status and parasite diversity in eastern and western barred owl populations, we found strong support for the Enemy Release Hypothesis. Northern spotted owls had higher parasite diversity and probability of infection than sympatric barred owls, offering some support for the Parasite Spillback and Dilution Effect Hypotheses. Overall, this study demonstrates the complexity of host-parasite relationships and high- lights some of the ways in which species’ range expansions may alter such relationships among both invasive and native hosts. Keywords Blood parasite Enemy Release Hypothesis Haemoproteus Invasive species Parasite Spillback Hypothesis Dilution Effect Hypothesis Range expansion Electronic supplementary material The online version of this article (doi:10.1007/s10530-014-0828-5) contains supple- mentary material, which is available to authorized users. K. E. Lewicki K. P. Huyvaert Department of Fish, Wildlife, and Conservation Biology, Colorado State University, 1474 Campus Delivery, Fort Collins, CO 80523, USA e-mail: [email protected]K. P. Huyvaert e-mail: [email protected]A. J. Piaggio A. B. Franklin (&) USDA/APHIS/WS/National Wildlife Research Center, 4101 LaPorte Avenue, Fort Collins, CO 80521, USA e-mail: [email protected]A. J. Piaggio e-mail: [email protected]L. V. Diller Green Diamond Resource Company, P.O. Box 68, Korbel, CA 95550, USA e-mail: [email protected]123 Biol Invasions (2015) 17:1713–1727 DOI 10.1007/s10530-014-0828-5
15
Embed
Effects of barred owl (Strix varia) range expansion on ......was located near the barred owl’s invasion front, where the effects of host range expansion on parasite assemblages are
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
ORIGINAL PAPER
Effects of barred owl (Strix varia) range expansionon Haemoproteus parasite assemblage dynamicsand transmission in barred and northern spotted owls(Strix occidentalis caurina)
Krista E. Lewicki • Kathryn P. Huyvaert •
Antoinette J. Piaggio • Lowell V. Diller •
Alan B. Franklin
Received: 4 March 2014 / Accepted: 9 December 2014 / Published online: 20 December 2014
� Springer International Publishing Switzerland (outside the USA) 2014
Abstract The range of the barred owl (Strix varia)
has expanded westward over the past century and now
entirely overlaps the range of the federally threatened
northern spotted owl (S. occidentalis caurina) in the
Pacific Northwest.We comparedHaemoproteus blood
parasite assemblages among northern spotted owls in
their native range and barred owls in both their native
and invasive ranges to evaluate predictions of five
hypotheses about parasites and biological invasions:
(1) Enemy Release, where hosts benefit from a loss of
parasites in their invasive range, (2) Enemy of My
Enemy, where invasive hosts introduce parasites to
naıve native hosts, (3) Parasite Spillback, where
invasive hosts act as a new reservoir to native
parasites, (4) Increased Susceptibility, where native
hosts introduce parasites to naıve invasive hosts, and
(5) Dilution Effect, where invasive species act as poor
hosts to native parasites and decrease the density of
potential hosts in their invasive range. We used
haplotype network analyses to identify one haplotype
common to both owl species throughout North Amer-
ica, three more haplotypes that appeared to be isolated
to the barred owl’s historic range, and a fifth haplotype
that was only found in California. Based on infection
status and parasite diversity in eastern and western
barred owl populations, we found strong support for
the Enemy Release Hypothesis. Northern spotted owls
had higher parasite diversity and probability of
infection than sympatric barred owls, offering some
support for the Parasite Spillback and Dilution Effect
Hypotheses. Overall, this study demonstrates the
complexity of host-parasite relationships and high-
lights some of the ways in which species’ range
expansions may alter such relationships among both
Electronic supplementary material The online version ofthis article (doi:10.1007/s10530-014-0828-5) contains supple-mentary material, which is available to authorized users.
K. E. Lewicki � K. P. HuyvaertDepartment of Fish, Wildlife, and Conservation Biology,
Covariates included in the model are described in Table 2. Only models with Akaike weights C0.05 are shown for each analysisa R2 = maximum re-scaled R2 generated through PROC LOGISTIC for logistic regression models; R2 values from generalized
linear models (PROC GLM) for linear regression modelsb DAICc = difference in AICc between a given model and the top-ranked modelc Akaike weight = probability that a given model is the best supported model given the model set and the datad Pr(Inf) = Probability of infectione A posteriori model
1722 K. E. Lewicki et al.
123
and a closely related (1 bp different) haplotype has
been found in owls from North America, Africa, and
Europe (Ishak et al. 2008), suggesting that this is a
common, cosmopolitan haplotype.
Interestingly, western barred owls had a lower
probability of infection with H4, contradicting the
prediction that probability of infection should be
similar among eastern and western barred owls for
shared haplotypes. It is unlikely that the observed
difference in probability of infection of H4 was driven
by this haplotype being naturally rarer in the Pacific
Northwest because northern spotted owls had a high
prevalence of this haplotype. Alternative explanations
may include (1) western barred owls may have new
behavioral adaptations and habitat associations that
have decreased their exposure to Haemoproteus
vectors, and/or (2) western barred owls may have
lower susceptibility to infection with these avian blood
parasites.
The observed higher prevalence in eastern versus
western barred owls is similar to the pattern observed
by Ishak et al. (2008) and supports the notion that
suitable habitat for and/or abundance of Haemopro-
teus vectors is heterogeneous and fragmented across
North America.Haemoproteus parasites require warm
temperatures for development in biting midge vectors
(Valkiunas 1996). Cold temperatures in the Rocky and
Cascade Mountain ranges may hinder rarer Haemo-
proteus haplotypes from accompanying barred owl
hosts with invasion of the west because of the
parasites’ dependence on these warmer temperatures;
sampling populations from the invasion corridor along
southern Canada would allow for a better test of this
hypothesis.
Post-invasion hypotheses
Of the two haplotypes detected in northern spotted and
western barred owls, one (H4) appears to be common
and cosmopolitan, while the other (H5) was detected
only in California. We found no evidence that either of
these haplotypes originated from eastern North Amer-
ica, which fails to support the EEH prediction that
barred owls would introduce novel Haemoproteus
haplotypes to northern spotted owls through range
expansion. The California-specific haplotype (H5) is
noteworthy because it has not been documented in
other studies (e.g., Perkins and Schall 2002; Ricklefs
and Fallon 2002) and is genetically distant (C13 bp
different) from Haemoproteus haplotypes detected in
previous studies. Discovery of this haplotype lends
support to a prediction of the ISH because it suggests
that barred owls may be acquiring new Haemoproteus
haplotypes as they expand their range (Table 1);
however, we observed that northern spotted owls had a
higher probability of infection and infection intensity
than western barred owls, providing strong support
against ISH predictions that these parasites would
more negatively affect western barred than northern
spotted owls (Table 1). Intuitively, these results are
not surprising; if Haemoproteus were having a strong,
negative impact on western barred owl populations, it
Fig. 2 Predicted probabilities of infection (a) and infection
intensities (b) of Haemoproteus in northern spotted and westernbarred owls from northwest California. Probabilities of infection
were estimated from a logistic regression model where owl
population and the natural log of distance to the coast were
covariates [Pr(Inf) & PO ? LnDC]; estimates for northern
spotted owls are solid lines and for barred owls are dashed
lines, and 95 % CI are shown in gray. Infection intensities were
estimated from a general linear model in which population, sex,
and a population by sex interaction were covariates
(Intensity & PO ? SX ? PO 9 SX)
Effects of barred owl (Strix varia) range expansion 1723
123
is doubtful that these populations would invade the
Pacific Northwest as successfully as has been observed
in the past several decades (US Fish and Wildlife
Service 2011).
Parasite similarity, probability of infection, and
infection intensity results all lend support to PSH
predictions that barred owls acquire new parasites as
they expand their range and that these parasites would
more negatively affect northern spotted owls
(Table 1); however, an alternative explanation for
the low probability of infection, parasite diversity, and
infection intensity observed in western barred owls is
that western barred owls are poor hosts to parasites in
their invasive range, which supports predictions of the
DEH (Table 1). Without historic data on Haemopro-
teus haplotypes of northern spotted owls, we could not
evaluate a central component of these hypotheses—
how Haemoproteus prevalence has changed in north-
ern spotted owls since the barred owl range expansion.
The conservation implications of these two hypotheses
differ markedly: under the PSH, the presence of barred
owls may negatively affect northern spotted owls by
perpetuating high Haemoproteus prevalence in north-
ern spotted owls while, under the DEH, the presence of
barred owls may indirectly benefit northern spotted
owls by decreasing Haemoproteus prevalence. Given
these implications, it will be important to continue
monitoring for changes inHaemoproteus assemblages
in these two host species to better understand the
impacts of barred owl range expansion on northern
spotted owl health.
Inferences about the transmission of parasites
between native and invasive hosts would also benefit
from future studies evaluating the role of additional
biological variables in shaping transmission dynam-
ics. In our study, the highest-ranked probability of
infection model included the natural log of distance to
coast (LnDC, Table 3), which may be related to the
regional distribution of Haemoproteus vectors. Gen-
erally, the biting midge vectors ofHaemoproteus have
higher reproductive success at warmer temperatures
(Mellor et al. 2000), and, in our study area, inland sites
were characterized by cool, wet winters and hot, dry
summers, while coastal sites experienced milder
temperatures and higher year-round precipitation
(Ting 1998; Franklin et al. 2000). It is possible that
the warmer temperatures inland during the peak
parasite transmission season led to increased vector
abundance at these sites and, in turn, higher levels of
parasite transmission. Vector abundance has also been
used to explain differences in parasite prevalence
between avian assemblages of disturbed and undis-
turbed sites. Chasar et al. (2009) found lower blood
parasite prevalence in areas with high disturbance,
possibly because areas of high disturbance provide
less suitable habitat for vectors than undisturbed areas.
Similar reasoning may explain why many of our
highest-ranking models included management inten-
sity (Table 3), which predicted a higher probability of
infection in lower intensity management areas. How-
ever, distance to coast and management intensity were
highly correlated because of our opportunistic sam-
pling; most of our samples on low-intensity manage-
ment areas were further inland than those on high-
management areas. Although distance to coast was
more strongly supported in our analyses, understand-
ing the differing effects of these two covariates on
Haemoproteus transmission would be improved by
more even sampling between sites with low and high
intensity management at coastal and inland areas. We
did not find a similar relationship between infection
intensity and distance to coast (Table 3); sex and
species appeared to have the strongest effects with
female northern spotted owls having a higher infection
intensities than males or either sex in barred owls.
Other studies have detected sex-biased parasitism of
blood parasites in owl hosts (e.g., Korpimaki et al.
1993), which may be driven by differences in life
history traits and exposure to environmental stressors
between the sexes.
Overall, our results suggest that Haemoproteus
assemblage dynamics of northern spotted owls are not
solely influenced by the presence or absence of invasive
barred owls, and evaluating the role of additional
biological variables may help broaden our global
understanding of the relationships among invasive and
native hosts, their parasites, and the environment.
The true cost of parasitism?
We found that northern spotted owlsweremore likely to
be infected with Haemoproteus haplotypes than sym-
patric barred owls, but our study did not directly
evaluate if and to what extent parasite infection status
and intensity influence northern spotted owl and barred
owl fitness. Although generally considered to be
relatively innocuous in their avian hosts,Haemoproteus
1724 K. E. Lewicki et al.
123
parasites can become pathogenic when coupled with
additional stressors (Remple 2004), such as competition
with barred owls.
We found strong support for the ERH for Haemo-
proteus parasites of invasive barred owls, but the true
cost of Haemoproteus infections also has implications
for invasive barred owl fitness. If Haemoproteus
parasites are relatively innocuous to their barred owl
hosts, the loss of these parasites among western barred
owls may not have much biological relevance. Nev-
ertheless, our results demonstrate an important pattern
that may be occurring among more cost-demanding
parasites that were not examined in this study. Thus,
we echo Ishak et al. (2008) suggestion that follow-up
studies should evaluate the relationship of infection
status with immunocompetence, estimated survival
and reproductive rates for infected compared to
uninfected birds, and competitive interactions of both
northern spotted and barred owls.
Finally, we compared parasite haplotype diversity in
this study under the assumption that host populations
infected with a low diversity of parasites were more
immunologically competent than host populations
infected with a high diversity of parasites. However,
Hudson et al. (2006) argue that high native parasite
diversity may be an indicator of the health of a given
ecosystem because it is often a result of long chains of
multispecies connections that can only be present in
healthy ecosystems. We detected greater haplotype
diversity among northern spotted owls than western
barred owls, and Ishak et al. (2008) reported a high
diversity of Leucocytozoon blood parasite lineages
among northern spotted owls relative to Leucocytozoon
assemblages of other owl species across theworld. If the
blood parasite infections we detected among northern
spotted owls are a result of host-vector-parasite inter-
actions that have co-evolved over a long period of time,
then our study suggests that Haemoproteus infections
may be benign if not beneficial in northern spotted owls.
Svensson-Coelho et al. (2013) found that avian host
species with high Haemoproteus prevalence showed
low Plasmodium prevalence and vice versa. One
explanation for this observed pattern is that infection
of parasites from one genus may inhibit infection of
parasites from another. In the context of our study
system, it is possible that northern spotted owls have
adapted to high Haemoproteus prevalence as part of a
defense mechanism against more virulent Plasmodium
parasites, which are seemingly rare among northern
spotted owls (Gutierrez 1989; Ishak et al. 2008; Lewicki
2013). Future studies on this concept in northern spotted
owls would help elucidate both the role that blood
parasites have on northern spotted owl fitness and the
complex relationships between blood parasites and
avian hosts in the face of invasion in general.
Acknowledgments We thank Peter Carlson and Jeremy
Rockweit of Klamath Biological Research Station; Mark
Higley and the Hoopa Valley Tribe; Green Diamond Resource
Company; Robert Feamster and Sierra Pacific Industries; and
Laurie Clark and the National Council for Air and Stream
Improvement, Inc., for allowing us to conduct research on their
lands and for assistance in collecting western owl samples. The
Avian Conservation Center (South Carolina), Wildcare
Foundation (Oklahoma), Avian Haven (Maine), Carolina
Raptor Center (North Carolina), Audubon of Florida (Florida),
The Raptor Center (Minnesota), Tri-State Bird Rescue &
Research (Delaware), and Alabama Raptor Center (Alabama)
collected all the eastern barred owl samples analyzed in this
study. Additional field and laboratory assistance came from
Constanza Toro, Annie Kellner, Matthew Hopken, Nikki Crider,
and JeremyDertien.Dr.ThomasGidlewski provided access to his
microscope; Nic Berrong assisted in installing and navigating the
i-Solution Lite software; and Drs. Ellen Martinsen and Robert
Ricklefs provided positive control samples. Finally, Drs. Liba
Pejchar, Brian Foy,KenBurnham, andAnnHess assisted in study
design and analysis and 2 anonymous reviewers provided helpful
comments and suggestions that greatly improved ourmanuscript.
This work was conducted under the auspices of the Colorado
State University Institutional Animal Care and Use Committee
protocol #10-1818A. Funding and additional support for this
project was provided by the U.S.D.A. Forest Service Region 5
contract 11-CS-11052007-319 and Colorado State University.
References
Atkinson CT, Van Riper CIII (1991) Pathogenicity and epizo-
otiology of avian haematozoa: Plasmodium, Leucocyto-
zoon, and Haemoproteus. In: Loye JE, Zuk M (eds) Bird-
parasite interactions: ecology, evolution, and behavior.