Introduction
Anthropogenic environmental changes continue to drive the increasing
frequency of zoonotic disease emergence – diseases transmitted between
wildlife and human populations (Allen et al. 2017; Gibb et
al. 2020). Zoonotic diseases emerge from a suite of processes that
require spatio-temporal overlap between humans and infected hosts or
vectors for spillover occurrence. Urban ecosystems are a hotbed for
emerging zoonoses due to opportunistic species adapted to anthropic
landscapes that serve as zoonotic hosts and the existence of interfaces
between humans, wildlife, and/or vectors where contact and pathogen
spillover can occur (Alirol et al. 2011; Hassell et al.2017). Through movement, wildlife hosts influence contact rates between
vectors, pathogens, and other hosts, ultimately shaping the distribution
of zoonotic hazards in complex multi-host-pathogen systems.
Urban wildlife movement patterns are driven by landscape-dependent and
independent fragmentation. Landscape-dependent fragmentation (LDF)
alters the physical configuration of habitat patches, preventing species
from moving freely across the landscape as patches become increasingly
isolated (Magle et al. 2014; Berger-Tal & Saltz 2019). As
natural land is converted to impervious surface, wildlife hosts
aggregate in remaining habitat fragments leading to increased contact
rates and spatial hotspots of transmission and spillover (Wilkinsonet al. 2018). Landscape-independent fragmentation (LIF) results
from anthropogenic disturbances which change animals’ perception and use
of their environment (Berger-Tal & Saltz 2019). For example, resource
subsidies lead to LIF whereby fertilized vegetation, gardens, or waste
byproducts (Becker et al. 2015; Langley et al. 2021)
provide stable, often predictable, resources which influence wildlife
host movement (Ossi et al. 2020; Ranc et al. 2020) and
contact patterns at the human interface, where zoonotic hazard becomes
risk.
Tick-borne disease emergence is intertwined with land use change and
habitat fragmentation (Diuk-Wasser et al. 2021) and tick-borne
diseases are the most common vector-borne zoonoses in temperate North
America, Europe, and Asia. In the United States, Lyme disease (LD), a
bacterial infection caused by Borrelia burgdorferi sensu
stricto , affects 400,000 people annually (Kugeler et al. 2021;
Schwartz et al. 2021). Historically, LD was associated with the
incursion of suburban and exurban development into rural areas (Barbour
& Fish 1993). More recently, ticks expanded their geographic range with
climate change (Sonenshine 2018; Ogden et al. 2021) and by
occupying diverse and novel landscapes such as cities (VanAcker et
al. 2019). The urban expansion of LD occurs through two pathways
increasing human risk of exposure to the LD hazard: (1) vegetation
increases in cities undergoing de-urbanization linked with population
decline and land abandonment (Eskew & Olival 2018; Richards & Belcher
2019), or through cities adopting urban greening strategies (Yanget al. 2014; Halsey et al. 2022); and (2) expanding urban
matrix into surrounding natural habitats. Both pathways increase tick
habitat and the prevalence of wildland-urban interfaces where species
richness is dominated by synanthropic wildlife species, often tick hosts
or pathogen reservoirs (Gibb et al. 2020), and human exposure to
ticks is high (Diuk-Wasser et al. 2021).
In the eastern and midwestern US, the establishment and persistence of
the LD vector Ixodes scapularis ticks is supported by large
mammal hosts. For adult I. scapularis , white-tailed deer
(Odocoileus virginianus , hereafter deer) are the primary
reproductive stage host (Barbour & Fish 1993; Rand et al. 2004;
Ostfeld et al. 2018). Importantly, while deer amplify the vector
for LD (Carpi et al. 2008; Cagnacci et al. 2012), deer are
not susceptible to B. burgdorferi infection and do not supportB. burgdorferi transmission (Telford III et al. 1988).
Studies predicting the distribution of I. scapularis ticks and LD
over large geographic areas typically utilize broad-scale climatic and
landcover variables (Estrada-Peña 1998; Diuk-Wasser et al. 2006;
Soucy et al. 2018). This approach, however, lacks power to
predict the zoonotic hazard at fine-spatial scales because it ignores
local LD ecology and the modulating role of host movement in fragmented
environments at different spatial (home range, fine scale movements) and
temporal (seasonal, diel) scales.
Available niches for parasites are shaped by movement and resource
selection of wildlife hosts (Ezenwa et al. 2016). Thus, the
probability of tick population establishment is affected by the hosts’
scale of response to the landscape – for example, the scale at which
deer select and establish their home range. Wildlife resource selection
spans multiple spatial scales and is hierarchically nested (Johnson
1980), where broad-scale selection constrains fine-scale selection
(Senft et al. 1987). Multi-scale selection is integral to
consider in urban landscapes where deer often display urban-adapted
behavior such as foraging close to households (Swihart et al.1995; Kilpatrick et al. 2000; Grund et al. 2002) while
resting in forest patches. If deer habitat selection encompasses
anthropogenic resources, residential areas may experience enhanced
exposure to infected ticks. Because ticks passively fall from hosts upon
engorgement, host movement speed and directionality through varying
landcovers affect tick spatial clustering and ticks’ likelihood of
survival post-feeding. Further, the burden of I. scapularis ’ life
stages on deer varies with tick phenology such that seasonal movement
patterns exhibited by deer may differentially affect which life stages
are dispersed, when, and to where.
This study examines fine-scale deer movement in an urban, fragmented
borough of New York City (NYC) to determine how deer movement links to
local distributions of I. scapularis ticks and heightened LD
hazard. We employ a multi-scale hierarchical resource selection
framework to (i) examine deer’s scale of response to anthropogenic
landscape features when establishing their home-range (i.e., responses
to LDF in 2nd order selection), (ii) determine fine-scale spatiotemporal
effects of the urban landscape on within-home range habitat selection
and avoidance (i.e., responses to LIF in 3rd order selection), and (iii)
assess the spatial congruence between deer seasonal habitat selection
and LD hazard. We expect high intensity development will restrict deer
home range selection at fine-spatial scales through limiting available
natural habitat and creating movement barriers. We expect deer to
exhibit diel variation when selecting for features within the home range
and to avoid highly anthropic areas during periods of heightened human
activity. We further hypothesize that vegetated neighborhoods nested
within areas connected to natural habitats will attract deer to
fine-scale foraging resources and support the microclimate for tick
survival, leading to higher tick occupancy compared to neighborhoods
that are less accessible to deer and/or present more hostile
microclimates for I. scapularis.
With the majority of the human population residing in cities (United
Nations, Department of Economic and Social Affairs 2018) and the recent
increase of tick-borne diseases in urban areas globally (Hamer et
al. 2012; Rizzoli et al. 2014; Hansford et al. 2017,
2021; Heylen et al. 2019; VanAcker et al. 2019; Simmonset al. 2020; Sormunen et al. 2020), there is an urgent need
to understand the urban ecology of tick-borne disease emergence and the
determinants of heterogeneity in tick-borne disease hazard. Here, we
pair deer movement and tick surveillance data to provide the first study
that links deer movement behavior to tick distribution across a highly
urban landscape. We conclude that resource selection at different
spatial scales enables urban deer to navigate fragmented habitat and
that sex and individual-based responses to human activity characterize
differences in tick distribution capacity.