Access to this full-text is provided by Springer Nature.
Content available from Scientific Reports
This content is subject to copyright. Terms and conditions apply.
1
Vol.:(0123456789)
Scientic Reports | (2024) 14:12479 | https://doi.org/10.1038/s41598-024-62384-6
www.nature.com/scientificreports
Broadscale spatial synchrony
in a West Nile virus mosquito vector
across multiple timescales
Lindsay P. Campbell
1,2*, Amely M. Bauer
1,2, Yasmin Tavares
3, Robert P. Guralnick
4 &
Daniel Reuman
5
Insects often exhibit irruptive population dynamics determined by environmental conditions. We
examine if populations of the Culex tarsalis mosquito, a West Nile virus (WNV) vector, uctuate
synchronously over broad spatial extents and multiple timescales and whether climate drives
synchrony in Cx. tarsalis, especially at annual timescales, due to the synchronous inuence of
temperature, precipitation, and/or humidity. We leveraged mosquito collections across 9 National
Ecological Observatory Network (NEON) sites distributed in the interior West and Great Plains region
USA over a 45-month period, and associated gridMET climate data. We utilized wavelet phasor
mean elds and wavelet linear models to quantify spatial synchrony for mosquitoes and climate and
to calculate the importance of climate in explaining Cx. tarsalis synchrony. We also tested whether
the strength of spatial synchrony may vary directionally across years. We found signicant annual
synchrony in Cx. tarsalis, and short-term synchrony during a single period in 2018. Mean minimum
temperature was a signicant predictor of annual Cx. tarsalis spatial synchrony, and we found a
marginally signicant decrease in annual Cx. tarsalis synchrony. Signicant Cx. tarsalis synchrony
during 2018 coincided with an anomalous increase in precipitation. This work provides a valuable step
toward understanding broadscale synchrony in a WNV vector.
Mosquito borne pathogens are a major threat to human and veterinary health1. ese pathogens are nested within
dynamic systems that include one or more arthropod vectors and vertebrate hosts that must interact in space and
time for transmission to be maintained in the natural environment2. Several factors contribute to the distribu-
tion and magnitude of transmission hazard in an area, including intrinsic population dynamics of vectors and
hosts, and extrinsic environmental conditions that can aect the timing, abundances, and distributions of disease
system components3. Mosquito vectors are one component in these systems, and like all insects, are ectotherms,
with their distributions and abundances closely linked to exogenous environmental variables4. Because of these
linkages and the role of mosquito vectors in pathogen transmission, environmental drivers of mosquito popu-
lation dynamics are oen the focus of investigations attempting to understand outbreaks or epizootic events5.
While informative, the majority of these studies have focused on environmental eects on local-scale variation
in abundances and do not take into account how mosquito populations uctuate together over space and time
at broader scales and extents, despite the potential for synchronous or asynchronous population dynamics in
dierent geographic locations to aect distributions of pathogen transmission6–8.
Spatial synchrony, i.e. similarities in temporal uctuations of populations occurring across geographically
distinct locations9, is an ecological phenomenon observed across multiple taxa and across local to broad geo-
graphic scales. In insects, spatial synchrony has most oen been studied in groups that have periodic “outbreaks”,
e.g. spongy moths or larch budmoths, and crop pests10–15. However, spatial synchrony has been less studied in
mosquito vectors of human and veterinary pathogens16. Mechanistic drivers of spatial synchrony consist of dis-
persal between populations, exogenous environmental conditions (through a process called the Moran eect),
and trophic interactions of the focal species with another species that exhibits synchrony or that is very mobile9.
Given known correlations between the environment, particularly temperature and precipitation, and mosquito
OPEN
1Florida Medical Entomology Laboratory, University of Florida, Vero Beach, FL 32962, USA. 2Department of
Entomology and Nematology, University of Florida, Gainesville, FL 32611, USA. 3Department of Ecology,
Evolution, and Environmental Biology, Graduate School of Arts and Sciences, Columbia University, New York,
NY 10025, USA. 4Florida Museum, University of Florida, Gainesville, FL 32611, USA. 5Department of Ecology and
Evolutionary Biology and Center for Ecological Research, University of Kansas, Lawrence, KS 66047, USA. *email:
lcampbell2@u.edu
Content courtesy of Springer Nature, terms of use apply. Rights reserved
2
Vol:.(1234567890)
Scientic Reports | (2024) 14:12479 | https://doi.org/10.1038/s41598-024-62384-6
www.nature.com/scientificreports/
population dynamics17,18, mosquito species are strong candidates for investigations of environmental causes of
spatial synchrony across dierent spatiotemporal scales19.
One challenge to previous studies of spatial synchrony in mosquito populations is the need for relatively long
and consistent collection records across disparate geographic locations. e result is that the majority of prior
population studies occur over small geographic areas or elevational gradients8,16,19–21. However, broadscale spatial
synchrony is not uncommon in other insect species. Sheppard etal.13 found broadscale spatial synchrony in
the timing of adult ight onset of aphid species driven by climate conditions at multiple timescales, and Haynes
and Walter12 highlight how understanding spatial synchrony in insects across scales can be useful to informing
pest management decisions. In addition, vector borne disease dynamics in multiple systems have been linked
to large-scale climate conditions, including anomalous or unusual conditions, for example: Ri Valley fever
virus and heavy precipitation resulting from El Niño in East Africa22; broadscale chikungunya incidence associ-
ated with unusually dry conditions23; and dengue incidence and El Niño-associated dry conditions or elevated
temperatures in ailand24,25, among others26. Needed are studies to determine if mosquito vector abundances
show strong spatially synchronous dynamics at broader extents and timescales and whether their dynamics are
associated with climate conditions which oen also vary over large spatial extents.
e National Ecological Observatory Network (NEON) conducts routine mosquito collections at 47 terrestrial
core and terrestrial gradient sites in the United States27,28. e sampling design, temporal resolution, and proto-
cols of mosquito collections at NEON sites were designed specically for the purpose of standardized monitor-
ing of mosquito population abundances, demography, diversity, and phenology, including comparisons among
NEON sites at regional to continental scales28. is monitoring program uniquely provides a means to generate
consistent and continuous time series collections to better understand mosquito population dynamics across
broad geographic scales. Here, we use this resource to investigate whether signicant spatial synchrony exists in
temporal population dynamics of the West Nile virus (WNV) vector species Culex tarsalis across 90 mosquito
traps across 9 widely distributed NEON sites that have the most continuous collection of data. In particular, we
quantify eects of exogenous climate variables on spatial synchrony between study locations.
Culex tarsalis Coquillett is a key vector of WNV (family Flaviviridae, genus Flavivirus), the leading cause of
mosquito-borne disease in humans in the United States (CDC 2021a). e virus is maintained in the natural
environment between mosquito vectors and avian hosts, and “spills over” to humans and other animals (Camp-
bell etal. 2002, Reisen 2013). Cx. tarsalis are multivoltine with a wide geographic distribution29. e species
is considered a particularly important vector of WNV in the midwest and western regions of the U.S. where it
is oen associated with agricultural irrigation or ditches30. Mosquitoes become active in late spring and early
summer. In temperate zones, adult females overwinter and have been observed in underground sheltered areas,
but they remain active year-round in warmer climates (Walter Reed Biosystematics Unit, 2024).
We hypothesized that we may nd signicant spatial synchrony in Cx. tarsalis abundances at multiple tem-
poral scales, including an annual eect. We also expected that average minimum or maximum temperature,
humidity values, and precipitation may contribute to annual spatial synchrony across the study period. We tested
this directly to determine which climate factors are most important. We also tested if unusual or extreme weather
events, such as widespread heavy precipitation, which occurred in Summer 2018, contribute to signicant spatial
synchrony at shorter timescales. Temperature is a well-known driver of insect population dynamics and has been
shown to aect Cx. tarsalis mosquito development timing31, while precipitation is critical in generating aquatic
habitats required for multiple mosquito life stages4. Humidity can aect mosquito hydroregulation, impacting
activity and survival32. Investigating the population dynamics of vector species and how they uctuate together
across large spatial scales has the potential to provide useful knowledge about risks mosquitoes pose, if outbreaks
are predictable, and how those risks can be mitigated.
Results
Following QA/QC assembled monthly time series of estimated abundances of Cx. tarsalis mosquitoes we included
nine NEON stations in our analyses over a 45month period from 2016 to 2019, and associated gridMET climate
data (Figs.1 and 2). We utilized wavelet phasor mean elds (wpmf) to quantify spatial synchrony at multiple
timescales for mosquitoes and climate (see Methods for full details of cleaning and analysis approach).
Results of wavelet phasor mean eld (wpmf) analyses indicated signicant spatial synchrony in Cx. tarsalis
abundances at an annual timescale across the study period; and during a short-term, isolated event at the 2 to
3month timescale band between August and September 2018 (Fig.3). Although the wpmf output also showed
seasonal spatial synchrony (i.e. ~ every six months), this result was likely a harmonic of the annual timescale
synchrony.
e plot of the wpmf for Cx. tarsalis showed a potential decrease in the strength of annual-timescale spatial
synchrony across the study period. Comparing this possibility to an appropriate null hypothesis based on a
bootstrapping technique (Methods) indicated a marginally signicant (e.g. p-value between 0.05 and 0.10)
decrease in the slope of Cx. tarsalis synchrony at the averaged 10 to 14month timescale (p-value = 0.096), but
when considering only the 12month timescale on its own (p-value = 0.242). Wavelet methods are well known
to commit “leakage,” whereby periodic variational content in time series at a given timescale is detected also at
a range similar timescales. is is because of the nite length of time series, and tradeos between a wavelet’s
abilities to localize a phenomenon in time and timescale space. For reasons of leakage, the results here using
the 10 to 14month band are probably more indicative of annual-timescale behavior than are the results using
solely the 12-month timescale. Analogous slopes for climate variables at the same timescales were not signicant
(Supplementary Table1), i.e., there was no evidence for changes in the strength of synchrony of climate variables.
e wpmf plots for four out of ve environmental variables showed patterns of signicant spatial syn-
chrony at similar timescales and time periods to Cx. tarsalis (Figs.4 and 5B; Fig.5A reproduces Fig.3 for visual
Content courtesy of Springer Nature, terms of use apply. Rights reserved
3
Vol.:(0123456789)
Scientic Reports | (2024) 14:12479 | https://doi.org/10.1038/s41598-024-62384-6
www.nature.com/scientificreports/
comparison). As expected, cumulative precipitation and temperature variables showed signicant annual syn-
chrony, with both mean minimum and maximum temperature demonstrating the strongest synchrony across
the entire study period. Average minimum and maximum mean VPD showed signicant synchrony across small
portions of the study period at the annual timescale. Cumulative precipitation also showed signicant spatial
synchrony between August and September 2018 (Fig.5B).
Results from wavelet linear model tests showed signicant eects of synchrony in mean minimum tem-
perature on synchrony in Cx. tarsalis at the 10 to 14month time scale (p-value = 0.049). Mean minimum vapor
pressure decit index, i.e., greater humidity, was marginally signicant (p-value = 0.090), and mean maximum
temperature (p-value = 0.125), mean maximum vapor pressure decit index (p-value = 0.472), and cumulative
precipitation (p-value = 0.725) were not signicant. Mean minimum temperature explained a large portion of
the synchrony (99.490%) with low cross-terms (-1.212), and residuals (1.722). A rank plot with corresponding
Figure1. Map of NEON sites included in analyses. Abbreviations for sites follow NEON conventions. Yellow
circles indicate terrestrial core sites and orange circles indicate terrestrial gradient sites. Background aerial map
Earth Start Geographics SIO, Microso Corporation© 2024, available through Bing Maps and accessed through
QGIS v 3.16.16.
Figure2. Time series of monthly mean number of Cx. tarsalis per trap hour for each NEON site.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
4
Vol:.(1234567890)
Scientic Reports | (2024) 14:12479 | https://doi.org/10.1038/s41598-024-62384-6
www.nature.com/scientificreports/
band tests for mean minimum temperature and wavelet linear model results for mean minimum temperature
are available in Supplementary Fig.1 and Supplementary Table2.
Although spatial synchrony in cumulative precipitation was nominally signicant between August and Sep-
tember 2018 at the 2 to 3month timescale, the wavelet linear model test was not signicant for 2 to 3month
timescales. is result was expected, given that synchrony in Cx. tarsalis was isolated to a short-term event and
periodicity was not present at the 2 to 3month timescale; wavelet linear models test for consistency of phase
Figure3. Wavelet phasor mean eld plot for Cx. tarsalis abundances. e study period is represented on
the x-axis, and the timescale of spatial synchrony is on the y-axis. Color corresponds to the strength of phase
synchrony in the data at each time and timescale; so areas in red indicate stronger synchrony. Contour lines
indicate statistical signicance at 95% condence level.
Figure4. Wavelet phasor mean eld (wpmf) plots for average minimum and maximum temperature and
average minimum and maximum vapor pressure decit values.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
5
Vol.:(0123456789)
Scientic Reports | (2024) 14:12479 | https://doi.org/10.1038/s41598-024-62384-6
www.nature.com/scientificreports/
relationships between response and putative predictor variables over the duration of the time series and across
sampling locations, and apparently what happened on 2 to 3month timescales in our Cx. tarsalis data was
instead a one-o event. However, observations of precipitation values coinciding with this time period revealed
that anomalously high precipitation occurred across several study sites, and a visual inspection of estimated
abundance values across NEON sites showed an increase in Cx. tarsalis during this time period across multiple
sites (Fig.5C, see Data Availability).
Discussion
is study provides some evidence of broad scale spatial synchrony in Cx. tarsalis abundances, and results
demonstrate that Cx. tarsalis populations can uctuate together at an annual timescale and during short-term,
isolated events. is work highlights the need for continued investigation to understand spatiotemporal and
synchronous dynamics of mosquito vectors and their drivers. ese dynamics, which our results suggest may
be predictable, may help in understanding broader WNV system dynamics.
Our nding of signicant spatial synchrony at an annual timescale was expected given the seasonal dynam-
ics of mosquitoes and insects in general33,34. Specically, we found that mean minimum temperatures were a
signicant predictor of synchrony in Cx. tarsalis abundances at an annual timescale, i.e., mosquito synchrony
Figure5. (A,B) Wavelet phasor mean eld plots for Cx. tarsalis monthly mean number of mosquitoes per
trap hour (reproducing Fig.3) and monthly cumulative precipitation for the same time periods and locations.
Time on the x-axis represents the monthly time series from April 1, 2016 to December 31, 2019. Timescale
on the y-axis represents the timescale, or period. Red values enclosed in black lines indicate signicant spatial
synchrony. C Precipitation anomaly map for August to October 2018 adapted from NOAA (https:// www. ncei.
noaa. gov/ access/ monit oring/ us- maps). Units are in inches representing the departure from average precipitation
over the 1901 to 2000 time period. Areas in green represent greater than average precipitation and areas in
brown represent less than average precipitation.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
6
Vol:.(1234567890)
Scientic Reports | (2024) 14:12479 | https://doi.org/10.1038/s41598-024-62384-6
www.nature.com/scientificreports/
on annual timescales was statistically attributable to synchrony in mean minimum temperatures on the same
timescale range. Temperatures are known to produce synchronizing eects on insects through multiple mecha-
nisms, including a cessation of activity during winter diapause, temperature dependent development where
minimum temperature thresholds must be met for development to occur at dierent life stages, and temperature
dependent mortality at dierent life stages35. Each of these factors or combinations of these factors could have
synchronizing eects on Cx. tarsalis activity and abundances, particularly across a large geographic area that
includes temperate environments where populations undergo winter diapause36,37. In overwintering mosquitoes,
termination of diapause begins with extended photoperiod and warming temperatures, which is then followed by
a post-diapause development period that can result in synchronous spring activity once a minimum temperature
threshold is reached38. In addition, minimum temperatures could produce synchronous abundances through
mortality if minimum temperatures are too cold.
We found that the strength of annual Cx. tarsalis synchrony showed a marginally signicant decrease across
the study period, but we did not observe a signicant decrease in the strength of annual synchrony of minimum
temperature across the study period. According to the wavelet Moran theorem of13, two factors contribute to the
strength of synchrony induced in a population variable by a synchronous environmental variable: the strength of
synchrony of the environmental variable; and the strength/consistency of the relationship between the population
variable and the environmental variable. us, annual timescale Cx. tarsalis synchrony could have been caused
by temperature synchrony, and could have declined over our study period even while temperature synchrony
held steady, if the inuence of temperature on mosquito annual dynamics became less pronounced over the
study period. is could have happened, for instance, if an overall warming trend in minimum temperatures
interacted with threshold dependencies in mosquito life history processes so that thresholds were less frequently
a limiting factor in the growth of some populations38,39. In general, warmer diapause temperature has been found
to desynchronize eclosion in the green-veined white Pieris napi Linnaeus (Lepidoptera:Pieridae) butteries40.
Other factors could also contribute to a decrease in annual synchrony in Cx. tarsalis even while temperature
synchrony remains strong: increased local stochasticity in population dynamics, altered biotic interactions, or
increasing variation in asynchronous regional environmental conditions that drive local Cx. tarsalis dynamics
could all result in reduced annual spatial synchrony across the study period.
An outstanding question that can only be addressed through continued long-term mosquito collections is
whether the marginally signicant result of a decrease in annual spatial synchrony in Cx. tarsalis observed here
is part of a longer-term, consistent trend, or instead is part of a periodic phenomenon, so that the decline in
synchrony will be reversed in due course12 provide a summary of insect populations where longer-term shis
in the strength of spatial synchrony have been observed, and highlight the need for longer-term monitoring
to understand synchronous population dynamics and the eects of global climate change on insects. Reuman
etal. (In Review) point out that, even for systems where time series length is long (those authors consider some
time series which are almost 2000years long), the evidence suggests that yet longer time series can reveal even
longer-timescale synchrony. For mosquito vector species, understanding these patterns is particularly relevant
because the strength of spatial synchrony across multiple timescales may be informative for understanding and
predicting for the timing and abundance of mosquitoes, which relates through a set of other complex processes
to transmission risk.
In addition to signicant annual synchrony, we found evidence of signicant synchrony during a single,
short-term event that occurred between August and September 2018. ese results demonstrate that spatial
synchrony in Cx. tarsalis is not conned to annual cycles alone and emphasizes the need to consider how short
durations or individual events can lead to one-o, spatially synchronous elevations in abundances. e timing
of this event is relevant because it occurred during the peak WNV transmission season and during a year with
broadscale spillover41. Here we show that precipitation between August and October 2018 was unusually high,
with record breaking values occurring across several sites in and around our study area (Fig.5C) (https:// www.
ncei. noaa. gov/ access/ monit oring/ us- maps).
Although synchrony in precipitation was not a signicant predictor of synchrony of Cx. tarsalis during this
event, this result was not unexpected given that spatial wavelet analyses focus on periodicity, and single events
that are short in duration may not trigger signicant wavelet coherence in wavelet linear models. However,
observations of strong spatial synchrony in Cx. tarsalis and knowledge of anomalous precipitation during the
same time period warrants further investigation. Understanding not only interannual patterns but eects of
the timing and distribution of intra-annual, widespread, anomalous weather events on mosquito population
synchrony may reveal environmental drivers that are precursors to potential elevated vector borne disease
transmission hazard. Investigation of such isolated events should probably proceed via statistical methods other
than the wavelet approaches we have used here.
We close by noting some limitations of this work and next steps. First, while the geographic coverage of
the NEON sites spanned a broad area, our study was limited to 90 mosquito traps across 9 locations owing to
sampling coverage and distribution of Cx. tarsalis mosquitoes. Second, we lack the longer time series needed to
determine synchrony on timescales longer than about 16 to 18months. ird, short time series limit statistical
power to detect environmental drivers of synchrony, especially on longer timescales, which is demonstrated
in our nding of a marginally signicant decrease in the strength of spatial synchrony in Cx. tarsalis abun-
dances over our study period. To establish signicance of a relationship between mosquitos and temperature,
the wavelet methods we used looked for consistency, across both space and time, in phase dierences between
annual-timescale uctuations in Cx. tarsalis and temperature. But, on annual timescales, only a few oscillations
of both variables occur during the duration of our data, limiting the potential of our methods to detect con-
sistent phase relationships. Longer datasets would mitigate this problem. In addition, although we conducted
a broadscale analysis of Cx. tarsalis mosquitoes, the distribution of the NEON sites encompasses a relatively
small portion of their geographic range, and our ndings are specic to the study region.Although, evidence
Content courtesy of Springer Nature, terms of use apply. Rights reserved
7
Vol.:(0123456789)
Scientic Reports | (2024) 14:12479 | https://doi.org/10.1038/s41598-024-62384-6
www.nature.com/scientificreports/
indicates mosquito-mediated dispersal of West Nile virus in the western United States42, we consider it unlikely
that dispersal is a driver of synchrony on the spatial scales we examine because of the large geographic distance
between NEON sites in our analyses.
is work sets the stage for integrating longer time series over more sites, including in regions outside the
United States where Cx. tarsalis are abundant and to other WNV mosquito vectors. Longer-term time series
of mosquito collections from monitoring programs including NEON or from mosquito trap collections con-
ducted by mosquito control programs may provide the opportunity to investigate more robustly environmental
correlations with unusual events and to identify common precursors, such as periodic ENSO oscillations, to
synchronous population uctuations.
Despite limitations, this work provides a rst step toward understanding broad scale population synchrony
in a key WNV mosquito vector. Our results highlight the need for additional investigations into eects of
synchronous population dynamics on disease system dynamics during normal and unusual environmental
conditions, over multiple timescales and dierent geographic areas. Future investigations will benet from a
systems approach that includes not only environmental correlates but other components of the WNV disease
system, including avian host populations, migration phenology, and detected spillover events. Understanding
the timing and distribution of inter- and intra-seasonal synchronous mosquito vector population dynamics
across geographic scales may provide needed insight and a piece of the puzzle toward overcoming outstanding
challenges in predicting the magnitude and extent of WNV transmission. Overcoming these challenges can
ultimately help better inform prevention and control agencies.
Materials and methods
National Ecological Observatory Network (NEON) routine mosquito collections (Product: DP1.10043.001)
between April 1, 2016 and December 31, 2019 were used in this analysis43. Mosquitoes at NEON sites are col-
lected using CO2 baited Centers for Disease Control (CDC) light traps, and within each NEON site, mosquitoes
are typically collected at ten trap locations, referred to as plots. Traps are set at a minimum distance of 310m, and
here, the maximum distance between 2 traps occurred at one NEON site (SRER) with a distance of ~ 14.36km.
Terrestrial core sites collect mosquitoes every two weeks during the eld season and then sample three trap
locations weekly during the o season27,28. Following three consecutive trapping events with zero mosquitoes
collected, terrestrial core sites change their sampling scheme to the o season protocol where traps are set weekly
at three plots. is increase in temporal sampling at a reduced number of sites is designed to capture the return
of adult mosquito ight activity following winter conditions. Terrestrial gradient sites collect mosquitoes every
four weeks during the eld season, cease collections when the terrestrial core site within their domain changes to
the o season protocol, and then resumes collections when mosquito activity begins at the terrestrial core site27,28.
Here, we conducted our analyses at a monthly temporal resolution and used the following steps to assemble time
series of Cx. tarsalsis abundances for analyses.
First, we reduced site locations to those that had collected at least one Cx. tarsalis mosquito between 2016
and December 2019 for a total of 26 candidate sites. Next, we performed QA/QC on individual sampling records
to ensure that trap issues did not compromise sampling. We considered a trap collection compromised if any
event occurred that could aect the mosquito counts collected or recorded. ese events included interference
with CO2 sublimation because of blockage, traps tipped over and on the ground, holes in catch cups, ants in
catch cups, samples frozen and irretrievable from the sides of catch cups, damaged samples, or lost and discarded
samples. In each of these cases, the number of mosquitoes or the magnitude of damage to the sample could not
be assessed, introducing an unknown level of uncertainty to abundances. Individual counts recorded in NEON
data reect the number of mosquitoes per species identied out of a subsample of the total trap collection. In
order to estimate the Cx. tarsalis abundance for each record, the individual count is divided by the proportion
of the total trap collection identied. We calculated the estimated count and then created a monthly template
beginning April 1, 2016 and ending December 31, 2019. If a trap collection did not include Cx. tarsalis, we entered
a zero for the count, and if the trap was not set during the month, we entered NA.
Next, we estimated trapping eort and calculated Cx. tarsalis counts per trap hour per month across the 10
locations within each of the 26 NEON candidate sites for each month of the study period, resulting in one value
per site per month. Following this step, we checked sample coverage across the study period for each NEON site
and reduced the analysis to nine sites with consistent sampling during active mosquito time periods. ese sites
consisted of six terrestrial core sites and three terrestrial gradient stations, and they are predominantly located
in the interior West and Great Plains region (Fig.1). Because the spatial wavelet analysis soware we will apply
(see below) requires continuous time series, we replaced NA values with zeros during winter and early spring
months for temperate sites when collections did not occur, using the assumption supported in the literature36,37,44,
that Cx. tarsalis mosquitoes were not active during winter time periods. We note that cessation of sampling dur-
ing winter must be preceded with 0 counts in abundance at the terrestrial core site within the NEON domain,
providing a strong basis for making the assumption that mosquitoes are not active. All sites with mean number
of mosquitoes per trap hour per month are available in supplementary materials, including NA values when
sampling did not occur (see Data Availability).
Aer assembling and cleaning site-level data, we then calculated the centroid of the ten mosquito trap loca-
tions within each NEON site for each of the nine sites to obtain a single geographic point reference for environ-
mental data preparation. GridMET daily cumulative precipitation, minimum and maximum temperature, and
minimum and maximum mean vapor pressure decit index (a measure of humidity) at a 4km spatial resolu-
tion were downloaded for each NEON site between April 1, 2016 and December 31, 2019 using the ‘climateR’
package45,46. Daily data was binned by month, and we calculated total cumulative precipitation for each month,
Content courtesy of Springer Nature, terms of use apply. Rights reserved
8
Vol:.(1234567890)
Scientic Reports | (2024) 14:12479 | https://doi.org/10.1038/s41598-024-62384-6
www.nature.com/scientificreports/
and average values for all other variables (e.g., minimum and maximum temperature and relative humidity)
using functions available in the ‘dplyr’ package in R47.
Once monthly abundance and climate data were assembled, we used the ‘wsyn’ package in R for spatial
synchrony analyses48. First, we used the ‘cleandat’ function with clev = 3 to individually de-mean, detrend, and
standardize the variance of the monthly time series for Cx. tarsalis and each environmental variable. Next, we
calculated wavelet phasor mean elds (wpmf) to quantify whether signicant spatial synchrony occurred at one
or more timescales for monthly estimated abundances of Cx. tarsalis and each environmental variable. Given
a collection of time series measured at the same times (e.g., our Cx. tarsalis time series), the wpmf technique
provides a plot which displays the strength of phase synchrony in the input time series as a function of time
and timescale, with signicance contours. Intense colors on the plot indicate strong synchrony at the given time
and timescale. is method and a suite of closely related and now well developed methods have been applied
numerous times to study synchrony of ecological time series15,49–56, and the methods are implemented, open
source, in the wsyn package on CRAN48. e wsyn package includes a “vignette” which gives a straightforward,
operational introduction to the methods implemented therein.
Next, we observed the wpmf plot for Cx. tarsalis to determine whether spatial synchrony was signicant
across one or more timescales. We then t univariate wavelet linear models using the ‘wlm’ and ‘wlmtest’ func-
tions in wsyn to quantify whether spatial synchrony in the climate variables were signicant predictors of spatial
synchrony in Cx. tarsalis at the same timescales. Wavelet linear models were originally developed by13,54, and
have now been applied several times in ecology15,49–53 to identify environmental causes of synchrony; they are
especially useful when causes of synchrony may dier by timescale. ese tools can identify not only which
environmental drivers likely help cause synchrony on a given timescale band, they can also indicate the fractions
of synchrony explained by each driver and by interactions between drivers. Wavelet linear model methods make
statements of statistical signicance based on resampling/randomization procedures; the number of randomiza-
tions in the ‘wlmtest’ function was set to 10,000. e ‘bandtest’ and ‘plotrank’ functions in wsyn were used to
quantify signicance. If a climate variable was identied as a signicant driver of synchrony in Cx. tarsalis, the
percent synchrony explained by the variable was then obtained from the wlm model, with associated cross-terms
and residuals.
Aer observing the wpmf output for Cx. tarsalis, we tested whether a signicant decrease in the strength of
spatial synchrony occurred across the study period at the annual timescale by generating 10,000 “synchrony pre-
serving surrogates” using the ‘surrog’ function in wsyn, with surrtype = ‘aa’54. Synchrony preserving surrogates
are so-called surrogate datasets, i.e., articial, bootstrapped datasets of the same structure as the original data
(same number, length, and sampling frequency of time series), but which have been randomized in such a way
that synchrony between the time series is maintained in its strength and timescale structure, but any directional
changes through time in synchrony are eliminated. us, comparing patterns of change through time in the
synchrony of real data against the same patterns computed for the surrogate datasets provides a test of whether
synchrony has directionally changed, to a signicant extent, against an appropriate null hypothesis.
We calculated the synchrony values for our observed Cx.tarsalis data at the 12month and averaged 10 to
14month timescales, regressed these quantities against year to obtain two slopes, and calculated the same
statistics for each of the 10,000 synchrony-preserving surrogate datasets. We then calculated the proportion of
the 10,000 surrogate slopes which were less than the observed slopes to obtain a p-value for the test of the null
hypothesis that observed decreases in annual-band Cx.tarsalis were no more than could have been expected by
chance. Signicance was measured with an ɑ < 0.05 and we report marginal signicance with an ɑ < 0.10. We
also tested whether a signicant decrease in the strength of annual spatial synchrony was present in the climate
variables, using the same methods.
Data availability
e data used in this study, including the data le prior to lling NA values, is available through GitHub (https://
github. com/ Campb ell- Lab- FMEL/ Culex- tarsa lis- synch rony).
Code availability
All code used to perform analyses and to create gures is available through GitHub (https:// github. com/ Campb
ell- Lab- FMEL/ Culex- tarsa lis- synch rony).
Received: 4 March 2024; Accepted: 16 May 2024
References
1. WHO. World Health Organization Fact Sheet. Vector-borne Diseases. https:// www. who. in t/ n ews- room/ fact- sheets/ detail/ vector-
borne- disea ses (2020).
2. Pavlovsky, E. N. e natural nidus of a disease as a pathobiocenose. In e natural nidality of transmissible disease (University of
Illinois, 1966).
3. Reisen, W. K. Landscape epidemiology of vector-borne diseases. Annu. Rev. Entomol. 55, 461–483 (2010).
4. Mullen, G. R. & Durden, L. A. Medical and veterinary entomology (Academic Press, 2009).
5. Davis, J. K. et al. Integrating environmental monitoring and mosquito surveillance to predict vector-borne disease: Prospective
forecasts of a West Nile virus outbreak. PLoS Curr. https:// doi. org/ 10. 1371/ curre nts. outbr eaks. 90e80 717c4 e67e1 a830f 17fee aaf85
de (2017).
6. Althouse, B. M. et al. Synchrony of sylvatic dengue isolations: A multi-host, multi-vector SIR model of dengue virus transmission
in Senegal. PLoS Negl. Trop. Dis. 6, e1928 (2012).
7. McDevitt-Galles, T., Moss, W. E., Calhoun, D. M. & Johnson, P. T. J. Phenological synchrony shapes pathology in host-parasite
systems. Proc. Biol. Sci. 287, 20192597 (2020).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
9
Vol.:(0123456789)
Scientic Reports | (2024) 14:12479 | https://doi.org/10.1038/s41598-024-62384-6
www.nature.com/scientificreports/
8. McMillan, J. R., Chaves, L. F. & Armstrong, P. M. Ecological predictors of mosquito population and arbovirus transmission syn-
chrony estimates. J. Med. Entomol. 60, 564–574 (2023).
9. Liebhold, A., Koenig, W. D. & Bjørnstad, O. N. Spatial synchrony in population dynamics. Annu. Rev. Ecol. Evol. Syst. 35, 467–490
(2004).
10. Liebhold, A. M. & Haynes, K. J. Spatial synchrony of insect outbreaks. Insect Outbreaks https:// doi. org/ 10. 1002/ 97811 18295 205#
page= 135 (2004).
11. Haynes, K. J., Bjørnstad, O. N., Allstadt, A. J. & Liebhold, A. M. Geographical variation in the spatial synchrony of a forest-
defoliating insect: Isolation of environmental and spatial drivers. Proc. Biol. Sci. 280, 20122373 (2013).
12. Haynes, K. J. & Walter, J. A. Advances in understanding the drivers of population spatial synchrony. Curr. Opin. Insect Sci. 53,
100959 (2022).
13. Sheppard, L. W., Bell, J. R., Harrington, R. & Reuman, D. C. Changes in large-scale climate alter spatial synchrony of aphid pests.
Nat. Clim. Chang. 6, 610–613 (2016).
14. Ward, S. F. & Aukema, B. H. Climatic synchrony and increased outbreaks in allopatric populations of an invasive defoliator. Biol.
Invasions 21, 685–691 (2019).
15 . Walter, J. A. et al. Weather and regional crop composition variation drive spatial synchrony of lepidopteran agricultural pests. Ecol.
Entomol. 45, 573–582 (2020).
1 6. Chaves, L. F. Mosquito species (Diptera: Culicidae) persistence and synchrony across an urban altitudina l gradient. J. Med. Entomol.
54, 329–339 (2017).
17. Nik Abdull Halim, N. M. H., Che Dom, N., Dapari, R., Salim, H. & Precha, N. A systematic review and meta-analysis of the eects
of temperature on the development and survival of the Aedes mosquito. Front. Public Health https:// doi. org/ 10. 3389/ fpubh. 2022.
10740 28 (2022).
18. Valdez, L. D., Sibona, G. J., Diaz, L. A., Contigiani, M. S. & Condat, C. A. Eects of rainfall on Culex mosquito population dynam-
ics. J. eor. Biol. 421, 28–38 (2017).
19. Chaves, L. F., Satake, A., Hashizume, M. & Minakawa, N. Indian Ocean dipole and rainfall drive a Moran eect in East Africa
malaria transmission. J. Infect. Dis. 205, 1885–1891 (2012).
20. Chaves, L. F., Friberg, M. D. & Moji, K. Synchrony of globally invasive Aedes spp. immature mosquitoes along an urban altitudinal
gradient in their native range. Sci. Total Environ. 734, 139365 (2020).
21. Chaves, L. F. et al. Modeling the association between Aedes aegypti ovitrap egg counts, multi-scale remotely sensed environmental
data and arboviral cases at Puntarenas, Costa Rica (2017–2018). Curr. Res. Parasitol. Vector Borne Dis. 1, 100014 (2021).
22. Anyamba, A. et al. Prediction of a Ri Valley fever outbreak. Proc. Natl. Acad. Sci. USA 106, 955–959 (2009).
23. Anyamba, A. et al. Meshing biosurveillance and climate data to inform chikungunya disease surveillance. Int. J. Infect. Dis. 116,
S97 (2022).
24. Cazelles, B., Chavez, M., McMichael, A. J. & Hales, S. Nonstationary inuence of El Niño on the synchronous dengue epidemics
in ailand. PLoS Med. 2, e106 (2005).
25. García-Carreras, B. et al. Periodic synchronisation of dengue epidemics in ailand over the last 5 decades driven by temperature
and immunity. PLoS Biol. 20, e3001160 (2022).
26. Anyamba, A. et al. Global disease outbreaks associated with the 2015–2016 El Niño event. Sci. Rep. 9, 1930 (2019).
2 7. Springer, Y. P. et al. Tick-, mosquito-, and rodent-borne parasite sampling designs for the National Ecological Observatory Network
[Special Feature: NEON Design]. (2016).
28. Hoekman, D. et al. Design for mosquito abundance, diversity, and phenology sampling within the National Ecological Observatory
Network. Ecosphere 7, e01320 (2016).
29. Gorris, M. E. et al. Updated distribution maps of predominant Culex mosquitoes across the Americas. Parasit. Vectors 14, 547
(2021).
30. Dunphy, B. M. et al. Long-term surveillance denes spatial and temporal patterns implicating Culex tarsalis as the primary vector
of West Nile virus. Sci. Rep. 9, 6637 (2019).
3 1. Reisen, W. K. Eect of temperature on Culex tarsalis (Diptera: Culicidae) from the Coachella and San Joaquin Valleys of California.
J. Med. Entomol. 32, 636–645 (1995).
32. Brown, J. J., Pascual, M., Wimberly, M. C., Johnson, L. R. & Murdock, C. C. Humidity—e overlooked variable in the thermal
biology of mosquito-borne disease. Ecol. Lett. 26, 1029–1049 (2023).
33. Wolda, H. Insect seasonality: Why?. Annu. Rev. Ecol. Syst. 19, 1–18 (1988).
34. Tauber, M. J. & Tauber, C. A. Insect seasonality: Diapause maintenance, termination, and postdiapause development. Annu. Rev.
Entomol. 21, 81–107 (1976).
35. Bjørnstad, O. N., Nelson, W. A. & Tobin, P. C. Developmental synchrony in multivoltine insects: Generation separation versus
smearing. Popul. Ecol. 58, 479–491 (2016).
36. Buth, J. L., Brust, R. A. & Ellis, R. A. Development time, oviposition activity and onset of diapause in Culex tarsalis, Culex restuans
and Culiseta inornata in southern Manitoba. J. Am. Mosq. Control Assoc. 6, 55–63 (1990).
37. Nelms, B. M., Macedo, P. A., Kothera, L., Savage, H. M. & Reisen, W. K. Overwintering biology of Culex (Diptera: Culicidae)
mosquitoes in the Sacramento Valley of California. J. Med. Entomol. 50, 773–790 (2013).
38. Denlinger, D. L. & Armbruster, P. A. Mosquito diapause. Annu. Rev. Entomol. 59, 73–93 (2014).
39. Forrest, J. R. Complex responses of insect phenology to climate change. Curr. Opin. Insect Sci. 17, 49–54 (2016).
40. Lehmann, P., Van Der Bijl, W., Nylin, S., Wheat, C. W. & Gotthard, K. Timing of diapause termination in relation to variation in
winter climate. Physiol. Entomol. 42, 232–238 (2017).
41. McDonald, E. et al. Surveillance for West Nile virus disease - United States, 2009–2018. Am. J. Transpl. 21, 1959–1974 (2021).
42. Venkatesan, M. & Rasgon, J. L. Population genetic data suggest a role for mosquito-mediated dispersal of West Nile virus across
the western United States. Mol. Ecol. 19, 1573–1584 (2010).
43. NEON (National Ecological Observatory Network). NEON. Mosquitoes sampled from CO2 traps (DP1.10043.001) htt ps: // doi. org/
10. 48443/ 3cyq- 6v47. (2024).
44. Mitchell, C. J. Diapause termination, gonoactivity, and dierentiation of host-seeking behavior from blood-feeding behavior in
hibernating Culex tarsalis (Diptera: Culicidae). J. Med. Entomol. 18, 386–394 (1981).
45. Johnson, M. climateR: climateR. Preprint at https:// github. com/ mikej ohnso n51/ clima teR (2023).
46. Abatzoglou, J. T. Development of gridded surface meteorological data for ecological applications and modelling. Int. J. Climatol.
33, 121–131 (2013).
47. Wickham, H., François, R., Henry, L. & Müller, K. Vaughan D. dplyr: A grammar of data manipulation. R package version 1.1.3,
<https:// CRAN.R- proje ct. org/ packa ge= dplyr> (2023).
48. Reuman, D. C., Anderson, T. L., Walter, J. A., Zhao, L. & Sheppard, L. W. wsyn: Wavelet approaches to studies of synchrony in
ecology and other elds. R package R package version 1.0.4, <https:// CRAN.R- proje ct. org/ packa ge= wsyn> (2021).
49. Walter, J. A. et al. Spatial synchrony cascades across ecosystem boundaries and up food webs via resource subsidies. Proc. Natl.
Acad. Sci. USA 121, e2310052120 (2024).
50. Castorani, M. C. N. et al. Disturbance and nutrients synchronise kelp forests across scales through interacting Moran eects. Ecol.
Lett. 25, 1854–1868 (2022).
51. Walter, J. A. et al. Micro-scale geography of synchrony in a serpentine plant community. J. Ecol. 109, 750–762 (2021).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
10
Vol:.(1234567890)
Scientic Reports | (2024) 14:12479 | https://doi.org/10.1038/s41598-024-62384-6
www.nature.com/scientificreports/
52. Anderson, T. L., Sheppard, L. W., Walter, J. A., Rolley, R. E. & Reuman, D. C. Synchronous eects produce cycles in deer popula-
tions and deer-vehicle collisions. Ecol. Lett. 24, 337–347 (2021).
53. Sheppard, L. W., Mechtley, B., Walter, J. A. & Reuman, D. C. Self-organizing cicada choruses respond to the local sound and light
environment. Ecol. Evol. 10, 4471–4482 (2020).
54. Sheppard, L. W., Defriez, E. J., Reid, P. C. & Reuman, D. C. Synchrony is more than its top-down and climatic parts: interacting
Moran eects on phytoplankton in British seas. PLoS Comput. Biol. 15, e1006744 (2019).
55. Walter, J. A. et al. e geography of spatial synchrony. Ecol. Lett. 20, 801–814 (2017).
56. Anderson, T. L. et al. e dependence of synchrony on timescale and geography in freshwater plankton. Limnol. Oceanogr. 64,
483–502 (2019).
Acknowledgements
We would like to acknowledge the National Ecological Observatory Network and personnel for the collection,
identication, and hosting of the mosquito data used in this study and Sara Paull for clarication on the NEON
mosquito collections data and sampling scheme. e National Ecological Observatory Network is a program
sponsored by the National Science Foundation and operated under cooperative agreement by Battelle. is
material is based in part upon work supported by the National Science Foundation through the NEON Program.
LPC is supported by the USDA National Institute of Food and Agriculture, Hatch project #1021482; AMB is
supported through the University of Florida Graduate Student Fellowship Award; and this work was partially
supported through the University of Florida Biodiversity Institute Seed Fund. DCR was partly supported by NSF
Bio-Oce grant #2023474 and by the McDonnell and Humboldt foundations. Guralnick acknowledges support
from NSF CIBR grant #2223512.
Author contributions
LPC and DR designed the study with help from RPG. AMB, LPC, and YT performed the analyses. LPC led the
manuscript writing with help from RPG, DR, AMB, and YT. AMB coordinated the deposition of data and code
on GitHub.
Competing interests
e authors declare no competing interests.
Additional information
Supplementary Information e online version contains supplementary material available at https:// doi. org/
10. 1038/ s41598- 024- 62384-6.
Correspondence and requests for materials should be addressed to L.P.C.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional aliations.
Open Access is article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. e images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.
© e Author(s) 2024
Content courtesy of Springer Nature, terms of use apply. Rights reserved
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com
Available via license: CC BY 4.0
Content may be subject to copyright.
