Current and future distribution of Ixodes scapularis ticks in Québec: Field validation of a predictive model.

The incidence of Lyme disease is increasing in Québec and is closely linked to the distribution of Ixodes scapularis ticks. A time-to-establishment model developed in 2012 by Leighton and colleagues predicted the year of tick population establishment for each municipality in eastern Canada. To validate if this model correctly predicted tick distribution in Québec, predicted tick establishment was compared to field data from active tick surveillance (2010-2018) using two criteria: i) the detection of at least one tick and ii) the detection of the three questing stages of the tick. The speed of tick establishment and the increase in the exposed human population by 2100 were predicted with the time-to-establishment model. Field observations were consistent with model predictions. Ticks were detected on average 3 years after the predicted year. The probability of tick detection is significantly higher after the predicted year than before (61% vs 27% of collections). The trend was similar for the detection of three tick stages (16% vs 9% of collections). The average speed of tick range expansion was estimated by the model to be 18 km/year in Québec, with 90% of the human population exposed by 2027. The validation of the time-to-establishment model using field data confirmed that it could be used to project I. scapularis range expansion in Québec, and consequently the increase in Lyme disease risk over the coming decades. This will help public health authorities anticipate and adapt preventive measures, especially in areas not yet affected by Lyme disease.

Introduction

Lyme disease, caused by Borrelia burgdorferi sensu lato complex (hereafter shortened to B. burgdorferi), is the most common vector-borne disease in North America [1]. In Canada, the number of locally acquired human cases per year increased from 144 in 2009 to 2636 cases in 2019 [2]. In Eastern Canada, the incidence of Lyme disease is progressing because of the northward range expansion of the blacklegged tick (Ixodes scapularis), the main vector of B. burgdorferi. This expansion is facilitated by the presence of favorable climatic conditions (temperature and humidity affect tick survival and activity), habitat types, (deciduous and mixed forests are considered suitable habitats for I. scapularis) and host communities (white-footed mice–Peromyscus leucopus and white-tailed deer–Odocoileus virginianus which are the main hosts of I. scapularis). Climate change and associated changes in environmental conditions and human activities have contributed to the increasing risk of Lyme disease in Canada [3,4]. In Québec, a province in eastern Canada, the integrated Lyme disease surveillance program provides standardized baseline data on both ticks and human disease to track the spread of Lyme disease [5,6]. Lyme disease has been a notifiable human disease in Québec since 2003. In Québec, the number of locally acquired human cases of Lyme disease has increased from 2 cases in 2008 to 381 by 2019 [7]. Passive tick surveillance programs have existed in Québec since 1990 and consist of submission of ticks found on people or pets, mainly by human and animal health professionals. Active surveillance has been conducted annually in Québec since 2007 and consists of collecting ticks in the environment during spring and summer by the drag sampling method [5,6]. Active surveillance data show that blacklegged ticks were initially only detected south of the St. Lawrence River but are now frequently detected north of the river [8-10].

The distribution of established I. scapularis tick populations is a well-known proxy of human Lyme disease risk [11-13]. The tick life cycle typically spans two years with larvae active in summer, nymphs next spring, and adults in autumn [11]. Finding the three stages at one collection site during the same year suggests that they are not of the same generation and provides strong evidence of establishment given that at least one reproduction cycle has occurred locally [12]. Another study suggested that even a single tick collected in the environment during three person-hours of drag sampling provides a strong indicator of local tick establishment [13]. In Québec, public health authorities use the criterion of the three stages in one year to determine if a tick population has established [5]. Currently, ticks can be found throughout Québec due to “adventitious” ticks dispersed by migrating birds or terrestrial animals but established tick populations have only been identified in the southern parts of Québec [14,15]. As Lyme disease risk continues to spread northwards, anticipating and rapidly detecting the location of newly established tick populations would help public health authorities implement local prevention measures.

Active tick surveillance provides useful information about the distribution of ticks, but will not detect and predict all established tick populations in space and time. A predictive risk map of future tick occurrence according to various climate change scenarios was produced based on a temperature threshold suitable for tick establishment [16,17]. Simon et al. [18] combined the same model of tick distributions with projected range expansion of white-footed mice to produce a future risk index for B. burgdorferi in Québec. These models give interesting information about potential tick distribution in the environment but are too complex to be easily used by public health authorities. Leighton et al. [19] took a different approach, analyzing a 20-year time series of passive surveillance data, along with environmental drivers including temperature, elevation, annual rainfall, and local and long-distance dispersal of ticks, to predict the year of future tick establishment for municipalities across eastern Canada. This approach resulted in a predictive map of tick establishment per year with an estimated speed of spread of 46 km/year. This “time-to-establishment” map provides a simple and useful tool for public health authorities to anticipate the change in risk by municipality and by year. Field data from Ontario found good agreement between model predictions and observed presence of I. scapularis 10 year later [20], but model projections have yet been evaluated with field data from Québec.

The objective of our study was to assess the concordance between field observations from active tick surveillance (2010–2018) and the timing of tick establishment predicted by the model of Leighton et al. [19]. We investigated how well the model was able to predict 1) the detection of the all three life stages of ticks (strict definition of established tick population), and 2) the detection of at least one tick (less restrictive definition). We then used the model projections to 2100 to investigate the pattern and the speed of tick spread and the potential increase of human exposure to ticks in the coming decades in Québec.

Materials and methods

Data

Study area

The province of Québec is in the eastern part of Canada, located between the provinces of Ontario to the west and New Brunswick to the east and bordering the United States to the south. It is the largest and the second-most populous province in Canada, with 8,164,361 inhabitants distributed over 1,356,625 km2 [21]. The province is divided into 1,226 municipalities, or census subdivisions (CSDs), with a median area of 100 km2 and an average population of 6,639 inhabitants (ranging from 5 to 1,650,000; median = 1,142). For each municipality included in this study, we estimated the population size and the area (km2) using data from the 2016 census [21].

Model predictions

The predicted year of tick establishment for I. scapularis for each municipality is the output of a parametric survival regression model developed for eastern Canada passive tick surveillance data (1990–2008) and environmental predictors such as temperature (degree days > 0°C), elevation, annual rainfall, and local and long-distance dispersal of ticks [19]. For each municipality, there was a mean predicted year, with a lower predicted year (reflecting the fastest progression of ticks) and an upper predicted year (reflecting the slowest progression of ticks), based on the maximum and minimum temperatures observed in each municipality between 1991 and 2008 to account for potential climate change [19].

Active tick surveillance data

Tick distribution data come from annual Lyme disease surveillance activities in Québec, carried out jointly by the Institut national de santé publique du Québec (INSPQ), the Ministère de la Santé et des Services Sociaux (MSSS), the Public Health Agency of Canada (PHAC), and the University of Montréal [5,6]. A standardized procedure was used from 2010 to 2018, with drag sampling carried out between June and September in different woodlands that are open to the public. Ticks are generally collected once a year per site but not every year. The sampling period targets nymphs and larvae, but adults can also be collected early in the season. Host-seeking (or questing) ticks were collected by dragging a 1 m2 white flannel cloth over the forest floor, in the morning or early afternoon [22]. Drag sampling did not occur on rainy days or when the forest floor was wet. Dragging method was used following four lines of 500m on either side of a trail, with two lines within 5m of the trail and two lines at 25m of the trail, covering a total of 2000 m2 per site [6]. Every 25 m, the flag was inspected, and collected ticks were sent to the Laboratoire de Santé Publique du Québec (LSPQ) for tick species and life stage identification, using taxonomic keys for larvae, nymphs and adults [23,24].

Model validation

Established tick population indicators

The model developed by Leighton et al. [19] was designed to predict the year of tick population establishment, based on a passive surveillance indicator of the detection of at least two adults or one immature tick by active surveillance for two consecutive years. This definition of tick population establishment is not the same as those currently used by Québec’s public health authorities. We therefore compared predictions with two active surveillance measures: 1) detection of the three tick stages in the same year, which is the Québec surveillance definition of an established tick population [5] and 2) detection of at least one tick, which is a broader definition of potential tick population establishment [25]. We examined the lower, mean and upper predicted year of establishment to test the sensitivity of the model to climate variation.

Statistical analysis

We investigated the concordance between field observations and model predictions in two steps: firstly, year by year, for each sampling event; secondly, over the study period, aggregating surveillance data from 2020 to 2018 for each municipality

: First, we investigated the concordance between predicted establishment and tick surveillance data collected during each sampling event (site visit) over the study period, year by year. Do we find tick populations when and where the model predicted? For each sampling event, in a given municipality and year, (n = 444 visits), we determined: i) whether all three tick stages were detected (yes/no), ii) whether at least one stage of tick was detected (yes/no) and iii) whether the model predicted the presence of an established tick population (yes/no) in the municipality on the year of sampling.

: Secondly, we investigated the concordance between predicted establishment and the presence of ticks in each municipality over the study period, aggregating all tick data from n 2010 to2018. Have we already found tick populations between 2010 and 2018 in municipalities where the model predicted tick population establishment prior to 2018? Because the drag sampling is known to have a low sensitivity to detect tick populations even when these are present [25,26], we aggregated surveillance data by municipality over the study period, cumulating active surveillance data from 2010 to 2018. For each municipality where active surveillance was carried out at least once between 2010–2018 (n = 231 municipalities), we determined: i) whether all three tick stages were detected at least once before 2018 (yes/no), ii) whether at least one stage of tick was detected prior to 2018 (yes/no) and iii) whether the model predicted the presence of an established tick population in 2018 (yes/no).

We compared predicted and observed presence/absence of ticks using the Chi-square test and the kappa coefficient of determination. A significant Chi-square test, with p<0.05, suggests that the distribution of sites with presence of ticks is not at random between areas with predicted or unpredicted established tick population according to the model. We used the matched McNemar test for the analysis of all the sampling because of some repeated samples at the same site. The kappa statistic was interpreted as follows: > 0.75 as excellent, 0.40 to 0.75 as fair to good, and < 0.40 as poor agreement. Analyses were carried out using the kappa and mcnemar.test functions in R version 4.0.2 [27]. We also calculated the time lag between the exact predicted year and the first detection of all three tick stages or the detection of at least one tick stage in a municipality by active surveillance.

Projected tick range expansion in Québec: 2019–2100

We used the Leighton et al. [19] model to project annual tick range expansion across all Québec municipalities between 2019–2100. The annual speed of tick range expansion in Québec was estimated as the difference between the square root areas of tick establishment in two consecutive years: speed [km/year] = √ (area year n)—√ (area year n-1). From 2006 to 2100, the annual proportion of the Québec population exposed was calculated as the number of people living in municipalities where the model predicted an established tick population divided by the human population of Quebec, based on the 2016 census data [21]. Finally, in order to identify potential regional differences in tick exposure in the coming years, we produced a map of projected tick range expansion to 2100 using the Free and Open Source QGIS software version 3.4.9 (QGIS Geographic Information System).

Results

Of Québec’s 1,226 municipalities, the model of Leighton et al. [19] was able to predict the year of tick population establishment in 1,180 municipalities since 46 (4%) were excluded from the study because of a lack of environmental data. The predicted spread of tick establishment was centrifugal from the southern border of Québec with Ontario and the United States (Fig 1). For 44% of the municipalities, tick population establishment was predicted to occur before 2018 (540 municipalities according to the mean predicted year; 284 and 651 municipalities according to the upper and lower predicted years).

Fig 1

Tick surveillance in Québec municipalities, 2010–2018, and predicted tick population establishment by 2018.

Active surveillance from 2010 to 2018: municipalities with detection of three stages are indicated by black stars; those with the presence of at least one tick by black triangles and those with no tick detection by white circles. Shaded areas of the map show municipalities with predicted tick population establishment by 2018 according to the lower, mean and upper predicted year (light, medium and dark grey zones, respectively).

Active tick surveillance

Between 2010 and 2018, ticks were sampled in 231 municipalities, for a total of 444 tick sampling events—municipalities were sampled twice on average during the study period (range: 1 to 7 samples per municipality). At least one I. scapularis tick was detected in 218 (49%) site visits, including 52 (12%) with detection of all three stages. In the remaining 226 (51%) site visits, no ticks were detected by the dragging method. At the municipality level, I. scapularis was detected in 122 municipalities (53% with at least one tick), including 35 (15%) with detection of the three stages in one year. No ticks were detected in the remaining 109 (47%) municipalities sampled.

Sampling event concordance

Detection of three tick life stages

The probability of detecting the three tick life stages during a single sampling event by active surveillance was significantly higher in municipalities where the model predicted the existence of an established tick population (i.e., sampling conducted in a municipality on or following the predicted year of tick establishment for that municipality). The three stages were detected in 16% (47/289) of the collections that took place after the mean predicted year, but only in 3% (5/155) of the collections that took place before the mean predicted year (lower year: 3% vs 1%; upper year: 16% vs 9%; p<0.001) (Table 1). The upper predicted year provided the best prediction: 64.2% (285/444) of the sampling events were predicted correctly and 35.8% (159/444) of the sampling events were predicted incorrectly (kappa = 0.09 [IC95% 0.01–0.17], p<0.001). The mean predicted year was the second-best prediction: 44.4% (197/444) of sampling events were predicted correctly versus 55.6% (247/444) predicted incorrectly (kappa = 0.09 [IC95% 0.05–0.13], p<0.001). The lower predicted year provided the least accurate prediction: 26.1% (116/444) of sampling events were predicted correctly versus 73.8% (328/444) predicted incorrectly (kappa = 0.03 [IC95% 0.02–0.05], p<0.001) (Table 3). The first detection of the three stages by active surveillance occurred on average three years after the mean predicted year (mean = 3.32; sd = 2.91; range = [-5; 8]) (Table 2). In only one instance were the three stages detected before the lower predicted year, and this was only one year prior to predicted establishment.

Table 1

Concordance between tick surveillance data and predicted year of establishment.

Observed presence or absence of ticks during active surveillance (no detection, tick presence, three stages) relative to model predictions (before or after the lower, mean and upper predicted year of establishment).

		Number (%) of sampling events (n = 444)				Kappa [IC 95%] A: Tick presence B: Three stages	Chi-square test p-value
Predicted year	Established tick population (Year of tick collection)	Total	No detection	Tick presence	Three stages 1
Lower	Expected (After predicted year2)	378	179 (48%)	199 (52%)	51 (3%)	A: 0.11 [0.05–0.18]	<0.001
	Not expected (Before predicted year)	66	47 (72%)	19 (28%)	1 (1%)	B: 0.03 [0.02–0.05]	<0.001
Mean	Expected (After predicted year2)	289	113 (39%)	176 (61%)	47 (16%)	A: 0.30 [0.22–0.38]	<0.001
	Not expected (Before predicted year)	155	113 (78%)	42 (27%)	5 (3%)	B: 0.09 [0.05–0.13]	<0.001
Upper	Expected (After predicted year2)	160	58 (37%)	102 (63%)	26 (16%)	A: 0.21 [0.12–0.30]	<0.001
	Not expected (Before predicted year)	284	168 (59%)	116 (41%)	25 (9%)	B: 0.09 [0.01–0.17]	<0.001
	Total	444	226	218	52

1 the number of sampling events with three stages are included in tick presence.

on or following the predicted year.

Table 3

Concordance between field data and predicted tick establishment by 2018.

Observed presence or absence of ticks during active surveillance (no detection, tick presence, three stages) in municipalities between 2010–2018 vs predicted tick establishment in 2018 (before or after 2018) according to lower, mean and upper model predicted year.

		Number (%) of municipalities in 2010–2018 (n = 231)				Kappa [IC 95%] A: Tick presence B: Three stages	Chi-square test p-value
Predicted year	Established tick pop (Predicted year)	Total	No detection	Tick presence	Three stages 1
Lower	Expected (Before 2018)	209	89 (43%)	120 (57%)	35 (16%)	A: 0.17 [0.09–0.25]	<0.001
	Not expected (After 20182)	22	20 (91%)	2 (9%)	0 (0%)	B: 0.03 [0.02–0.06]	<0.001
Mean	Expected (Before 2018)	171	56 (32%)	115 (68%)	35 (20%)	A: 0.43 [0.33–0.54]	<0.001
	Not expected (After 20182)	60	53 (88%)	7 (12%)	0 (0%)	B: 0.11 [0.07–0.16]	<0.001
Upper	Expected (Before 2018)	118	28 (24%)	90 (76%)	30 (25%)	A: 0.47 [0.36–0.59]	<0.001
	Not expected (After 20182)	113	81 (72%)	32 (28%)	5 (4%)	B: 0.20 [0.11–0.29]	<0.001
Total		231	109 (47%)	122 (53%)	35 (14%)

1 the number of collections with three stages are included in tick presence.

2 on or following 2018.

Table 2

Time lag between predicted tick establishment and detection of ticks by active surveillance.

Mean, range and standard deviation of the time lag in years between predicted tick establishment and observation of a ticks (three stages or tick presence) in a municipality according to lower, mean and upper predicted year of establishment.

	Time lag (years) between prediction and observation (mean, range, sd)*
	Lower year	Mean year	Upper year
Three stages	5.69 (-1;10) sd = 2.71	3.32 (-5;8) sd = 2.91	-0.19 (-9;5) sd = 3.31
Tick presence	5 (-7;12) Sd = 3.40	2.4 (-12;9) Sd = 3.73	-1.26 (-19;5) Sd = 4.16

*Time lag = observation year—predicted year; range = (minimum; maximum); sd = standard deviation.

Detection of at least one tick

Detection of at least one tick showed similar trends as with the detection of all three stages. Ticks were detected in 61% (176/289) of the sampling events that took place on or following the mean predicted year, and 27% (42/155) of tick collections that took place before the mean predicted year (lower year: 52% vs 28%; upper year: 63% vs 41%; p<0.001) (Table 1). The mean predicted year provided the best prediction in this case: 65.1% of sampling events were predicted correctly (289/444) and 34.9% (155/444) of sampling events were predicted incorrectly (kappa = 0.30 [IC95% 0.22–0.38], p<0.001). The upper predicted year was the second-best prediction: 60.8% (270/444) of sampling events were predicted correctly versus 39.2% (174/444) predicted incorrectly (kappa = 0.21 [IC95% 0.12–0.30], p<0.001). The lower predicted year provided the least accurate prediction: 55.4% (246/444) of sampling events were predicted correctly versus 44.6% (198/444) predicted incorrectly (kappa = 0.11 [IC95% 0.05–0.18], p<0.001) (Table 3).

The first detection of ticks by active surveillance occurred on average two years after the mean predicted year (presence of ticks: 2.4; sd = 3.73; range = [-12;9]) (Table 2).

Concordance over the study period

Detection of three stages

The probability of detecting the three tick stages by active surveillance between 2010 and 2018 was significantly higher if the model predicted tick establishment before 2018 (Table 3). The three stages were detected in 20% (35/171) of municipalities with a mean predicted year before 2018, but 0% (0/60) of municipalities with a mean predicted year after 2018 (16% vs 0% for lower year and 34% vs 4% for upper year; p<0.001). The upper predicted year was the best prediction: 59.7% (138/231) of the municipalities were predicted correctly and 40.2% (93/231) incorrectly (kappa = 0.21 [IC95% 0.11–0.29], p<0.001). The mean predicted year was the second-best prediction: 41.1% (95/231) of the municipalities were predicted correctly and 58.8% (136/231) incorrectly (kappa = 0.11 [IC95% 0.07–0.16], p<0.001). The lower predicted year was the worst prediction: 24.6% (57/231) of the municipalities were predicted correctly and 75.3% (174/231) incorrectly (kappa = 0.03 [IC95% 0.02–0.25], p<0.001) (Table 3).

Ticks were detected in 68% (115/171) of municipalities with mean predicted year of establishment before 2018 and in 12% (7/60) of municipalities with mean predicted year after 2018 (74% vs 29% for lower predicted year and 58% vs 10% for upper predicted year; p<0.001). The upper predicted year was the best prediction: 74.0% (171/231) of the municipalities were predicted correctly and 26.0% (60/231) incorrectly (kappa = 0.47 [IC95% 0.36–0.59], p<0.001). The mean predicted year was the second-best prediction: 72.7% (168/231) of the municipalities were predicted correctly and 27.3% (63/231) incorrectly (kappa = 0.43 [IC95% 0.33–0.54], p<0.001). The lower predicted year was the worst prediction: 60.6% (140/231) of the municipalities were predicted correctly and 39.4% (91/231) incorrectly (kappa = 0.17 [IC95% 0.09–0.25], p<0.001) (Table 3).

Projected tick range expansion in Québec

Speed of tick spread

The average speed of projected tick range expansion between 2020–2100 (based on mean predicted year of establishment) was 18 km/year (Fig 2). The speed was similar for upper and lower predicted years (15 and 23 km/year). For the 2020–2030 projections, tick range expansion speed was 22 km/year using the mean predicted year (14 and 21 km/year for the lower and upper years). Speed of range expansion was relatively constant over time, except for two strong accelerations in the 2046–2047 and 2067–2068 projections, which correspond to the prediction of tick population establishment in remote municipalities with large surface areas. Removing these municipalities did not significantly change speed estimate.

Fig 2

Predicted increase in the area of tick establishment and the proportion of the human population living in municipalities with established tick population in Québec from 2008 to 2100.

Dashed lines show the percentage of the surface area of the province of Québec predicted to contain an established tick population by the lower, mean and upper predicted years (light, medium and dark grey, respectively). Solid lines show the percentage of the Québec human population living in areas predicted to have an established tick population by the lower, mean and upper predicted years (light, medium and dark grey lines, respectively).

Exposed human population

According to the model predictions and based on the mean predicted year, 58% of the Québec population were predicted to live in a municipality with an established tick population by 2012, 85% by 2020 and 90% by 2027 (Fig 2). Only one region was concerned in 2008, then 10 regions predicted in 2020 and 16 in 2030 among the 19 regions of Quebec (Fig 3).

Fig 3

Prediction of tick population establishment in Québec from 2008 to 2100 by 5-year intervals.

Predicted year of tick population establishment for each Québec municipality based on the upper predicted year of the model of Leighton et al. (2012) [19].

Regional specificity

Predictions up to 2100 show a northern expansion of tick populations from southern Québec and along the St. Lawrence River (Fig 3). However, in some municipalities, tick establishment varies from surrounding municipalities, with cluster of earlier or later tick establishment. Most regions currently affected by Lyme disease have established tick populations predicted by the model. In the other regions, tick establishment should be expected in the coming years.

Discussion

In our study, we showed that the “time-to-establishment” model of Leighton et al. [19] successfully captured the overall pattern of presence of at least one I. scapularis tick and the establishment of a tick population observed through active surveillance in Québec from 2010–2018. Concordance between observations and predictions was relatively low but statistically significant, with no major geographical inconsistencies. The mean predicted year of establishment seems to be a good predictor of tick presence, while the upper predicted year provided a better predictor of the detection of all three tick stages in the field. The predictions correctly reflect the known epidemiological situation in 2018, and the model seems sufficiently reliable to estimate the progression of tick range expansion in the province of Québec, with a predicted average speed of 18 km/year. The model suggests a continued northward expansion of I. scapularis’ range in the coming decades, with 90% of Québec’s population living in a tick-endemic area by 2027.

Lyme disease risk is associated with the presence of an established tick population [28–30], but “establishment” remains difficult to define

An important goal for public health authorities is to predict where and when tick populations will establish in order to inform and protect human populations. Established tick populations reproduce locally, but different criteria have been used to operationally define “establishment” based on surveillance data. The first Canadian definition of established tick populations was the detection of all three tick life stages for at least two consecutive years [12]. However, this definition required significant human and financial resources to repeatedly collect ticks in the environment over a two-year period. In Québec, the province has used the detection of all three tick stages within the same year [6] to define an established tick population. However, Québec also used the criterion of “the presence of at least one tick” to determine Lyme disease risk level by municipality [5]. In contrast, the model developed by Leighton et al. [19] was designed to predict the year of tick population establishment based on a passive surveillance indicator of the detection of at least two adult ticks or one immature tick by active surveillance for two consecutive years. It was therefore important to compare these two definitions with the model predictions in order to understand how to use this model in public health in Québec.

Concordance between field data and model predictions is low but statistically significant, probably as a consequence of limits inherent to active surveillance and modelling tools

On the one hand, the sensitivity of drag sampling for detecting ticks in the environment is low because the efficiency of dragging depends on meteorological conditions during sampling [13,26] and because ticks have a heterogeneous distribution at a fine geographical scale [30-34]. Detection of I. scapularis larvae, nymph and adults during the same visit is highly unlikely since they are generally active during different seasons [11]. Not observing a tick during a single visit to a municipality may also lead to the incorrect conclusion of absence. On the other hand, detection of any tick stage, especially if only one tick is found, does not always indicate an established tick population. Instead, such observations could be adventitious ticks carried by vertebrate hosts (migratory birds, rodents, deer) into areas without established tick populations [32,35,36]. In our active surveillance data, for 30% of municipalities with a first tick detection, ticks were not systematically detected during sampling in subsequent years. These limits likely explain the low concordance we observed: with an agreement of 65% between predictions and observations, the model is better than flipping a coin to predict tick presence or detection of all three stages in active surveillance in a given municipality and year, but shows better agreement (74%) when cumulating surveillance data over the study period. In fact, multiple sampling events for each municipality reduce the sampling error due to the failure of detecting ticks at a site where they are in fact established. These limits also explain the better results with the criterion of “presence of at least one tick” rather than the criterion of “presence of three stages”, which is harder to attain with active surveillance methods in Québec. Moreover, because of limitations in logistical capacity, active surveillance is not carried out every year in all municipalities in Québec; this incomplete sampling underestimates tick establishment and contributes to the time lag between the predicted year of establishment and detection of ticks by field sampling (three years on average in our study).

Model predictions provide an accurate overall portrait of tick range expansion, and fine-scale variation in local establishment is expected in areas of emerging risk

Some divergence between predictions and field data may be explained by factors that influence local tick establishment such as forest type, micro-climate, and distribution and behavior of vertebrate hosts. These factors were not included in the model of Leighton et al. [19]. This could explain some divergence between predictions and observations, as will also influence the real progression of ticks over the coming decade. Temperature is thought to be the most important factor determining suitable areas for tick establishment [16,37], but other factors could play an important in northern Québec where the habitat is dominated by coniferous forest, considered to be marginal habitat for ticks, and where vertebrate host communities are different from southern Québec. Leighton et al. [19] suggested that rainfall and elevation may also influence the speed of tick population establishment. Moreover, previous studies conducted at finer spatial scales highlighted the heterogeneous distribution of ticks [30-34], probably as a result of differences in local environmental conditions and host communities that support local and long-distance dispersal of ticks. At the Québec scale, the predicted established range of the tick in 2018 correctly overlaps the confirmed distribution of ticks based on active surveillance (presence or three stage), suggesting that model predictions are reliable at this scale. In addition, the small number of municipalities with detection of the three tick stages that are located outside of the predicted area of establishment are situated very close to the predicted area. Moreover, we noticed local particularities in the predictive map (e.g., zones with tick establishment earlier or later than surrounding area) which were also observed in the acarological surveillance data (results not presented here). Overall, the model successfully predicted the progression of tick range expansion over the past decade, and observed fine-scale variation in local establishment expected within this emerging risk area. Further studies could improve the existing model and refine predictions by integrating more recent environmental and surveillance data.

Because field observations were consistent with model predictions, we explored the speed at which tick population establishment progressed across Québec

According to the model, the overall speed of expansion of the established range of I. scapularis in Québec was estimated to be 18 km/year [95% CI; 15–23 km/year], which is lower than the value of 46 km/year for eastern Canada suggested by Leighton et al. [19]. The original model was for all eastern Canadian provinces (i.e., Québec, Ontario, etc.), but differences in the speed of range expansion among provinces were not explored in that study. A recent study carried out in southern Ontario corroborated the speed of 46 km/year [20], by comparing the predicted year of establishment of a tick population within a radius of 46 km from a site with an established tick population; however, the authors did not calculate the speed predicted by the model in Ontario as was done here. Using a model based on degree days > 0°C [16,17] and tick abundance, Simon et al. [18] estimated a northward expansion of I. scapularis of 300 km from 2011 to 2050 in southern Québec, representing a speed of tick range expansion of 7 km/year, which is more consistent with the estimate from the present study. The model by Simon et al. [18] was based on environmental variables and was not biased by administrative boundaries. In our study, because we calculated speed based on the difference between the total areas of municipalities with established tick populations area in two consecutive years, larger municipalities, especially in northern Québec, can artificially increase estimates of the speed of tick progression. The real speed of establishment also depends on suitable habitat and the composition of host communities, and won’t be consistent over time and space even within a municipality. Finally, beyond the general acceleration of range expansion expected with warming climate conditions [19], climate change is likely to impact the observed speed of tick range expansion in the coming decades through its multiple effects on the different components of tick ecology.

Application in public health

Despite the aforementioned limitations, the Leighton et al. [19] model reliably estimated the recent pattern of tick range expansion in Québec, and by extension the risk of human Lyme disease, providing a useful projection of risk for the coming decades. The presence of at least one tick detected by active surveillance is a criterion used to identify “at-risk” municipalities in Québec [5] and may reflect the presence of newly established tick populations. The detection of the three tick stages during the same season by active surveillance, with detection of B. burgdorferi in at least one nymph, reflects the presence of an established tick population and is used to identify “endemic” municipalities in Québec [5]. Consequently, the mean predicted year of establishment could be a useful indicator “at-risk areas” while the upper predicted year could be used as an as indicator of “endemic areas”, currently and for the coming decades. This approach helps to estimate the present and future risk levels of all municipalities in Québec, and could guide future active surveillance activities. Since the distribution of B. burgdorferi is expected to follow a similar trajectory to that of I. scapularis [22], the projection maps could help anticipate future changes in the epidemiology of Lyme disease in Québec. The distribution of tick also provides an estimate of the exposed human population, with potentially 90% of Québec’s population living in an endemic area by 2027, which could be refined using the demographic projections for Québec. Even though other factors influence the number of human cases of Lyme disease (e.g., human behavior, prevention, diagnosis) [4], reliable projections of future risk could nevertheless help public health authorities develop preventive measures, particularly in regions not yet affected by Lyme disease.

Conclusion

Our study demonstrated that a predictive model of tick range expansion can be used to provide reliable projections of changes in the distribution of tick-endemic areas over a 10-year period, and useful estimates of the changing risk of Lyme disease at the municipality level in Québec for the coming years. Assessing the concordance between a predictive model and field data is an important step in evaluating such prediction tools for use by public health authorities. This model should be updated as new field surveillance data become available in Québec in the coming years, and could be similarly validated for use in predicting future spread of ticks and Lyme disease risk within other Canadian provinces.

(DOC)

Click here for additional data file.

3 Jun 2021

PONE-D-21-14260

Current and future distribution of Ixodes scapularis ticks in Québec: field validation of a predictive model previously developed in Canada

PLOS ONE

Dear Dr. Ripoche,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Both reviewers felt this study is a worthwhile contribution to the field. Please see attached reviewer's comments and pay careful consideration to grammatical errors.

Please submit your revised manuscript by Jul 18 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see:  http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols . Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at  https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols .

We look forward to receiving your revised manuscript.

Kind regards,

Catherine A. Brissette, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

3. We note that Figures 1 and 3 in your submission contain map images which may be copyrighted. All PLOS content is published under the Creative Commons Attribution License (CC BY 4.0), which means that the manuscript, images, and Supporting Information files will be freely available online, and any third party is permitted to access, download, copy, distribute, and use these materials in any way, even commercially, with proper attribution. For these reasons, we cannot publish previously copyrighted maps or satellite images created using proprietary data, such as Google software (Google Maps, Street View, and Earth). For more information, see our copyright guidelines: http://journals.plos.org/plosone/s/licenses-and-copyright.

We require you to either (1) present written permission from the copyright holder to publish these figures specifically under the CC BY 4.0 license, or (2) remove the figures from your submission:

3.1.    You may seek permission from the original copyright holder of Figures 1 and 3 to publish the content specifically under the CC BY 4.0 license.

We recommend that you contact the original copyright holder with the Content Permission Form (http://journals.plos.org/plosone/s/file?id=7c09/content-permission-form.pdf) and the following text:

“I request permission for the open-access journal PLOS ONE to publish XXX under the Creative Commons Attribution License (CCAL) CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). Please be aware that this license allows unrestricted use and distribution, even commercially, by third parties. Please reply and provide explicit written permission to publish XXX under a CC BY license and complete the attached form.”

Please upload the completed Content Permission Form or other proof of granted permissions as an "Other" file with your submission.

In the figure caption of the copyrighted figure, please include the following text: “Reprinted from [ref] under a CC BY license, with permission from [name of publisher], original copyright [original copyright year].”

3.2.    If you are unable to obtain permission from the original copyright holder to publish these figures under the CC BY 4.0 license or if the copyright holder’s requirements are incompatible with the CC BY 4.0 license, please either i) remove the figure or ii) supply a replacement figure that complies with the CC BY 4.0 license. Please check copyright information on all replacement figures and update the figure caption with source information. If applicable, please specify in the figure caption text when a figure is similar but not identical to the original image and is therefore for illustrative purposes only.

The following resources for replacing copyrighted map figures may be helpful:

USGS National Map Viewer (public domain): http://viewer.nationalmap.gov/viewer/

The Gateway to Astronaut Photography of Earth (public domain): http://eol.jsc.nasa.gov/sseop/clickmap/

Maps at the CIA (public domain): https://www.cia.gov/library/publications/the-world-factbook/index.html and https://www.cia.gov/library/publications/cia-maps-publications/index.html

NASA Earth Observatory (public domain): http://earthobservatory.nasa.gov/

Landsat: http://landsat.visibleearth.nasa.gov/

USGS EROS (Earth Resources Observatory and Science (EROS) Center) (public domain): http://eros.usgs.gov/#

Natural Earth (public domain): http://www.naturalearthdata.com/

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Title: Current and future distribution of Ixodes scapularis ticks in Québec: field validation of a predictive model previously developed in Canada

Authors: Marion Ripoche, Catherine Bouchard, Alejandra Irace-Cima, Patrick Leighton, Karine Thivierge

Summary of the study: The spread of Ixodes scapularis ticks into Canada and the resultant increase in the incidence of Lyme borreliosis in the Canadian population is a well-documented example of how climate change can increase the public health burden of vector-borne diseases. A model by Leighton and colleagues predicted the speed at which blacklegged ticks would expand their range in eastern Canada and establish new tick populations (1). The purpose of the present study is to test how this model performs in southern Quebec where the authors have spent 9 years (2010 – 2018) doing active surveillance for blacklegged ticks by using the dragging method. The authors conducted 444 tick sampling sessions in 231 census subdivisions (CSDs) between 2010 and 2018. The authors used two different criteria to determine whether a particular site had an established tick population: (1) sampling of all three tick stages (strict definition) and (2) sampling of a single blacklegged tick (less restrictive definition). The authors used the model by Leighton and colleagues to predict the year of blacklegged tick population establishment for each of the sampling sessions and CSDs (they used a mean, lower, and upper estimate). They then compared these predicted values (3 predictions) to the observed values from their active surveillance (2 criteria for an established tick population). The general result was that active surveillance was more likely to find ticks in sites that were predicted to have established tick populations according to the model compared to sites that were not predicted to have established tick populations. For the 444 sampling occasions, the model based on the mean year predicted 65.1% correctly (289/444) and 34.9% (155/444) incorrectly (Table 1). For the 231 CSDs, the model based on the upper year predicted 74.0% correctly (171/231) and 26.0% (60/231) incorrectly (Table 3). The model predicts that 90% of the human population in Quebec will be leaving in an area with an established tick population by 2027 (prediction based on the mean).

Major comments:

-The authors have used a straightforward approach to answering their question, and I don’t have too many criticisms of the study.

-The authors show that active surveillance is more significantly more likely to find ticks in areas that were predicted to have established tick populations compared to sites that were not. However, it would be useful if the authors came up with an overall measure of the ability of their model to predict the correct answer. For example, in Table 1 for the lower predicted year, the model predicted 55.4% of the sampling occasions correctly [(199 + 47)/444 = 246/444] and 44.6% of the sampling occasions incorrectly [(179 + 19)/444 = 198/444]. In other words, for this particular combination of prediction (lower year) and outcome (1 tick present is an established tick population), the model is not much better than flipping a coin. It would be good if this was pointed out in the manuscript. For the 231 CSDs, the model does better and for the upper year, it predicts 74.0% (171/231) of the CSDs correctly and 26.0% of the CSDs incorrectly (60/231). Again, it would be good to point out that for this combination of prediction (upper year) and outcome (1 tick present is an established tick population), the model performs better than a random coin toss.

-In the Discussion, the authors point out that the original model by Leighton and colleagues (1) predicted a speed of progression of established tick populations of 46 km per year. The estimated speed in the current study is considerably slower, 18 km per year over the time period of 2020 – 2100 and 22 km per year over the time period 2020 – 2030. A study in southern Ontario confirmed the speed of progression of established tick populations of 46 km per year, whereas a study in southern Quebec found a speed of progression of 7 km per year. In the Discussion, it is not really clear why the speed of progression differs between Quebec and Ontario, and some more clarification would be helpful.

-In the Discussion, the authors discuss the application of their model for public health. However, according to their model predictions, 90% of the human population of Quebec will be living in an area with an established tick population by 2027. If this prediction comes true, what is there left to say for public health authorities other than that almost the entire population in Quebec is at risk for Lyme disease and ask everyone to be careful?

-The manuscript was formatted with bullets. Is this a new formatting requirement for PLOS ONE manuscripts? If not, please don’t use bullets, as it gives reviewers the impression that the manuscript was hastily prepared.

-The manuscript contained many small grammatical errors. The authors should do a better job of checking the manuscript for such errors before submitting it to a journal. I have attached a Word document with ‘Track changes’ that documents a number of these grammatical errors.

Advice for future studies:

-The authors point out that the dragging method does not always detect the ticks at a particular site, even if the site is believed to have an established tick population. In future studies, the authors should consider using site occupancy models to address this problem. For example, you may sample the same site in 2016, 2017, and 2018, and find ticks in 2016 and 2018, but not in 2017. If we assume that the ticks were there in 2017, but were not detected, we can estimate the probability of not detecting ticks at a site with an established tick population. By separately estimating the probability of detection versus the probability of an established tick population, you will get better estimates of the latter. In Table 1, the best model (based on the mean year) predicts 65.1% of the sampling occasions correctly, whereas in Table 3, the best model (based on the upper year) predicts 74.0% of the CSDs correctly. This comparison suggests that failing to detect ticks in a site where they are established is less of a problem for the CSDs than the sampling occasions. This observation makes sense, because there are multiple sampling occasions for each CSD, which reduces the error due to the failure of detecting ticks at a site where they are established. If you conducted a very large number of sampling occasions at each site (e.g., hundreds), you would eventually have perfect knowledge of which sites had established tick populations and which did not.

-The authors do not give much insight as to why their model is not working better. They point out that a number of factors are not included in the original model by Leighton and colleagues (1). For example, the model is heavily reliant on temperature, but does not include elevation and rainfall. This study would have been more interesting if the authors had dropped components from the original model to quantify the contribution of those components to the accuracy of the predictions. Thus, an unsatisfying feature of this study is that the authors and the readers don’t really learn anything new about how the model by Leighton et al. could be improved. If the authors plan on future studies investigating the predictive ability of the model by Leighton et al., they should consider comparing different sub-models that drop different components of the original model, so that we learn which components of the models are actually important for making the correct predictions.

References

1. Leighton PA, Koffi JK, Pelcat Y, Lindsay LR, Ogden NH. 2012. Predicting the speed of tick invasion: an empirical model of range expansion for the Lyme disease vector Ixodes scapularis in Canada. J Appl Ecol 49:457-464.

Reviewer #2: GENERAL REMARKS

The study validates a predictive model on tick distribution in Canada, by comparing collected field data with the model predictions. These types of model evaluations are useful to assess and improve on the predictive model. However, although I would consider this study to be worthwhile, there are many linguistic errors to be addressed before the publication of this manuscript. Please find more specific remarks below.

ABSTRACT

If the use of title words such as Background, Objective etc. was a conscious decision, these should be clearer. Perhaps adding “:” or a line break can make this distinction between different subheadings more readable.

Line 26: “The incidence of Lyme disease”, or “The recorded number of Lyme disease cases” is increasing in Québec…

Line 27: remove “,” after “2012”

Line 28: put “until 2100” after “establishment”

Line 44: I would add “Ixodes scapularis” to the list of key words

INTRODUCTION

Line 49: Locally acquired cases (per year? In total?) please specify. You could say “increased from 144 annual cases in 2009, to 2636 in 2019”. Put “in Canada” at the beginning of the sentence.

Line 51: “is progressing”

Line 58: check commas and punctuation in general

Line 77: “Active acarological surveillance”

Line 95: all three stages

METHODS

Study area: As this is a study assessing the viability of a model based on climatic data: Some general climatic information (mean yearly temperatures, precipitation), would be welcome. This is also true for the elevational range (as this is factored into the model), and overall vegetation composition (what type of forests).

Some general information on I. scapularis would be welcome, including the vectorial capacity of this species, as well as its ecology (questing behavior and host-parasite relations).

Line 104: “province”

Line 117: tick distribution “data”, comes from …

Line 123: which configuration? Lines, quadrats, one single 2000m line? Specify

Line 130: spacing after citation

Line 131: year of establishment

Line 132: years

Line 139: “tick populations were”, or “a tick population was” supposed to be established

Line 138: remove first “of the model”, remove first comma

Line 146: “A tick population” or “tick populations”, there where and when…

Line 147: Predicted for it

Line 151: “if a tick population is predicted to be established in the area, then ticks are also expected to be found in the field data”

Line 169: for clarity, I suggest inserting a full equation in word in the form of:

Speed[km/year]=√(area*year*n)- √(area*year*n-1)

Line 172: “exposed to ticks”, remove first comma,

Line 173: “predicted an established tick population”

Line 175: “across the population of Quebec”

Line 176: “…of established tick populations”, “was produced/generated”

RESULTS

Line 184: add percentage

Line 199: “twice on average”

Line 209: “Concordance per collection”

Table 1: Is there a reason why the three stages column data is in italic? If so, specify.

Table 2: If you use “Sd” in the table, the *text should be “Sd” as well, not “sd”.

Table 1 and 3: footnote: “are included”

Line 240: remove second “detection”

Line 242: “the three stages”

Line 245-246: If you include the statistical result, please also include the p-value

DISCUSSION

I am missing some recommendations on if/how the findings of this study (field validation) could be used to enhance the model developed by Leighton (2012).

Line 298: “time-to-establishment”

Line 310: where and when “A tick population” or “tick populations” will establish

Line 315: Collect ticks

Line 320: Years

Line 321: remove comma. Risk of what? Specify Lyme disease risk

Line 322: “with the model predictions”

Line 338-339: Because of “limitations” in “logistical” capacity

Line 349: “influential factor”

Line 350: northern, sometimes “Quebec” is written “Québec”. Please be consistent.

Line 351: host communities are different how?

Line 352: “A previous study”, or “previous studies”

Line 363: remove comma

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: QuebecTickEstablishment_submission_20210428_v02.doc

Click here for additional data file.

16 Oct 2021

Dear reviewers,

Thanks for your pertinent comments and suggestion, it was a real contribution for the manuscript. Thank you for your understanding about my English, which is unfortunately far from perfect, and for making the effort to overpass or correct it by reviewing the article, it is much appreciated.

You will find below and in the revised manuscript the response to your comments.

Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Title: Current and future distribution of Ixodes scapularis ticks in Québec: field validation of a predictive model previously developed in Canada

Authors: Marion Ripoche, Catherine Bouchard, Alejandra Irace-Cima, Patrick Leighton, Karine Thivierge

Summary of the study: The spread of Ixodes scapularis ticks into Canada and the resultant increase in the incidence of Lyme borreliosis in the Canadian population is a well-documented example of how climate change can increase the public health burden of vector-borne diseases. A model by Leighton and colleagues predicted the speed at which blacklegged ticks would expand their range in eastern Canada and establish new tick populations (1). The purpose of the present study is to test how this model performs in southern Quebec where the authors have spent 9 years (2010 – 2018) doing active surveillance for blacklegged ticks by using the dragging method. The authors conducted 444 tick sampling sessions in 231 census subdivisions (CSDs) between 2010 and 2018. The authors used two different criteria to determine whether a particular site had an established tick population: (1) sampling of all three tick stages (strict definition) and (2) sampling of a single blacklegged tick (less restrictive definition). The authors used the model by Leighton and colleagues to predict the year of blacklegged tick population establishment for each of the sampling sessions and CSDs (they used a mean, lower, and upper estimate). They then compared these predicted values (3 predictions) to the observed values from their active surveillance (2 criteria for an established tick population). The general result was that active surveillance was more likely to find ticks in sites that were predicted to have established tick populations according to the model compared to sites that were not predicted to have established tick populations. For the 444 sampling occasions, the model based on the mean year predicted 65.1% correctly (289/444) and 34.9% (155/444) incorrectly (Table 1). For the 231 CSDs, the model based on the upper year predicted 74.0% correctly (171/231) and 26.0% (60/231) incorrectly (Table 3). The model predicts that 90% of the human population in Quebec will be leaving in an area with an established tick population by 2027 (prediction based on the mean).

Major comments:

-The authors have used a straightforward approach to answering their question, and I don’t have too many criticisms of the study.

MR:ok

-The authors show that active surveillance is more significantly more likely to find ticks in areas that were predicted to have established tick populations compared to sites that were not. However, it would be useful if the authors came up with an overall measure of the ability of their model to predict the correct answer. For example, in Table 1 for the lower predicted year, the model predicted 55.4% of the sampling occasions correctly [(199 + 47)/444 = 246/444] and 44.6% of the sampling occasions incorrectly [(179 + 19)/444 = 198/444]. In other words, for this particular combination of prediction (lower year) and outcome (1 tick present is an established tick population), the model is not much better than flipping a coin. It would be good if this was pointed out in the manuscript. For the 231 CSDs, the model does better and for the upper year, it predicts 74.0% (171/231) of the CSDs correctly and 26.0% of the CSDs incorrectly (60/231). Again, it would be good to point out that for this combination of prediction (upper year) and outcome (1 tick present is an established tick population), the model performs better than a random coin toss.

MR:I agree, it is clearer like that, I added that in the result and discussion.

-In the Discussion, the authors point out that the original model by Leighton and colleagues (1) predicted a speed of progression of established tick populations of 46 km per year. The estimated speed in the current study is considerably slower, 18 km per year over the time period of 2020 – 2100 and 22 km per year over the time period 2020 – 2030. A study in southern Ontario confirmed the speed of progression of established tick populations of 46 km per year, whereas a study in southern Quebec found a speed of progression of 7 km per year. In the Discussion, it is not really clear why the speed of progression differs between Quebec and Ontario, and some more clarification would be helpful.

MR: We calculated the speed with the data from the model. I don’t know the predicted speed for Ontario.

-In the Discussion, the authors discuss the application of their model for public health. However, according to their model predictions, 90% of the human population of Quebec will be living in an area with an established tick population by 2027. If this prediction comes true, what is there left to say for public health authorities other than that almost the entire population in Quebec is at risk for Lyme disease and ask everyone to be careful?

MR:It is true. But it is important to explain to region with no established tick population for the moment, that it will be coming in the next decades, even for region in the north with a rude winter

-The manuscript was formatted with bullets. Is this a new formatting requirement for PLOS ONE manuscripts? If not, please don’t use bullets, as it gives reviewers the impression that the manuscript was hastily prepared.

MR:the bullet was to mark the section and sub-section. I deleted them.

-The manuscript contained many small grammatical errors. The authors should do a better job of checking the manuscript for such errors before submitting it to a journal. I have attached a Word document with ‘Track changes’ that documents a number of these grammatical errors.

MR: Thanks for having corrected grammatical errors. I really appreciate that, and I will find a way to check before submission the next time

Advice for future studies:

-The authors point out that the dragging method does not always detect the ticks at a particular site, even if the site is believed to have an established tick population. In future studies, the authors should consider using site occupancy models to address this problem. For example, you may sample the same site in 2016, 2017, and 2018, and find ticks in 2016 and 2018, but not in 2017. If we assume that the ticks were there in 2017, but were not detected, we can estimate the probability of not detecting ticks at a site with an established tick population. By separately estimating the probability of detection versus the probability of an established tick population, you will get better estimates of the latter. In Table 1, the best model (based on the mean year) predicts 65.1% of the sampling occasions correctly, whereas in Table 3, the best model (based on the upper year) predicts 74.0% of the CSDs correctly. This comparison suggests that failing to detect ticks in a site where they are established is less of a problem for the CSDs than the sampling occasions. This observation makes sense, because there are multiple sampling occasions for each CSD, which reduces the error due to the failure of detecting ticks at a site where they are established. If you conducted a very large number of sampling occasions at each site (e.g., hundreds), you would eventually have perfect knowledge of which sites had established tick populations and which did not.

MR: I agree with you, but active surveillance in Quebec is currently limited to one visit per municipality and per year, and aa we are note able to visit all municipality each year, there is a turn over. But we have some sentinels sites, sampled each year, that could be used to estimate the probability of detection or not detection in a site with established tick population.

-The authors do not give much insight as to why their model is not working better. They point out that a number of factors are not included in the original model by Leighton and colleagues (1). For example, the model is heavily reliant on temperature, but does not include elevation and rainfall. This study would have been more interesting if the authors had dropped components from the original model to quantify the contribution of those components to the accuracy of the predictions. Thus, an unsatisfying feature of this study is that the authors and the readers don’t really learn anything new about how the model by Leighton et al. could be improved. If the authors plan on future studies investigating the predictive ability of the model by Leighton et al., they should consider comparing different sub-models that drop different components of the original model, so that we learn which components of the models are actually important for making the correct predictions.

MR:Yes, in further study we could try to improve the model with more recent environemental and surveillance data than in the original model.

References

1. Leighton PA, Koffi JK, Pelcat Y, Lindsay LR, Ogden NH. 2012. Predicting the speed of tick invasion: an empirical model of range expansion for the Lyme disease vector Ixodes scapularis in Canada. J Appl Ecol 49:457-464.

Reviewer #2: GENERAL REMARKS

The study validates a predictive model on tick distribution in Canada, by comparing collected field data with the model predictions. These types of model evaluations are useful to assess and improve on the predictive model. However, although I would consider this study to be worthwhile, there are many linguistic errors to be addressed before the publication of this manuscript. Please find more specific remarks below.

ABSTRACT

If the use of title words such as Background, Objective etc. was a conscious decision, these should be clearer. Perhaps adding “:” or a line break can make this distinction between different subheadings more readable.

Line 26: “The incidence of Lyme disease”, or “The recorded number of Lyme disease cases” is increasing in Québec…

Line 27: remove “,” after “2012”

Line 28: put “until 2100” after “establishment”

Line 44: I would add “Ixodes scapularis” to the list of key words

MR: done – changed in the text

INTRODUCTION

Line 49: Locally acquired cases (per year? In total?) please specify. You could say “increased from 144 annual cases in 2009, to 2636 in 2019”. Put “in Canada” at the beginning of the sentence.

Line 51: “is progressing”

Line 58: check commas and punctuation in general

Line 77: “Active acarological surveillance”

Line 95: all three stages

MR: done – changed in the text

METHODS

Study area: As this is a study assessing the viability of a model based on climatic data: Some general climatic information (mean yearly temperatures, precipitation), would be welcome. This is also true for the elevational range (as this is factored into the model), and overall vegetation composition (what type of forests).

MR:

Some general information on I. scapularis would be welcome, including the vectorial capacity of this species, as well as its ecology (questing behavior and host-parasite relations).

MR: changed in introduction l. 52-58

Line 104: “province”

Line 117: tick distribution “data”, comes from …

Line 123: which configuration? Lines, quadrats, one single 2000m line? Specify

Line 130: spacing after citation

Line 131: year of establishment

Line 132: years

MR: done – changed in the text

Line 139: “tick populations were”, or “a tick population was” supposed to be established

Line 138: remove first “of the model”, remove first comma

Line 146: “A tick population” or “tick populations”, there where and when…

Line 147: Predicted for it

Line 151: “if a tick population is predicted to be established in the area, then ticks are also expected to be found in the field data”

Line 169: for clarity, I suggest inserting a full equation in word in the form of:

Speed[km/year]=√(area*year*n)- √(area*year*n-1)

MR: done – changed in the text

Line 172: “exposed to ticks”, remove first comma,

Line 173: “predicted an established tick population”

Line 175: “across the population of Quebec”

Line 176: “…of established tick populations”, “was produced/generated”

MR: done – changed in the text

RESULTS

Line 184: add percentage

Line 199: “twice on average”

Line 209: “Concordance per collection”

MR: done – changed in the text

Table 1: Is there a reason why the three stages column data is in italic? If so, specify.

MR: yes, because the number of collections with three stages are included in tick presence (cf. footnote of table), it’s a sub-part of collection with tick presence

Table 2: If you use “Sd” in the table, the *text should be “Sd” as well, not “sd”.

Table 1 and 3: footnote: “are included”

Line 240: remove second “detection”

Line 242: “the three stages”

Line 245-246: If you include the statistical result, please also include the p-value

MR: done – changed in the text

DISCUSSION

I am missing some recommendations on if/how the findings of this study (field validation) could be used to enhance the model developed by Leighton (2012).

MR: added in the text Line 430

Line 298: “time-to-establishment”

Line 310: where and when “A tick population” or “tick populations” will establish

Line 315: Collect ticks

Line 320: Years

Line 321: remove comma. Risk of what? Specify Lyme disease risk

Line 322: “with the model predictions”

Line 338-339: Because of “limitations” in “logistical” capacity

Line 349: “influential factor”

Line 350: northern, sometimes “Quebec” is written “Québec”. Please be consistent.

MR: done – changed in the text

Line 351: host communities are different how?

MR: vertebrate host communities are different from southern Québec

Line 352: “A previous study”, or “previous studies”

Line 363: remove comma

MR: done – changed in the text

Submitted filename: 20211004_Response to Reviewers.doc

Click here for additional data file.

25 Nov 2021

7 Jan 2022

________________________________________

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2:

GENERAL REMARKS

The proposed revised manuscript is a substantial improvement on the initial submission. Particularly, the discussion now thoroughly outlines the limitations of the findings and the methods, and how these may be improved in the future. Overall, I would consider this manuscript fit for publication, if a few minor changes are implemented.

Line 1: “TITLE”

MR: done

INTRODUCTION

Line 51: “climatic” conditions

Line 52: …deciduous and mixed forests are “considered” suitable “habitats”…

Line 54: remove comma

Line 61: …drag sampling “method”.

Line 76: “northwards”

MR: done

Line 97 and lower: If chosen to make et al. italic, please do so throughout the manuscript.

MR: done

METHODS

Line 128-129: During the initial review, a request was made to specify the configuration of the 2000m2 area that was sampled using the flag dragging method. This was not done. It would be suitable to specify whether the sampling was done in simple line of 2000m, or if quadrats were used and if so, in which sampling structure.

MR: done. I added explanation in text, line 130-131

Line 155-156: This question is repeating the same declarative statement of the sentence before. I suggest removing this question.

MR: I prefer to keep the question to clarify for the lector, explaining the analysis in another way

Line 184: Remove empty line to keep the spacing consistent with the other chapters

MR: I standardized between sections

RESULTS:

Figure 2: Having black x and y labels would greatly enhance readability.

MR: done

DISCUSSION

Of all the chapters, the discussion has improved the most compared to the initial submission.

MR: Thanks

Line 368: “limitations”

MR: done

Reviewer #3:

This study uses field data to validate a time-to-establishment model developed in 2012 by one of the authors to predict the Ixodes scapularis population establishment for each municipality in eastern Canada. This field data consisted of active tick surveillance data collected betweem 2010-2018, where two criteria were used to define tick establishment: i) either the detection of at least one tick or ii) the detection of the three questing stages of the tick.

Overall I really like the premise of this paper, dragging/flagging data are valuable but rarely get published, and I like that the authors have used a more creative/informative way of doing this by using it to validate a model (which in itself is also worthwhile).

MR: Thanks

However there are some potentially significant issues with the use and interpretation of statistics, and some of the methods need more detail, for example how the human population increase was calculated and exactly when flagging/dragging (it is unclear which technique was used) was performed.

MR: I tried to clarify that in the text

Introduction

The introduction does a good job of introducing the topic, highlights the importance of the paper and clearly sets out the aims of the study.

MR: Thanks

Line 48-54: All nice points! I would also argue that improved awareness of Lyme disease in Canada, and therefore improved diagnostics and reporting are also likely to have caused the number of human cases to increase.

MR: I agree with you

Line 62: “In Québec, the number of 62 locally acquired human cases has increased from 2 cases in 2008 to 381 by 2019 (7)”

Human cases of what? Tick bite or Lyme disease? Clarify, and if Lyme disease I suggest moving this to line 58 after “a notifiable human disease in Québec since 2003.” for better information flow.

MR: done.

Lines 67-68: Which months are considered spring, summer and autumn in the study area (this is important when considering the months you sampled ticks).

MR: done.

Materials and Methods

Line 107: “The province of” should not be in bold

MR: done.

Line 122-132: There is some inconsistency in the description of tick surveillance, was it dragging with a drag cloth or flagging with a flag cloth (they can target different life stages of the tick).

MR: I added details about dragging methods

How frequently was flagging/dragging performed at each site?

MR: I added details and there is also information in results, line 209-210

Line 126 you mention tick sampling in June-September, which life stages would you realistically be likely to encounter in each of these months at the study location? This information should be included as finding each of the life stages is an important underpinning of your analyses?

MR: I added details

It is very hard for me to understand how the methods for sampling event concordance (line 146-152) and concordance over the study period (line 153-163) differ. After re-reading a few times I think I understood that concordance over the study period aggregated all tick data from 2010-2018 whereas sampling event concordance looked year by year? I think it would help the reader if this was made more explicit.

MR: yes, that is the difference. I added some details in the text.

Line 165: Please clarify your use, justification, and interpretation of McNemar’s. this is not truly paired data and McNemar does not assess for concordance.

MR: done. Line 172: “A significant Chi-square test, with p<0.05, suggests that the distribution of sites with presence of ticks is not at random between areas with expected or unexpected established tick population according to the model. We used the matched McNemar test for the analysis of all the sampling because of some repeated samples at the same site”

Line 177-179: “We cumulated the annual exposed human population from 2006 to 2100, and calculated the annual proportion of the Québec population exposed, based on the 2016 census data”. Cumulated and calculated how? Please add more detail.

MR: done

Results

Line 202-204: This should be in the methods. Do I therefore understand correctly that sites were not sampled every year? In which case this also should be made more explicit in sampling event concordance (line 146-152) and concordance over the study period (line 153-163). It would be nice to have a table of when sites were sampled, with the raw tick data and what the model predicted (lower, mean and upper year).

MR: I added some details in methods. I let description of data in result. Data are available is a request to specific person.

Line 211-214: This really depends when sampling was performed (and another reason for wanting to see a table of exactly when sites were sampled and the raw tick data), as you mention in lines 67-68 the different life stages are active at very different times of year. So, if for example you happened to sample a bunch of municipalities in early spring when larvae and adults are unlikely to be active you are going to have a very different chance of finding all three life stages than if you sample in the summer, possibly biasing your results.

MR: Sites are sampling generally once, between may and September. In endemic area, we are able to catch nymph, adults and larvae in spring. In discussion, this is the main limitation of active surveillance data (line 369), and that is why we assess the concordance with presence of tick and not only presence of confirmed established tick population.

Line 306-311: It would be nice to add some further details here, for example, how many regions affected by Lyme disease? Is the coming years 1, 2, 5 or 10 years? By how much to neighbouring municipalities differ? Is there any pattern to this?

MR: I added some details in the text (Line 310). We didn’t analyse the pattern of expansion, because it was outside the scope of our study.

Figure 2: I propose changing the figure legend title to “proportion of the human population living in municipalities with established tick populations in Québec from 2008 to 2100” or similar, as any of the population could be “exposed” just by going one day for hike in a location with ticks.

MR: done

Discussion

The general content of the discussion looks good and relevant, but I am hesitant at this stage to review it more thoroughly as I have concerns with the use and interpretation of statistics, and will be happy to read this in more detail once the authors have had a chance to respond to these concerns.

Line 323-325: “Concordance between observations and predictions was relatively low but statistically significant, with no major geographical inconsistencies” – I think you have interpreted the statistics incorrectly. The matched McNemar test does not test for concordance and I don’t see how these data are truly paired? Furthermore: “If the statistical significance level (i.e., p-value) is less than 0.05 (i.e., p < 0.05), you have a statistically significant result and the proportion of X before and after Y is statistically significantly different”, suggesting the opposite to concordance.

MR: see explanation in methods. We assessed concordance of presence of tick population between prediction and observation. We tested concordance in predicted area vs unpredicted area, for random distribution (chi-square test) ans level of concordance (kappa).

________________________________________

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: 20211215_PlosOne_Response to Reviewers.docx

Click here for additional data file.

18 Jan 2022

Current and future distribution of Ixodes scapularis ticks in Québec: field validation of a predictive model

PONE-D-21-14260R2

Dear Dr. RIPOCHE,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Catherine A. Brissette, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

25 Jan 2022

PONE-D-21-14260R2

Current and future distribution of Ixodes scapularis ticks in Québec: field validation of a predictive model

Dear Dr. Ripoche:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Catherine A. Brissette

Academic Editor

PLOS ONE