Background: Some tick species are invasive and of high consequence to public and veterinary health. Socioeconomic development of rural parts of the USA was enabled partly through the eradication by 1943 of cattle fever ticks (CFT, Rhipicephalus (Boophilus) annulatus and R. (B.) microplus). The southern cattle fever ticks (SCFT, R. (B.) microplus) remain a real and present threat to the USA animal agriculture because they are established in Mexico. Livestock-wildlife interactions in the Permanent Quarantine Zone (PQZ) established by the century-old Cattle Fever Tick Eradication Programme (CFTEP) in south Texas endanger its operations. Methods: We describe a spatially-explicit, individual-based model that simulates interactions between cattle, white-tailed deer (WTD, Odocoileus virginianus), and nilgai (Boselaphus tragocamelus) to assess the risk for SCFT infestations across the pathogenic landscape in the PQZ and beyond. We also investigate the potential role of nilgai in sustaining SCFT populations by simulating various hypothetical infestation and eradication scenarios. Results: All infestation scenarios resulted in a phase transition from a relatively small proportion of the ranch infested to almost the entire ranch infested coinciding with the typical period of autumn increases in off-host tick larvae. Results of eradication scenarios suggest that elimination of all on-host ticks on cattle, WTD, or nilgai would have virtually no effect on the proportion of the ranch infested or on the proportions of different tick habitat types infested; the entire ranch would remain infested. If all on-host ticks were eliminated on cattle and WTD, WTD and nilgai, or cattle and nilgai, the proportions of the ranch infested occasionally would drop to 0.6, 0.6 and 0.2, respectively. Differences in proportions of the ranch infested from year to year were due to primarily to differences in winter weather conditions, whereas infestation differences among tick habitat types were due primarily to habitat use preferences of hosts. Conclusions: Infestations in nilgai augment SCFT refugia enabled by WTD and promote pest persistence across the landscape and cattle parasitism. Our study documented the utility of enhanced biosurveillance using simulation tools to mitigate risk and enhance operations of area-wide tick management programmes like the CFTEP through integrated tactics for SCFT suppression.
Keywords: Cattle Fever Tick Eradication Program; Host-parasite interaction; Individual-based model; Spatially-explicit model; Stochastic; Integrated tick management research; Rhipicephalus microplus
Graph
Cattle fever ticks (CFT), Rhipicephalus (Boophilus) annulatus and R. (B.) microplus, pose a significant threat to the economic security of the USA cattle industry as vectors of Babesia bigemina and B. bovis, which cause bovine babesiosis, and Anaplasma marginale that causes anaplasmosis [[
Historically, success of the CFTEP depended primarily on the host specificity of CFT [[
The presence of alternate or secondary host species in the Texas-Mexico border region and beyond poses significant challenges to the CFTEP [[
Recent field studies have incriminated nilgai, Boselaphus tragocamelus, as an alternate host species for SCFT in the Texas-Mexico transboundary border region, which poses a risk for the re-emergence of CFT-borne diseases in the USA [[
Continued success of the CFTEP requires an integrated strategy based on an interdisciplinary systems approach, which specifically includes consideration of management risks and opportunities associated with the livestock-wildlife interface [[
In this study, the model developed by Wang et al. [[
Graph: Fig. 1 Conceptual model representing the potential role of nilgai (Boselaphus tragocamelus) in the maintenance of southern cattle fever tick (Rhipicephalus (Boophilus) microplus) populations in south Texas
The model of Wang et al. [[
The experimental design for our sensitivity analysis consisted of three sets of scenarios in which we altered (i) the combination of host species present, (ii) the relative habitat preferences of the host species, and (iii) the relative densities of the host species. For each scenario, we ran 10 replicate stochastic (Monte Carlo) simulations. During each simulation, we monitored mean densities of off-host (potentially host-seeking) tick larvae in each of the habitats used by hosts, as well as mean numbers of adult ticks on each host species present.
Scenarios involving different combinations of host species included (i) cattle, WTD, and nilgai (baseline scenario), (ii) only cattle, (iii) only WTD, and (iv) only nilgai. Scenarios involving different relative habitat use preferences of hosts included (i) baseline preferences of all host species (Additional file 1: Table S1), as well as scenarios in which all host species used either (ii) good, (iii) fair, or (iv) poor tick habitat. Scenarios involving different densities of hosts included (i) baseline densities of all host species (Additional file 1: Table S1), as well as scenarios in which baseline densities of all host species were reduced simultaneously to (ii) 1/2, (iii) 1/4, (iv) 1/8, (v) 1/16, (vi) 1/32, (vii) 1/64, (viii) 1/128, and (ix) 1/256 of their baseline levels. We also ran scenarios in which the baseline density of each host species was reduced sequentially as just described, assuming only that single species was present.
The experimental design for evaluating our proof of concept consisted of two approaches. The first was to evaluate three scenarios for initiating a tick infestation of the landscape and assessing the spatial and temporal rates of spread. The second approach was to evaluate three scenarios of SCFT eradication in which acaricide-treated hosts removed ticks from the system. For each scenario, we again ran 10 replicate stochastic (Monte Carlo) simulations. During each simulation, we monitored the number and spatial distribution of 1-ha landscape patches infested with off-host tick larvae. For the infestation scenarios, we infested the hypothetical 10,000-ha ranch by introducing one infested animal of (i) cattle, (ii) WTD, or (iii) nilgai during week 25 of 2009. We assumed the ranch was SCFT-free at the time of this infestation, and contained baseline densities of each of the three host species (Additional file 1: Table S1). Each of the three sets of scenarios involved introducing the infested host with its activity area (Additional file 1: Table S1) centered in a large patch of (i) good, (ii) fair, or (iii) poor habitat for off-host tick larvae, and also (iv) centered at the intersection of these three habitat types. For the eradication scenarios, we simulated the impact of a single acaricide treatment capable of complete and continuous elimination of all SCFT that attach to (i) cattle, (ii) WTD, (iii) nilgai, (iv) cattle and WTD, (v) WTD and nilgai, and (vi) cattle and nilgai. That is, all three host species continued to collect host-seeking larvae from the landscape, but all on-host larvae were eliminated immediately from the host species to which the acaricide was applied. The single acaricide application was applied during week 25 of 2009. Thus, in week 25 the role of individuals of the treated host species was changed from tick dispersal to tick removal.
General trends in mean densities of off-host tick larvae and mean numbers of adult ticks on hosts resulting from different (i) combinations of host species (Additional file 1: Figure S3), (ii) relative habitat preferences of host species (Additional file 1: Figure S4), and (iii) relative densities of the host species (Additional file 1: Figure S5) all might be summarized in the following manner. Seasonal fluctuations in mean densities of off-host tick larvae followed a bimodal pattern, with a spring increase and summer decline, followed by an autumn increase and winter decline. Increasing densities were correlated with periods of precipitation, low saturation deficits, and moderate temperatures favorable for larval survival during spring and autumn. Decreasing densities were correlated with unfavorable periods of high temperatures and saturation deficits during summer, and with unfavorable periods of low temperatures during winter, which sometimes are worsened by low precipitation and relatively higher saturation deficits. Mean numbers of adult ticks on all host species remained relatively high and constant during spring, summer and autumn (e.g. ≈ 50, ≈ 4.5 and ≈ 25 for cattle, WTD, and nilgai, respectively, during the baseline scenario), but declined to relatively low and more variable levels during winter (e.g. < 20, < 2 and < 10 for cattle, WTD, and nilgai, respectively, during the baseline scenario). Winter declines in on-host ticks were correlated primarily with off-host tick responses to low environmental temperatures, which retarded oviposition, and prolonged egg incubation, larval emergence, and host-seeking activity. The specific amplitudes and timing of seasonal fluctuations in abundances of both on- and off-host ticks varied from year-to-year in response to specific combinations of these weather parameters, and also were affected by specific characteristics of the community of host species present, which we point out below.
Alteration of the combinations of host species present affected off-host tick larval densities noticeably, but effects on numbers of adult ticks per host were negligible. Mean larval densities were highest with all three host species present, slightly lower when only cattle were present, and much lower when only nilgai or only WTD were present (Additional file 1: Figures S3a and S5). These relative differences did not vary among the different tick habitat types (Additional file 1: Figure S6). Mean numbers of adult ticks per host were the same as those described in the previous paragraph, regardless of alterations of the host species present (Additional file 1: Figure S7b, c and d).
Alteration of the relative habitat preferences of host species also affected off-host tick larval densities noticeably, but, again, effects on numbers of adult ticks per host were negligible. Mean larval densities were highest when all three host species used the good tick habitat exclusively, with densities decreasing slightly with exclusive use of fair tick habitat, and more noticeably with exclusive use of poor tick habitat (Additional file 1: Figure S4). Mean numbers of adult ticks per host were the same as those described above, regardless of alterations of the relative habitat preferences of host species (Additional file 1: Figure S4b, c and d).
Alteration of the relative densities of all host species simultaneously affected both off-host tick larval densities and numbers of adult ticks per host noticeably. Seasonality of fluctuations and relative variations from year-to-year were not affected. However, overall, mean larval density levels decreased proportionally with decreases in host densities (Additional file 1: Figure S8a), regardless of tick habitat type (Additional file 1: Figure S7). Mean levels of numbers of adult ticks per host decreased at generally increasing rates with decreases in host densities beyond host-specific threshold levels (Additional file 1: Figure S8b, c and d). Alteration of the relative host densities when only a single host species was present affected both off-host tick larval densities and numbers of adult ticks per host in a similar manner as when all host species were present. Once again, seasonality of fluctuations and relative variations from year-to-year were not affected. However, mean larval density levels decreased proportionally with decreases in densities of the single host present (Additional file 1: Figure S9), regardless of habitat type. Mean levels of numbers of adult ticks per host also decreased at generally increasing rates with decreases in densities of the single host present, beyond host-specific threshold levels (Additional file 1: Figure S10). Minimum densities of cattle, WTD, and nilgai capable of sustaining tick populations in the absence of the other two host species were 1, 26 and 8 individuals per hectare, respectively.
The introduction of one infested head of cattle or nilgai into an otherwise tick-free landscape during mid-summer (week 25) resulted in the infestation spreading to virtually all landscape patches by late autumn (week 50) (Fig. 2a, c). Infestations introduced by one infested WTD spread more slowly (Fig. 2b) but reached virtually all landscape patches by mid-spring (week 8) of the following year. The slower rate of spread was due primarily to the relatively lower number of ticks carried by WTD, as well as the relatively lower fecundity of engorged female ticks that feed on WTD [[
Graph: Fig. 2 Assessment of initiation of infestation. Simulated mean proportions of landscape cells infested with off-host (potentially host-seeking) tick larvae per hectare on a hypothetical 10,000-ha ranch under weather conditions recorded in Willacy County, Texas, USA, from January 2009 through December 2018 (only week 24, 2009 through week 2, 2010 shown here): (i) on whole ranch (grey dash line), and in (ii) good (black line), (iii) fair (black dash line), and (iv) poor (grey line) tick habitats. Simulations assumed one infested head of cattle (a), white-tailed deer (b) and nilgai (c) was introduced at the intersection of patches of good, fair, and poor tick habitat in an otherwise SCFT-free ranch during week 25 of 2009. Tick hosts present on the ranch included cattle, white-tailed deer and nilgai. Thirty-one percent, 28 percent and 41 percent of the ranch was considered good, fair and poor habitat, respectively, for off-host tick larvae. Relative habitat use preferences of hosts for good, fair and poor tick habitats, respectively, were 0.30, 0.10 and 0.60 for cattle, 0.20, 0.40 and 0.40 for white-tailed deer, and 0.30, 0.10 and 0.60 for nilgai
The elimination of all on-host ticks on either foraging/grazing (i) cattle, or (ii) WTD, or (iii) nilgai had virtually no effect on the proportions of infested landscape patches in good, fair, and poor tick habitat types (Fig. 3). For all practical purposes, the entire ranch remained completely infested. The elimination of all on-host ticks on (iv) cattle and WTD, (v) WTD and nilgai, or (vi) cattle and nilgai had more noticeable effects (Fig. 4). Differences in proportions of infested landscape patches from year to year were due to primarily to differences in winter weather conditions, whereas infestation differences among tick habitat types were due primarily to habitat use preferences of hosts. The proportions of infested landscape patches occasionally dropped as low as approximately 0.6 when all on-host ticks were eliminated from cattle and WTD, or from WTD and nilgai. When all on-host ticks were eliminated from cattle and nilgai, the proportions of infested landscape patches periodically dropped below 0.4, occasionally dropping as low as 0.2, and infestation differences among tick habitat types were altered (Fig. 4). With WTD being the only distributor of ticks across the landscape, and cattle and nilgai acting to eliminate host-seeking larvae, the habitat use patterns of the hosts in effect created a refuge for off-host ticks in the fair tick habitat, particularly during periods of unfavorable conditions for off-host tick survival (Figs. 5, 6). An expanded time series of maps illustrating the appearance and disappearance of these refuges can be found in Additional file 1: Figure S15.
Graph: Fig. 3 Assessment of treated hosts by single species. Simulated mean proportions of landscape cells infested with off-host (potentially host-seeking) tick larvae per hectare on a hypothetical 10,000-ha ranch under weather conditions recorded in Willacy County, Texas, USA from January 2009 through December 2018: (i) on whole ranch (grey dash line), and in (ii) good (black line), (iii) fair (black dash line), and (iv) poor (grey line) tick habitats. Simulations assumed complete and continuous elimination of all on-host ticks on cattle (a), white-tailed deer (b) and nilgai (c) beginning week 25 of 2009. Tick hosts present on the ranch included cattle, white-tailed deer and nilgai. Thirty-one percent, 28 percent and 41 percent of the ranch was considered good, fair and poor habitat, respectively, for off-host tick larvae. Relative habitat use preferences of hosts for good, fair and poor tick habitats, respectively, were 0.30, 0.10 and 0.60 for cattle, 0.20, 0.40 and 0.40 for white-tailed deer, and 0.30, 0.10 and 0.60 for nilgai
Graph: Fig. 4 Assessment of treated hosts by paired species. Simulated mean proportions of landscape cells infested with off-host (potentially host-seeking) tick larvae per hectare on a hypothetical 10,000-ha ranch under weather conditions recorded in Willacy County, Texas, USA from January 2009 through December 2018: (i) on whole ranch (grey dash line), and in (ii) good (black line), (iii) fair (black dash line), and (iv) poor (grey line) tick habitats. Simulations assumed complete and continuous elimination of all on-host ticks on cattle and white-tailed deer (a), white-tailed deer and nilgai (b) and cattle and nilgai (c) beginning week 25 of 2009. Tick hosts present on the ranch included cattle, white-tailed deer and nilgai. Thirty-one percent, 28 percent and 41 percent of the ranch was considered good, fair and poor habitat, respectively, for off-host tick larvae. Relative habitat use preferences of hosts for good, fair and poor tick habitats, respectively, were 0.30, 0.10 and 0.60 for cattle, 0.20, 0.40 and 0.40 for white-tailed deer, and 0.30, 0.10 and 0.60 for nilgai
Graph: Fig. 5 Assessment of habitat infestations when cattle and nilgai are treated for tick elimination. Simulated mean proportions of landscape cells infested with off-host (potentially host-seeking) tick larvae per hectare on a hypothetical 10,000-ha ranch during the indicated weeks of 2014: (i) on whole ranch, and in (ii) good, (iii) fair, and (iv) poor tick habitats. Simulations were run under weather conditions recorded in Willacy County, Texas, USA, from January 2009 through December 2018, and assumed complete and continuous elimination of all on-host ticks on cattle and nilgai beginning week 25 of 2009. Tick hosts present on the ranch included cattle, white-tailed deer and nilgai. Thirty-one percent, 28 percent and 41 percent of the ranch was considered good, fair and poor habitat, respectively, for off-host tick larvae. Relative habitat use preferences of hosts for good, fair and poor tick habitats, respectively, were 0.30, 0.10 and 0.60 for cattle, 0.20, 0.40 and 0.40 for white-tailed deer, and 0.30, 0.10 and 0.60 for nilgai
Graph: Fig. 6 Time series of maps illustrating spatial dynamics of a tick infestation within the hypothetical 10,000-ha ranch containing good (green), fair (red), and poor (blue) tick habitat types. Acaricide applications capable of complete and continuous elimination of all on-host ticks applied to cattle and nilgai (but not white-tailed deer) were initiated during week 25 of 2009. Yellow represents infested landscape cells
Over the last four decades, ungulate wildlife such as WTD and nilgai have adversely affected success of the CFTEP [[
Our evaluation and confirmation of the present model [[
Regarding ability of the model to produce useful output, results of baseline simulations seem reasonable and are irrefutable based on available field data and observations, results of infestation simulations are interpretable ecologically, and results of eradication simulations have clear management implications with regard to the potential role of nilgai in sustaining SCFT populations. Not surprisingly, published data on abundances of both off- and on-host life stages of SCFT in south Texas rangelands are rare, in part due to quarantine regulations, difficulty of sampling the environment for off-host larval SCFT, and intricacies of sampling SCFT on wildlife hosts. Thus, assessment of reasonableness of quantitative aspects of simulation results must be guided primarily by comparative reasoning. For example, baseline simulations (i.e. with baseline densities and relative habitat preferences of cattle, WDT and nilgai; Additional file 1: Table S1) generate annual variations in abundance of host-seeking larvae (following the bimodal pattern described in the previous paragraph) which are similar to those observed in New Caledonia [[
Within a management context, eradication simulations suggest that nilgai do, indeed, have the potential to sustain SCFT populations, even in the absence of other hosts. It appears the role of nilgai in sustaining SCFT infestations may be complimentary, rather than strictly analogous, to the role of WTD. Overlapping habitat use (tick deposition) patterns of these hosts may create refugia for SCFT during periods of unfavorable conditions for off-host tick survival, and facilitate local spread from these refugia during more favorable periods. Nilgai habitat use patterns may greatly facilitate the widespread redistribution and maintenance of SCFT during more favorable periods. Activity ranges of nilgai are almost an order of magnitude greater than those of WTD, with the maximum axis of the home ranges of radio-tracked individuals exceeding 30 km [[
Modeling efforts, such as the present one, in coordination with field studies on interactions among cattle, WDT, nilgai and other potential host species, are crucial for development of integrated strategies for sustainable SCFT eradication in the USA [[
Infestations in nilgai augment SCFT refugia enabled by WTD and promote pest persistence across the landscape and cattle parasitism. Our study demonstrated the utility of enhanced biosurveillance using simulation tools to mitigate risk and enhance operations of area-wide tick management programmes like the CFTEP through integrated tactics for SCFT suppression.
This publication reports the research outcomes of USDA-ARS project 3094-32000-039-93S titled "Nilgai Modeling" that was funded by FY-2018 USDA-APHIS Cattle Fever Tick Eradication Program (CFTEP). The research of AAPDL was supported through the appropriated project Cattle Fever Tick Control and Eradication (3094-32000-039-00-D). The USDA is an equal opportunity employer and provider.
We would like to thank the anonymous reviewer, Section Editor, Dr Emanuele Brianti and Editor-in-Chief, Dr Aneta Kostadinova for their time and effort. The manuscript is greatly improved as a result of their comments.
PDT, KHL and AAPDL conceived of the study. H-HW and WEG developed the simulation models. H-HW conducted statistical analyses. H-HW and WEG led the writing. PDT, KHL and AAPDL edited the manuscript. All authors read and approved the final manuscript.
The simulated data during and/or analyzed during the present study are available from the corresponding author upon reasonable request.
Not applicable.
Not applicable.
The authors declare that they have no competing interests.
Graph: Additional file 1: Figure S1. Map of historical infestations of the southern cattle fever tick. Figure S2. Weather profiles. Figure S3. Assessment of host contribution. Figure S4. Assessment of habitat usage. Figure S5. Simulated mean numbers of off-host tick larvae per hectare from January 2009 through December 2018. Figure S6. Assessment of host contribution in three habitats from January 2014 through December 2015. Figure S7. Assessment of host density in three habitats in December 2014. Figure S8. Assessment of host density on numbers of off-host ticks and adult ticks on hosts in December 2014. Figure S9. Assessment of single host contribution on numbers of off-host ticks in December 2014. Figure S10. Assessment of single host contribution on numbers of adult ticks on hosts in December 2014. Figure S11. Assessment of contribution of single infested host introduced in the middle of a patch of fair tick habitat from June 2009 through January 2010. Figure S12. Assessment of contribution of single infested host introduced in the middle of a patch of good tick habitat from June 2009 through January 2010. Figure S13. Assessment of contribution of single infested host introduced in the middle of a patch of poor tick habitat from June 2009 through January 2010. Figure S14. Time series of maps illustrating spatial spread of a tick infestation with one infested head of cattle introduced in June 2009. Figure S15. Time series of maps illustrating spatial dynamics of a tick infestation with acaricide applications initiated in June 2009. Table S1. List of the parameters used to represent nilgai, cattle, and white-tailed deer as hosts of cattle fever ticks, their baseline values, and their information sources.
• CFT
- Cattle fever tick
• SCFT
- Southern cattle fever tick
• PQZ
- Permanent Quarantine Zone
• WTD
- White-tailed deer
• PCR
- Polymerase chain reaction
• USDA
- United States Department of Agriculture
• APHIS
- Animal Plant Health and Inspection Service
Supplementary information accompanies this paper at 10.1186/s13071-020-04366-x.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
By Hsiao-Hsuan Wang; William E. Grant; Pete D. Teel; Kimberly H. Lohmeyer and Adalberto A. Pérez de León
Reported by Author; Author; Author; Author; Author