Tick paralysis is a common tick-borne illness in humans and animals throughout the world, caused by neurotoxins produced in the salivary glands of ticks and secreted into a host during the course of feeding by females and immature stages19. Fifty-nine ixodid and fourteen argasid ticks are currently believed to be involved in the transmission of tick paralysis worldwide19, 20. In Australia, I. holocyclus is considered to be the leading tick species implicated in the transmission of tick paralysis primarily in dogs, but also other species, viz. cats, sheep, cattle, goats, swine and horses. Humans are also occasionally affected, and the disease can be fatal2, 21. A second tick species, I. cornuatus has also been implicated in the transmission of tick paralysis in Australia; however, it is also considered a minor player in this disease22. Given the differences in their biology, distribution, and natural history of these two species, we focused on estimating the spatial distribution of I. holocyclus in the present study. We recognize, however, that it is important to consider the distributions of both species for proper epidemiological planning and management of tick paralysis in Australia.
Ecological niche modeling is a well-tested approach for estimating species distributions based on abiotic factors13, 23. Several new recommendations have been made in recent years for proper construction of niche models; such as the appropriate thinning of occurrence data24, consideration of an accessible area for a species being studied (M)25, thorough exploration of model complexity26, 27, and use of multiple statistical criteria for model selection28, 29. We carefully considered all these recommendations to produce a robust spatial distribution model for I. holocyclus. The resulting replicated models were fairly consistent in predicting suitability for I. holocyclus, as indicated by moderate range estimates (Fig. 2B). Further, the MOP analysis indicated satisfactory performance of the present-day model with extrapolation only in small areas outside the predicted suitable areas. These qualities, along with the model’s very low omission rate (0.044%) gives high confidence in the predicted suitable area for this species in Australia. It will be essential, however, to confirm the actual presence of I. holocyclus outside the traditionally known areas through acarological surveys to assess our findings.
The present-day spatial distribution predicted in this study (Fig. 2A) indicates that the geographic areas suitable for I. holocyclus match the currently known distribution of this species along the eastern seaboard, but the suitability also extends through most of the coastal areas in the south, and up to the Kimbolton Peninsula in Western Australia in the north. Highly suitable areas are present around and south of Perth, extending towards Albany, Western Australia. Most areas in Tasmania are also highly suitable for this species. The current distribution in the Eastern Seaboard may be wider than the traditionally known extents in some areas compared to Roberts30. It is likely that I. holocyclus will succeed in establishing permanent populations if introduced into areas that are currently free of them along the southern and northern coasts, and along the southwestern coast of Western Australia and Tasmania. Appropriate prevention of tick movement including pet inspections and quarantine will be necessary to avoid introductions.
Future potential distribution of I. holocyclus in year 2050 based on both low- and high-emissions scenarios indicate moderate increases in climatic suitability from the present-day prediction (Fig. 4A,B); but noticeably also moderate to low loss of climatically suitable areas in 2050. This loss could be at least partly attributed to potential future temperature and precipitation conditions exceeding suitable ranges for these ticks in these areas, limiting their ability to survive. Predicted loss of suitable areas in future can also be observed to be irregular, and in some areas, particularly along northern Queensland and in Northern Territory, enveloped between stretches of suitable areas. Our use of relatively coarse resolution data (1 km2) limits our ability to thoroughly interpret such phenomenon, but this is likely due to variations in the geography in these areas that respond differently to future climate, as well as the potential increase in ocean temperature and subsequent influences on areas along the coast that may render them unsuitable for this species. Despite the noticeable loss in climatically suitable areas, likely no net loss in area will accrue for this species by 2050.
Teo et al.31 assessed present and future potential distribution for I. holocyclus using both CLIMEX32, 33 and a novel, as-yet unpublished “climatic-range” approach to determine the suitability on monthly intervals. CLIMEX allows users to specify different upper and lower thresholds for climatic parameters, some of which were derived for their study from laboratory evaluations of I. holocyclus34. The present-day distribution reported in that study resembles our results in identification of a relatively narrow area along the East Coast as suitable; however, much of the northern and northeastern areas along the coast, the coasts of South Australia and southwestern Australia, and Tasmania are reported unsuitable. Their future predictions (2050) of the species’ potential distribution were based on two GCMs (CSIRO MK3 and MIROC-H) climate models, were also markedly different from our predictions, anticipating rather dramatic distributional loss for the species. Such model transfers are challenging, with many factors potentially producing inconsistencies35. However, the two studies reflect two fundamentally different classes of ecological niche models; CLIMEX is deterministic, whose predictions are largely constrained by user supplied threshold values for model inputs of physiological tolerance limits of a species33, whereas Maxent is a machine-learning correlative approach, in which known occurrences of a species is used in conjunction with environmental layers to determine conditions that meet a species’ environmental requirements, and therefore the suitability of geographic spaces. Although the former (CLIMEX) approach is appealing conceptually, scaling environmental dimensions between the micro-scales of physiological measurements and the macro-scales of geography is well-known to present practical and conceptual challenges36.
Different ixodid ticks employ different life-history strategies in response to adverse environmental conditions, including behavioral adaptations, active uptake of atmospheric moisture, restriction of water-loss, and tolerance towards extreme temperatures37. Precisely which of these mechanisms I. holocyclus utilizes, if any at all, for its survival during diverse temperature and humidity conditions is not clearly known, but it is likely to involve multiple mechanisms. In this sense, the threshold values used by Teo et al.31, based purely on laboratory observations may have been overly restrictive, leading to a conservative distributional estimate for this species. Further, because relationships between abiotic variables and species’ occurrences are fairly complex and highly dimensional, a physiological thresholding approach wherein values are set independently for different abiotic parameters may not capture species’ relationships with environments adequately. The correlative approaches employed in the present study are data-driven, and as such may capture more of this complexity, with fewer problems of scaling across orders of magnitude of space and time.
In conclusion, ticks are poikilothermic ectoparasites, whose survival, reproduction and other biological functions are regulated by ambient climatic conditions. Although ixodid ticks are known to regulate their body temperatures by moving about their habitat (vegetation), attempts to model their spatial distribution has resulted in models largely based on climate variables. Nevertheless, other factors such as host availability play a significant role in tick distribution, which unfortunately cannot be readily included in correlative ecological niche models largely because such data are rarely available. These suitability predictions, in addition to being entirely based on large-scale climate, also do not reveal the highly likely heterogeneity in abundance or density in different geographic areas within the realized climatically suitable areas. For these reasons, the distribution maps produced in this study must be used with some caution, and perhaps as a guide to target sampling and not as a substitute for thorough acarological surveys.
Source: Ecology - nature.com
