Email updates

Keep up to date with the latest news and content from Parasites & Vectors and BioMed Central.

Open Access Highly Accessed Research

Effects of global changes on the climatic niche of the tick Ixodes ricinus inferred by species distribution modelling

Daniele Porretta, Valentina Mastrantonio, Sara Amendolia, Stefano Gaiarsa, Sara Epis, Claudio Genchi, Claudio Bandi, Domenico Otranto and Sandra Urbanelli*

Parasites & Vectors 2013, 6:271  doi:10.1186/1756-3305-6-271

PubMed Commons is an experimental system of commenting on PubMed abstracts, introduced in October 2013. Comments are displayed on the abstract page, but during the initial closed pilot, only registered users can read or post comments. Any researcher who is listed as an author of an article indexed by PubMed is entitled to participate in the pilot. If you would like to participate and need an invitation, please email info@biomedcentral.com, giving the PubMed ID of an article on which you are an author. For more information, see the PubMed Commons FAQ.

Correlative models, ticks and climate trends

Sandra Urbanelli   (2013-11-18 11:18)  University of Rome "La Sapienza"










We thank Augustin Estrada-Pena [1] for his comments on our article [2]. One of the major constraints he raised in his letter pertained the differences between changes of climatic niche and in the areas of climatic suitability for Ixodes ricinus and, therefore, between map of climatic suitability and map of distribution and abundance of this tick species. Indeed, in our article [2] Species Distribution Modelling has been used to model areas of climatic suitability for this tick species and therefore for projecting them under two scenarios of climatic changes, in order to predict future changes in their distribution. We are aware of the differences between changes in climatic niche and in areas of climatic suitability, and between map of climatic suitability and those of distribution or abundance of a species. This is clear when we refer in the text to ‘areas of climatic suitability’ and, later on, when discussing the expansion of climatic niche (i.e., expansion of climatically suitable areas). Furthermore, most part of the discussion section is permeated by the caution needed to interpret our results under the context of possible changes in geographic distribution of I. ricinus. Indeed, the climatic features are only one of the ecological factors affecting the geographic distribution of tick species, since the presence of suitable hosts play an important role for the occurrence and dispersal of I. ricinus, and we also thoroughly discussed the limits of our model at smaller geographic scale [1]. Finally, the literature on the modelling approach we used has been also referred for providing the readers with the theoretical background and the meaning of the terminology used (i.e., climatic niche, species distribution and areas of climatic suitability). A criticism to the validity of our models [2] concerns the use of the AUC that, however, was not used for selecting our model, but the AIC index. Notably, the climatically suitable area predicted by our model under current conditions encompasses the known geographic distribution of I. ricinus and that previously inferred for this species using both climatic features and vegetation index by Estrada-Pena and colleagues [3].
In our paper we attempted to answer whether areas today unfavourable for I. ricinus, could become climatically suitable for this tick species in the future. Our results have clearly shown that under future climatic scenarios, the areas climatically suitable for I. ricinus will increase at continental geographic scale. Considering the intrinsic limits of each modelling approach, we concluded that the use of SDM models at a higher resolution should be integrated by a more refined analysis of further abiotic and biotic data.
We agree with Estrada-Pena in that ‘The field of modelling health affecting arthropods needs serious reflection about the conceptual basis and about the true meaning of the predictions, and that we should thus take care to not transmit the same and always alarming message.’  He is in line with the take home message of the article we published [2].

References

[1] http://www.parasitesandvectors.com/content/6/1/271/comments
[2] Porretta D., Mastrantonio V., Amendolia S., Gaiarsa S., Epis S., Genchi C., Bandi C., Otranto D. Urbanelli S. (2013). Effects of global changes on the climatic niche of the tick Ixodes ricinus inferred by species distribution modeling. Parasit Vectors, 6:271.  
[3] Estrada-Peña A, Ayllón N, de la Fuente J: Impact of climate trends on tick-borne pathogen transmission. Front Physiol 2012, 3:1-12.

Competing interests

The authors declare that they have no conflict of interest

top

Correlative models, ticks and climate trends: what we want to measure? A candid comment on Porretta et al. (2013)

Agustin Estrada-Pe��a   (2013-10-21 16:56)  Veterinary Faculty

Predicting the consequences of climate change for the distribution range and phenology of ticks and the severity of the pathogens they transmit remains a persistent challenge surrounded by much controversy. Studies are necessary to accurately capture the contributions of multiple, interacting, and often nonlinear responses of the tick to both the microclimate and the availability of hosts [1]. This has kindled interest in capturing the basic patterns of environmental features regulating the geographical ranges of ticks and their associated pathogens, using the so-called correlative models. I would like to comment some points of the paper by Porretta et al. [2] [hereinafter referred to as DP] which I consider might convey a biased concept about the probable impacts of the climate on the life cycle of the tick Ixodes ricinus. The approach by these authors assumes that the most important factors driving the distribution of the tick are related to the weather. Thus, its climate niche can be reconstructed by relating data on the occurrence of the tick with datasets summarising climate, topographic, edaphic, and other `abiotic�� or `ecological�� variables. I consider that this approach is a very valid one, provided that the explanatory variables are selected on ecological grounds, and that the modelling inference is adequately interpreted.

The capture of the climate niche is a strict estimation of the similarity of a series of weather variables at the sites in which the organism has been collected [3]. The projection of such inference on the geographical space produces a weighted similarity at each point of the variables evaluated compared with the values recorded at the sites where the tick has been reported. Such projection is not an estimate of either distribution or abundance, although it is incorrectly assumed to be a `risk map�� in which climate similarity is interpreted as a direct estimator of abundance. The main assumption of DP is that the climate niche of I. ricinus `changes�� as a consequence of the forecasted climate trends, which I believe is a misconception. The potential distribution of a species varies geographically with the oscillation of climatic conditions, but is environmentally invariant. Prevailing opinions state that the climate niche is the representation of the abiotic "preferences" of such organism [4-5]: it is static because niche conservatism seems to be a rule for most organisms [6-8]. The trends of climate could change the amount of territory that statistically match the distribution of values at which the organism has been recorded, but not the physiological response of the organisms [9].

A particularly important point in correlative modelling is the adequate choice of the explanatory variables. There is a tendency to make an automatic selection of the best covariates by the modelling algorithm, without further checking their ecological significance, a point already mentioned [10,11]. Such `blind selection�� procedures will construct statistically reliable models, but ecologically non-significant associations; these models will therefore produce unreliable results when projected onto different geographical or environmental backgrounds. For example, it is well known that rainfall is not an adequate estimator of the life cycle processes of ticks, which are primarily driven by the relative humidity and saturation deficit of the air [12-15]. Saturation deficit and relative humidity are not the same as rainfall, which, together with other variables related to the temperature, is a trait used by DP to outline the climate niche of I. ricinus. Patterns of precipitation undoubtedly have an effect on the relative humidity at the regional scale [16,17]. However, the effects are not the same in different biomes. There is no universal relationship between air humidity (or saturation deficit) and rainfall patterns. Ignoring this basic feature of the biology of ticks will probably produce poor correlations and flawed conclusions. The model is not more or less accurate if estimates of rainfall or water vapour are included. It is simply unreliable, because the tick has not an empirically demonstrated universal response to rainfall.

Capturing the climate niche of an organism includes the evaluation of the reliability of the selected modelling approach. This is commonly addressed by the use of the Area under the Receiver Operating Characteristic curve (AUC), a statistical measurement of the predictive accuracy of the model. AUC has been criticised [18] because the modeller is unaware of its variability in cases of serendipitous associations between the modelled organism and the descriptive variables. AUC is not an objective evaluation [19-21] and can produce a false estimation of good fitting by reasons unrelated to the descriptive variables. In example, a simple way to increase the values of AUC is to use a spatial background that is several times larger than the reported distribution of the species to be modelled. This always produces a high value of AUC, which gives a false sensation of adequate fitting of the model [22]. Studies have concluded that the geographical background on which project the inference of the climate niche must to strictly adhere to its reported distribution [23]. The choice of a large geographical background, like in the paper by DP, undoubtedly influenced the relatively high AUC values reported for I. ricinus.

Another potential source of inaccuracies in this paper is the selection of climate variables based on its availability and not on its reliability for the modelling purposes. Environmental datasets matter and a model is as good as is the set of explanatory variables. The easily available climate datasets based on interpolations from climate stations over large areas are being increasingly used in a variety of studies dealing with the climate niche and distributions [24]. However, the estimation of the effects of the internal coherence of the climate dataset on the potential errors of the model is rarely addressed. Series of data are commonly affected by autocorrelation [25] and by collinearity [26,27]. Both factors are known to strongly modify the apparent influence of variables that affect the distribution of an organism, resulting in the false perception of a well-fitted model because they inflate the AUC and therefore produce a bias of the model if residuals are not explicitly evaluated [28,29]. The point here is that internal issues of the explanatory variables have been not verified by DP, and it is assumed that values of AUC are bona fide representations of the reliability of the models. Without an evaluation of how the autocorrelation of the dataset affects the values of AUC [30] conclusions derived from the model will remain unreliable.

Probably as a consequence of the gaps I mentioned before, the results provided by DP show areas of matching climate for I. ricinus along regions of the Mediterranean basin, including steppe areas of Spain, Italy, or even the driest and almost desert areas of central Turkey, in northern parts of Scandinavia or in the mountain range of Norway. All these sites are known to not match the preferred range of abiotic conditions in which I. ricinus has been reported [31,32]. This is the point in correlative modelling: the modeller is blind to the inaccuracies introduced in the process because a high AUC value is not necessarily interpreted as an accurate projection of the sites where the organism has an adequate degree of climate similarity. Another point concerns to the reliability of the projections of a correlative model on scenarios of future climate. Such models are static and probabilistic in nature, since they statistically relate the geographical distribution of species to their present environment [22]. To state it in a simple way, a correlative model cannot be projected into another set of descriptive variables (either in the spatial or temporal domain), because descriptive variables are commonly correlated in different ways. The model cannot "respond" reliably to these new conditions, providing that new correlations among variables are not captured by additional terms of the model [33]. The specialists in the field have already alerted about the issue [11,34,35] but this has not been addressed in the literature on the topic. As a result, published record accumulates the use of predicted climate scenarios for estimating potential future distributions of health affecting arthropods without adequate evaluation of inaccuracies. This is not an issue when process-driven models are projected on scenarios of climate to evaluate areas of potential climate similarity [36-38] because these are lineal responses and not a complex relationships among variables used by modelling algorithms.

Other than the methodological gaps I briefly explained before, I would like to take a special point on the conclusions outlined in this paper. There is no way to conclude about the phenology of a species of tick if only a measure of the climate similarity is addressed by a correlative model. My modest opinion is that conclusions by DP are inadequate, because both the lack of reliability of the models built around the described procedural omissions, and because the lack of supporting results. A good use of species distribution models requires a clear distinction of the differences between potential and realized distributions [20]. I think that as the availability of distribution data and the pressure to obtain results increase, so does the danger of developing distribution models without a solid conceptual background. The field of modelling health affecting arthropods needs serious reflection about the conceptual basis and about the true meaning of the predictions. We simply ignore how the basic tick��s life cycle could become affected by the trends of climate, and how the transmission of pathogens will be altered because we lack the fundamental factor of the equation: the hosts [10]. The complexity and plasticity of the tick��s life cycle makes a linear response to climate change unlikely. We should thus take care to not transmit the same and always alarming message, because such simple message linking climate and ticks is not yet possible. Selecting the explanatory variables on ecological grounds, and abandoning the idea that a statistical index will drive our conclusions would, anyway, improve our knowledge on the inherent plasticity of the tick's life cycle. References.

1. Eisen L, Eisen RJ: Using geographic information systems and decision support systems for the prediction, prevention, and control of vector-borne diseases. Ann Rev Entomol 2011, 56:41-61.
2. Porretta D, Mastrantonio V, Amendolia S, Gaiarsa S, Epis S, Genchi C, Bando C, Otranto D, Urbanelli S: Effects of global changes on the climatic niche of the tick Ixodes ricinus inferred by species distribution modelling. Parasit Vectors 2013, 6:271.
3. Kearney M, Porter W: Mechanistic niche modelling: combining physiological and spatial data to predict species�� ranges. Ecology letters 2009, 12:334-350.
4. Colwell RK, Rangel TF: Hutchinson's duality: the once and future niche. Proc Nat Acad Sci 2009, 106:19651-19658.
5. Lavergne S, Mouquet N, Thuiller W, Ronce O: Biodiversity and climate change: integrating evolutionary and ecological responses of species and communities. Ann Rev Ecol Evol Syst 2010, 41:321-350.
6. Peterson AT, Sobern J, Sanchez-Cordero V: Conservatism of ecological niches in evolutionary time. Science 1999, 285:1265-1267.
7. Losos JB: Phylogenetic niche conservatism, phylogenetic signal and the relationship between phylogenetic relatedness and ecological similarity among species. Ecology letters 2008, 11:995-1003.
8. Wiens JJ: Speciation and ecology revisited: phylogenetic niche conservatism and the origin of species. Evolution 2004, 58:193-197.
9. Kearney M: Habitat, environment and niche: what are we modelling?. Oikos 2006, 115:186-191.
10. Randolph SE: Predicting the risk of tick-borne diseases. Int J Med Microbiol 2002, 291:6-10.
11. Heikkinen RK, Luoto M, Arajo MB, Virkkala R, Thuiller W, Sykes M: Methods and uncertainties in bioclimatic envelope modelling under climate change. Progress in Physical Geography 2006, 30:751-777.
12. Perret JL, Guigoz E, Rais O, Gern L: Influence of saturation deficit and temperature on Ixodes ricinus tick questing activity in a Lyme borreliosis-endemic area (Switzerland). Parasitol Res 2000, 86:554-557.
13. Randolph SE, Green RM, Hoodless AN, Peacey MF: An empirical quantitative framework for the seasonal population dynamics of the tick Ixodes ricinus. Int J Parasitol 2002, 32:979-989.
14. Kahl O, Knolle W: Water vapour uptake from subsaturated atmospheres by engorged immature ixodid ticks. Exp Appl Acarol 1988, 4:73-83.
15. Kahl O, Alidousti I: Bodies of liquid water as a source of water gain for Ixodes ricinus ticks (Acari: Ixodidae). Exp Appl Acarol 1997, 21:731-746.
16. Rao VB, Cavalcanti IFA, Hada K: Annual variation of rainfall over Brazil and water vapor characteristics over South America. J Geophys Res Atmos 1996, 101:26539-26551.
17. Thornton PE, Hasenauer H, White: Simultaneous estimation of daily solar radiation and humidity from observed temperature and precipitation: an application over complex terrain in Austria. Agric Meteorol 2000, 104:255-271.
18. Lobo JM, Jimenez-Valverde A, Real R: AUC: A misleading measure of the performance of predictive distribution models. Glob Ecol Biogeog 2008, 17:145-151.
19. Elith J, Graham CH: Do they? How do they? WHY do they differ? On finding reasons for differing performances of species distribution models. Ecography 2009, 32:66-77.
20. Jimenez-Valverde A, Lobo JM, Hortal J: Not as good as they seem: the importance of concepts in species distribution modelling. Diversity and Distributions 2008, 14:885-890.
21. Jimenez-Valverde A: Insights into the area under the receiver operating characteristic curve (AUC) as a discrimination measure in species distribution modelling. Glob Ecol Biogeog 2012, 21:498-507.
22. Jimenez-Valverde A, Acevedo P, Barbosa AM, Lobo J, Real R:. Discrimination capacity in species distribution models depends on the representativeness of the environmental domain. Glob Ecol Biogeog 2013, 22:508-516.
23. Acevedo P, Jimenez-Valverde A, Lobo J, Real R: Delimiting the geographical background in species distribution modelling. J Biogeog 2012, 39:1383-1390.
24. Peterson AT, Nakazawa Y: Environmental data sets matter in ecological niche modelling: an example with Solenopsis invicta and Solenopsis richteri. Glob Ecol Biogeog 2008, 17:135-144.
25. Legendre P: Spatial autocorrelation: trouble or new paradigm?. Ecology 1993, 74:1659-1673.
26. Mackinnon MJ, Puterman ML: Collinearity in generalized linear models. Comm Statistics-theory and Meth 1989, 18:3463-3472.
27. Dormann CF, Elith J, Bacher S, Buchmann C, Carl G, Carre G, Lautenbach S: Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography 2012 36:027-046. doi: 10.1111/j.1600-0587.2012.07348.x.
28. Dormann CF, McPherson JM, Araujo MB, Bivand R, Bolliger J: Methods to account for spatial autocorrelation in the analysis of species distributional data: a review. Ecography 2007, 30:609-628.
29. Hawkins BA: Eight (and a half) deadly sins of spatial analysis. J Biogeogr 2012, 39:1-9.
30. Estrada-Pena A, Estrada-Sanchez A, Estrada-Sanchez D, de la Fuente J: Assessing the effects of variables and background selection on the capture of the tick climate niche. Int J Hlth Geog 2013, 12:1-13.
31. Estrada-Pena A, Farkas R, Jaenson TG, Koenen F, Madder M, Pascucci I, Jongejan F: Association of environmental traits with the geographic ranges of ticks (Acari: Ixodidae) of medical and veterinary importance in the western Palearctic. A digital data set. Exp Appl Acarol 2013, 59: 351-366.
32. Medlock JM, Hansford KM, Bormane A, Derdakova M, Estrada-Pena A, George JC, Van Bortel W: Driving forces for changes in geographical distribution of Ixodes ricinus ticks in Europe. Parasit Vectors 2013, 6:1-11.
33. Araujo MB, Guisan A: Five (or so) challenges for species distribution modelling. J Biogeog 2006, 33:1677-1688.
34. Aarujo MB, Luoto M: The importance of biotic interactions for modelling species distributions under climate change. Global Ecol Biogeog 2007, 16:743-753.
35. Pearson RG, Thuiller W, Araujo MB, Martinez-Meyer E, Brotons L, McClean C, Lees DC: Model-based uncertainty in species range prediction. J Biogeog 2006, 33:1704-1711.
36. Ogden NH, Maarouf A, Barker IK, Bigras-Poulin M, Lindsay LR, Morshed MG, Charron DF: Climate change and the potential for range expansion of the Lyme disease vector Ixodes scapularis in Canada. Int J Parasitol 2006, 36:63-70.
37. Dobson ADM, Rrandolph SE: Modelling the effects of recent changes in climate, host density and acaricide treatments on population dynamics of Ixodes ricinus in the UK. J Appl Ecol 2011, 48:1029-1037.
38. Estrada-Pena A, Estrada-Sanchez D: Deconstructing Ixodes ricinus: a partial matrix model allowing mapping of tick development, mortality and activity rates. Med Vet Entomol 2013, n/a. doi: 10.1111/mve.12009.

Competing interests

The author declares that the text before was prepared in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

top

Post a comment