Three Maps Hiding Behind One Risk Color
The blacklegged tick map becomes easier to interpret once its colors are separated into the questions they can actually answer. This chapter compares the Ohio study’s occurrence and relative-abundance estimates, then adds the pathogen evidence from Biltmore Forest in western North Carolina. The practical judgment is simple: a place can be favorable for encountering ticks without having the same abundance, pathogen prevalence, or human disease risk that a single “high-risk” color might imply.
That distinction is not a technical nicety. Public-health decisions depend on the denominator and the outcome. A surveillance crew deciding where to search needs a different estimate from a laboratory testing a tick, and both differ from a clinician assessing a patient. If these layers are collapsed, a map can appear more informative while actually hiding the evidence needed to act.
A probability of finding one tick is not a count
The Ohio researchers fitted two models because occurrence and abundance describe different stages of a surveillance question. The occurrence model asked whether a standardized search would find any blacklegged tick at a site. Its output was a probability of encountering at least one tick during a 750-metre drag. The abundance model asked about the number of tick encounters under that same general search definition. Because the counts combined larval encounters, nymphs, and adults, the result was relative abundance rather than a life-stage-specific population count.
A simple analogy is the difference between asking whether a shop has at least one customer and asking how busy it is. The first question can be answered with presence or absence. The second needs a count, and the count depends on when and how it was measured. A place can be likely to produce one encounter but not many, or it can have a high expected count in a habitat where the chance of detecting a tick varies with conditions. The two Ohio surfaces are related, but one cannot be substituted for the other.
The model inputs also behaved differently across these outcomes. The top occurrence model retained precipitation seasonality, aridity, and several forest-patch measures. Proportion of forest had a positive relationship with occurrence, while aridity and mean patch area had negative relationships. For some other effects, including the mean effects of largest patch index, number of patches, and precipitation seasonality, the 95% credible interval included zero. That means the estimated direction was not strong enough, in this model, to rule out a range that includes no relationship.
The top abundance model retained a different combination: mean temperature of the driest quarter, aridity, forest-patch measures, robin density, and deer density. Deer density had a positive relationship with relative abundance. Robin density had a negative relationship, but the authors did not treat that as evidence that robins reduce ticks. Robin density was lower in forested areas, and its strong negative correlation with forest proportion suggested that it was acting as a proxy for less suitable habitat in this model system.
These differences are a reminder that a variable can be associated with one outcome without determining another. Deer can matter greatly to reproduction and to the number of adult ticks while not being necessary to explain whether blacklegged ticks occur at all. A generalist vector may establish through many hosts, but once established, local host density or habitat configuration may influence how many ticks a standardized search encounters.