Suburbanization has greatly altered the epidemiological landscape. The steady rise in human Lyme disease incidence in the northeast over the last three decades is related to landscape modification through suburbanization (Maupin et al., 1991; Barbour and Fish, 1993; Frank et al., 1998). Eastern deciduous forest communities were fragmented by suburban development, resulting in a habitat matrix that is ideal for deer and some small rodents, especially the white-footed mouse (Daily and Ehrlich, 1995). Although the net amount of forest has not changed significantly in the past two decades (Brownstein et al., 2003), continued building has created a higher number of forest patches interspersed with residential development. The proximity of humans, wildlife, and ticks in a habitat that is suitable for all three virtually ensures exposure of people to tick bites and the agents of Lyme disease, human granulocytic anaplasmosis, and babesiosis (Falco and Fish, 1988a; Ehrlich and Ehrlich, 2002). Furthermore, deer are often protected from hunting - a situation that not only permits the maintenance of a large tick population, but has also led to increased tension among residents who do and do not advocate controlling the number of deer in areas where Lyme disease is endemic (see below). While suburban growth has fostered I. scapularis population expansion, the spread is not uniform and the resultant tick distribution is patchy, with varying densities in adjacent areas. In general, all landscapes can be thought of as mosaics composed of discrete bounded patches that are differentiated by a number of biotic (e.g. vegetation features, host abundance) and abiotic (e.g. weather) factors (Pickett and Cadenasso, 1995). Patches can arise because of natural or human-caused disturbance, fragmentation of the land, regeneration of a habitat type, or persistent differences in environmental resources (Forman and Godron, 1986). In the case of Lyme disease, the deciduous forest matrix contains small habitat patches, some of which may be suitable for I. scapularis and some not. The relationship among habitat patches across a range of spatial scales provides important information on factors influencing the distribution of I. scapularis.
Past studies have used a landscape approach to identify habitat features associated with the presence of I. scapularis. One, conducted in a residential area of Westchester County, New York, focused on a single neighborhood and found considerable variation among sampled properties. Although Maupin et al. (1991) noted a general pattern in which progressively fewer ticks were found on moving from the woodland to the wooded edge and onto residential lawns, it appears that the edge or ecotone habitat probably was underestimated. Edge effects may extend some distance into adjacent woodlots (Harris, 1984), though Maupin etal. (1991) limited consideration of edge to "non-ornamental, unmaintained edge which abutted woodlots..." (p. 1106), and this may have resulted in too few ticks being counted.
Dister et al. (1993) conducted a regional analysis of habitat features throughout Westchester County, New York, using LANDSAT Thematic Mapper imagery and a GIS (geographic information system) relating land-cover composition to canine seroprevalence for antibodies to B. burgdorferi. They noted that the percentage of land cover represented by deciduous forest was the variable most highly correlated with canine seroprevalence, a measure of Lyme disease risk (Daniels et al., 1993; Falco et al., 1993). However, Dister et al. (1993) concluded that further analysis of spatial context with finer-scaled databases was needed to understand the influence that surrounding land cover classes may have on Lyme disease risk.
This approach offers a systematic way of discerning which ecological features of the landscape influence tick abundance, and how temporal changes in those features play a role at the population level. Because the suburban landscape is a disjunct pattern of lawns, homes, and woodland defined by human activity, anthropogenic effects on habitat structure are important to consider. Among the anthropogenic effects that may have a strong influence on tick distribution is the creation of edge habitat or ecotone, which is characteristic of the fragmentation that accompanies the building of homes in previously uninhabited woodland (Temple and Wilcox, 1985; Verner, 1985). Such areas appear to be particularly important for adult I. scapularis. For example, Stafford and Magnarelli (1993) noted that adult ticks were recovered predominantly from lawn and ecotone areas, and Schmidtmann et al. (1994) found adults most commonly at the woodland-pasture interface (i.e., edge) on their study sites. Kramer et al. (1993) observed that drag-sampling along trails in parkland (essentially narrow corridors bordered by edge) yielded relatively high numbers of adults. Using county case reports, Glass et al. (1995) found that in Baltimore County, Maryland, the risk of human Lyme disease decreased with increasing distance from forest edge (Dennis et al., 1998).
As noted earlier, the significance of edge habitats lies in their attractiveness to white-tailed deer (Harlow, 1984) and a variety of rodent species (see, for example, Diffendorfer et al., 1995), all of which are fed upon routinely by I. scapularis (Fish and Dowler, 1989). When wildlife hosts utilize these habitats, engorged ticks may be dropped that will later molt to host-seeking ticks with the potential to transmit pathogens to their next host - possibly a human. This is especially relevant in the case of P. leucopus, which commonly inhabits ecot ones. Differences in the abundance of white-footed mice from one site to another may be related to the distribution of forest fragments as a result of land development (Krohne and Hoch, 1999).
Although fragmentation of the landscape has been shown to increase both tick density and infection prevalence (the percentage of ticks infected with a specific pathogen, B. burgdorferi in this case) (Ostfeld and Keesing, 2000; Schmidt and Ostfeld, 2001), two components of the entomologic risk of Lyme disease, work by Brownstein et al. (2005) indicates that the incidence of human disease is actually lower in fragmented contexts. These investigators speculate that human contact with infected ticks may be lower with increased fragmentation because of a reduction in forest patch size and an increase in isolation. In that case, fewer properties will contain or adjoin woodlots that support tick populations, resulting in an overall reduction in peridomestic exposure (Brownstein et al., 2005). It might be concluded that although development will increase the risk of Lyme disease to a point, further development resulting in increased isolation of a residential property (farther distance from a woodland patch) will then reduce risk. Since it is unlikely that most people residing in Lyme disease-endemic areas would support increased development as a means of evading infection, the way that landscape structure affects risk should be considered in future residential planning.
Was this article helpful?