Pathogens that spill over between species cause a significant human and animal health burden. Here, we describe characteristics of animal reservoirs that are required for pathogen spillover. We assembled and analyzed a database of 330 disease systems in which a pathogen spills over from a reservoir of one or more species. Three-quarters of reservoirs included wildlife, and 84% included mammals. Further, 65% of pathogens depended on a community of reservoir hosts, rather than a single species, for persistence. Among mammals, the most frequently identified reservoir hosts were rodents, artiodactyls, and carnivores. The distribution among orders of mammalian species identified as reservoirs did not differ from that expected by chance. Among disease systems with high priority pathogens and epidemic potential, we found birds, primates, and bats to be overrepresented. We also analyzed the life history traits of mammalian reservoir hosts and compared them to mammals as a whole. Reservoir species had faster life history characteristics than mammals overall, exhibiting traits associated with greater reproductive output rather than long-term survival. Thus, we find that in many respects, reservoirs of spillover pathogens are indeed special. The described patterns provide a useful resource for studying and managing emerging infectious diseases.

Summary1. Variation in wood specific gravity (WSG) within and across species of tropical trees is poorly studied in relation to vegetation change during tropical forest succession. We investigated WSG of 91 species in eight long-term plots along a successional chronosequence in north-eastern Costa Rica. 2. Radial changes in WSG were described by the rate of change (slope) along the stem radius. Significant radial changes were found in 42 of 74 species, with 37 species exhibiting increases from inner to outer wood, and five exhibiting decreases. Radial increases were commonly observed in species with inner WSG below 0Á5, whereas radial decreases were observed in species with inner WSG above 0Á7. 3. Wood specific gravity weighted by cross-sectional area (wWSG) varied four-fold among species. Species classified as second-growth specialists had lower wWSG, lower inner and outer WSG, and higher slopes than old-growth specialists; successional generalists showed intermediate trait values. 4. Among 18 species sampled in both second-and old-growth forests, four species (22%) showed significant variation in wWSG between forest types. Of 33 widely sampled species, seven species (21%) showed a significant effect of stem diameter on wWSG. 5. Second-growth plots had lower stand-level wWSG and more pronounced radial increases than old-growth plots. Individual tree biomass for species with radial increases and low WSG was substantially underestimated when based on unweighted WSG compared with wWSG. 6. Wood specific gravity varied with successional stage at multiple levels: within species, among successional specialist groups, and across plots of varying ages. Radial increases in WSG are common among trees in early and intermediate stages of tropical forest succession. This trait may confer growth advantages early in succession and increased resistance to physical or biotic damage during later successional stages.

Through their potentially devastating impacts on the environment, wildfires may impact pathogen, vector, and host interactions, leading to changing risks of vector‐borne disease in humans and other animals. Despite established risks for tick‐borne disease and increasing frequency and severity of wildfires in the United States, impacts of wildfire on ticks and tick‐borne pathogens are understudied. In 2015, the large Wragg fire extensively burned a long‐term field site at Stebbins Cold Canyon University of California Reserve (CC). We characterized the tick, reservoir host and pathogen community over a two‐year period after the burn, comparing our findings to pre‐fire data and to data from Quail Ridge Reserve (QR), a nearby unburned site. After the fire, there were 5.5 times more rodent, primarily Peromyscus spp., captures at CC than QR (compared to 3.5 times more pre‐fire). There were significantly fewer dusky‐footed woodrats (Neotoma fuscipes) at both sites post‐fire, likely due to drought but not fire. Pre‐fire tick infestation prevalence on rodents was comparable across sites (12.5% at CC and 9.9% at QR) and remained low at CC post‐fire (13.7%) but was significantly higher at QR (48.0%), suggesting that ticks or their habitat were destroyed during the burn. Normalized difference vegetation indices documented a 16‐fold loss of vegetation post‐ compared to pre‐fire at CC; loss of vegetation and direct impacts on fauna are likely the main drivers of the post‐fire differences in ticks we saw at CC. These data contribute to our understanding of tick‐associated disease risks in our increasingly disturbed landscapes.

Timber harvest may impact tick-borne disease by affecting small mammal and tick community structures. We assessed tick and small mammal populations in older second-growth redwood (Sequoia sempervirens (D. Don) Endl) habitat at two harvested sites in Santa Cruz County, California, where local risk of tick-borne disease is high and determined the prevalence of tick-borne pathogens in ticks. After single-tree removal harvest in 2014, there was a modest reduction in canopy, primarily toward the end of the study. Harvested sites showed strong reductions in California mouse (Peromyscus californicus, (Gambel)) captures 2-yr after harvest, resolving such that treatments and controls were comparable by the end of the study. Following harvest, treated sites experienced a transient decreased tick infestation while control plots experienced an increase. Ixodes angustus (Neumann) infestation probability on harvested plots decreased immediately after harvest, increasing with time but remaining lower than control plots, whereas I. pacificus (Cooley and Kohls) prevalence was higher shortly after the harvest on harvested plots, and continued to increase. Mean abundance of ticks on vegetation increased on control plots. We detected Borrelia burgdorferi ((Johnson et al.) Baranton) and Anaplasma phagocytophilum ((Foggie 1949) Dumler) in 3.8 and 3.1% of ticks on rodents, but no differences were associated with harvest. Impacts of forest harvest on tick-borne disease depend on removal practice and intensity, whether or not hosts are habitat specialists, and whether or not ticks are host specialists.

No abstract