Free-ranging banded mongooses (Mungos mungo) in northeastern Botswana are infected by a novel Mycobacterium tuberculosis complex pathogen, M. mungi, which putatively infects mongooses through lesions in the skin (often the planum nasale) from an environmental reservoir. To understand the epidemiology of the yearly and highly seasonal outbreaks of M. mungi in this population of banded mongooses, researchers need to understand what factors influence banded mongoose exposure to M. mungi and banded mongoose susceptibility to M. mungi infection.

Researchers have no baseline data on the behavioral ecology of this population of banded mongooses - such as home range dynamics, denning ecology, movement ecology, and foraging ecology, all of which may play a role in banded mongoose exposure to M. mungi. Further, researchers have highlighted the potential role of prolonged elevations of glucocorticoids in impairing cell-mediated immunity, which would play a significant role in determining susceptibility to a mycobacterium such as M. mungi, however, researchers have no data on the endocrinology of banded mongooses. Finally, researchers have not detected M. mungi infection in any other population of banded mongooses. Our study population has a gradient of troops (social groups) that vary from troops with extremely close association with humans in a town, to troops associated with humans at tourist lodges within the Chobe National Park, to troops with no discernible association with humans within the national park and surrounding forest reserve. Researchers have few data on how synanthropy (living with humans) affects banded mongoose behavioral ecology and no data on how synanthropy affects banded mongoose endocrinology. Researchers do not know whether or how the high level of synanthropy in this population of banded mongooses plays a role in the epidemiology of M. mungi outbreaks.

Thus, we document here some aspects of banded mongoose home range dynamics, movement metrics, denning ecology and foraging behavior for our study population in northeastern Botswana. We present a novel method for screening data from global positioning system (GPS) collars for large measurement error and we present a detailed home range study. We also document the spatio-temporal dynamics of glucocorticoid production among several banded mongoose study troops across our study site, using a non-invasive assay for fecal glucocorticoid metabolites, which we validated and also present here.  We tested to see which factors, including nutritional limitation, predation risk, and reproduction (and associated competition, agonistic encounters, and predation), best explained the variation in glucocorticoid production among our study troops over several years.

We found that the metrics traditionally used to screen data from GPS collars, horizontal dilution of precision (HDOP) or fix dimension (2-D or 3-D), performed poorly relative to a new screening metric that we propose, the estimated elevation error (EEE). We propose that researchers use our screening method, which combines test data and a model-averaging information-theoretic framework that uses a priori candidate models of telemetry measurement error. Although we recommend including EEE in a priori candidate models, it may not describe telemetry error in other systems as well as it did in our own.

Banded mongooses in our study population formed troops of a median of 13 adults (IQR: 11 to 21 adults) and these troops used home ranges of a median of 68 ha (IQR: 39 to 134 ha) with core areas of a median of 15 ha (IQR: 9 to 28 ha). These cores (statistically-clumped space use) occurred at a median volume contour of 66 % (IQR: 58 to 71 %). Synanthropic troops showed more clumped area use than apoanthropic troops (those living away from humans). Synanthropic troops also used man-made structures for den sites in SI{81}{percent} of nights, fed from refuse sites in 13 % of foraging observations, and drank from anthropogenic water sources in 78 % of drinking observations.

From our conducted adrenocorticotropic hormone challenge, we detected valid increases in fecal glucocorticoid metabolite concentrations in mongoose feces using our four tested enzyme-immunoassays. An 11-oxoetiocholanolone assay detecting 11,17-dioxoandrostanes (11,17-DOA) performed best. Using this assay, we detected expected decreases in fecal glucocorticoid metabolite concentrations 48 h after administering dexamethasone sodium phosphate. We also validated this assay using biological events as challenges, in which captive mongooses showed higher fecal glucocorticoid metabolite concentrations during reproductive activity, agonistic encounters, and depredation events. The time delay of fecal glucocorticoid metabolite excretion approximately corresponded with food transit time, at a minimum of approximately 24 h. Fecal glucocorticoid metabolite metabolism was minimal up to 8 h post-defecation.

Reproduction and its associated challenges dramatically increased glucocorticoid production, which otherwise remained low and stable in a captive troop with a constant food supply and lowered predation risk. Variation in glucocorticoid production in free-ranging banded mongooses was best explained by food limitation as described by current nutritional limitation (proportion of fecal organic matter), recent rainfall (which increases soil macrofauna availability), and access to concentrated anthropogenic food resources. Habitat differences in soil macrofauna density and reproductive events also explained variation in glucocorticoid production in free-ranging mongooses, but to a much lower degree. Predation risk, as measured by canopy cover (escape from aerial predators) and group size (decreased per capita vigilance) explained very little of the variation in glucocorticoid production. In the late dry season, banded mongooses in our population may face a "perfect storm" of nutritional limitation, agonistic encounters at concentrated food resources, aggressive evictions, estrus, competition for mates, parturition, and predation pressure on pups. We suspect that this prefect storm may push glucocorticoid responses into homeostatic overload and may impair cell-mediated immunity in banded mongooses.