Search

Abstract: Rabies is a fatal zoonotic disease of mammals; it exacerbates the uncertainty of conserving populations of some threatened mammals (TM). Modelling affords an inexpensive, a priori way of studying key parameters of wildlife rabies transmission, rabies management economics, and TM conservation. Numerous models of rabies transmission have been published. Linear density dependent models predicted that a threshold density (K T ⩽1.0), possibly attained by culling or contraception, would eliminate an epizootic through reduced contacts among host animals. Density independent models predicted less advantage of culling and contraception in rabies control due to limited contacts among territorial host animals. Recent stochastic, mixed models offer novel predictions about the role of culling, fertility control, and oral rabies vaccination (ORV) in disease management. Use of a “threshold successful contact” rate (C T) as a parameter in these models predicts that density reduction of host animals will enhance ORV campaigns in non-TM contexts via more efficient bait delivery and vaccination. Economic analyses of medical, public health, and veterinary costs have shown post-exposure prophylaxis (PEP) and increased pet vaccinations (PV) to be major rabies-caused expenses during and after epizootics in North America. No modelling efforts have examined either the benefits-costs of rabies management strategies to conserve TM or the use of ORV, per se, to conserve TM – an omission due in part to the lack of methodologies for properly valuing TM (potential savings) and the expense or lower priority of using ORV for TM protection. This paper: (1) describes key aspects of rabies-transmission models in wildlife, (2) posits the use of C T to predict disease persistence, (3) reviews selected ORV strategies, economic studies, and benefit–cost models associated with the use of ORV as a means of rabies control in non-TM situations, (4) discusses implications of these models to the conservation of TM, and (5) recommends five steps to improve modelling of rabies transmission (wildlife disease in general), rabies-control economics, and TM conservation. [Copyright &y& Elsevier]

Abstract: A mixture of 16.47±0.038L concentrated capsicum oleoresin with 163.8±3.63L water and a total of 17.2L concentrated soybean oil with 177.1L water was dispensed over five 4.87m×4.87m plots via a grid of 128 augured holes (≈15cm-diameter×≈30cm deep) in a compacted, sandy loam soil. During the next 9 days, a total of 37 soil samples were obtained from these plots and other off-plot areas. Spectrophotometric analyses of dual aliquots of soil samples were performed to detect absorbance of capsaicin (λ max at 282nm). Analysis of variance showed that absorbance readings from points of capsicum application (augured holes) were 30-fold greater than for samples obtained 30–40cm distant (, df 1/4.22, ). Soybean-oil plot and off-plot soil samples yielded negligible absorbance readings. Thus, capsicum oleoresin persisted for ⩾10 days in this type of soil, but migration was minimal. [Copyright &y& Elsevier]

During 2000–2001, small mammals, birds, and potential corn/soybean damage were studied at a low-tillage, non-irrigated agricultural research site in the Colorado Piedmont. A small mammal survey involved four trapping sessions and 18, 12-live-trap grids each. Within years, two grids each were placed at random, fixed locations in experimental corn, fallow, millet, pea, soybean, sunflower, and wheat plots at the site; two off-plot grids each were set at random, fixed locations <100 m from the north and south edge of these plots. In 2001, periodic bird observations were conducted, and damage to corn and soybean plants was assessed. Capture rates were low during all trap sessions (range 0.1%–3.3%,X¯=2.2%). Sixty-three small mammals were captured and 39 were recaptured. Captures included deer mouse (Peromyscus maniculatus), northern grasshopper mouse (Onychomys leucogaster), thirteen-lined ground squirrel (Spermophilus tridecemlineatus), and western harvest mouse (Reithrodontomys megalotis). Deer mice accounted for 56 of the initial captures (88.9%). In-crop captures (n=15) and recaptures (n=16) were most frequent in wheat plots. Bird counts were low and included horned lark (Eremophila alpestris), killdeer plover (Charadrius vociferus), lark bunting (Calamospiza melanocorys), mountain plover (Charadrius montanus), mourning dove (Zenaida macroura), and western meadowlark (Sturnella neglecta). No direct seed removal, sprout removal, or plant clipping by small mammals or birds was observed, but some clipping of soybean plants was attributed to deer (Odocoileus virginianus and O. hemionus) and jack rabbits (Lepus townsendii or L. californicus). Plant debris accumulation is viewed as a key factor determining small mammal abundance and potential damage in low-till agriculture. [Copyright &y& Elsevier]

Accurate estimation of crop damage by wildlife often requires labor-intensive sampling procedures. Variable area transect (VAT) sampling has been identified as a potential labor-saving alternative to quadrat sampling. We fully enumerated all damaged corn plants in a field to assess accuracy of sampling by small quadrats 5 m long by one or two row-widths, and by VAT sampling with searches from random starting points to the 3rd, 4th, 5th, and 6th damaged plant. Two widths for strip transects also were considered for VAT sampling: one and two crop rows. Two analytical formulae were considered for the VAT data. Thus, 18 estimates of damage were produced and compared for accuracy with the true damage level. Sampling and enumeration were carried out the same day, so that the damage level could not change between enumeration and sampling. Estimation from VAT sampling using the original formula showed promise as a low-labor assessment method by producing highly accurate estimates (<1.5% absolute relative bias) when strip transects of two row-widths were searched to the 4th or 5th damaged plant. The formula designed to improve estimation for VAT sampling in aggregated populations had high biases. Quadrat sampling is well-known to produce unbiased results when optimized, and we restricted our research to only the low-effort situation, which also produced biased results. [Copyright &y& Elsevier]