A regression-based equivalence test for model validation: shifting the burden of proof

Model validation is often realized as a test of how well model predictions match a set of independent observations. One would think that the burden of proof should rest with the model, to force it to show that it can make accurate predictions. Further, one would think that increasing the sample size ought to increase the model's ability to demonstrate its utility. Traditional statistical tools are inappropriate for this because they default to the case that the model and the data are no different, and their ability to detect differences increases with the sample size. These traditional tools are optimized to detect differences, rather than similarities. We present an alternative strategy for model validation that is based on regression and statistical tests of equivalence. Equivalence tests reverse the usual null hypothesis: they posit that the populations being compared are different and use the data to prove otherwise. In this sense, equivalence tests are lumping tests, whereas the traditional statistical tests are splitting tests. To date, model validation with equivalence tests has focused on comparisons of means. Our proposed test checks not only for similarity of means, but also for similarity between individual predictions and observations. The strategy is demonstrated using three case studies that differ in their modeling objectives, and for varied sample sizes. The proposed strategy provides a formal means of model validation that is superior to traditional statistical tests in each case.     

Coarse root elongation rate estimates for interior Douglas-fir

Accurate estimates of root growth rates are important for root system modeling, and the spread of root systems may be an important determinant of belowground site occupancy. Estimating root system growth rates is complicated because missing, discontinuous, and false annual growth rings make root cross sections difficult to age. These irregularities can occur even in roots of dominant conifers with rare or absent stem growth ring abnormalities. Incomplete rings were noted in the root growth rings of nine co-dominant interior Douglas-fir (Pseudotsuga menziesii var. glauca (Beissn.) Franco) trees. Coarse root (> 1.0-cm diameter) elongation rates were estimated by fitting a geometric mean regression line to ring count and lateral distance data. In all nine roots examined, the geometric mean regression slope was well within the range of the 95% confidence interval for the ordinary least squares regression of lateral distance versus age, suggesting that measurement error may have been negligible. Coarse root elongation rates (which ranged from 2.8 to 15.3 cm year(-1) and averaged 7.4 cm year(-1)) in the interior Douglas-fir trees studied were much lower than those reported by others. This discrepancy may be a result of limited soil water availability, soil heterogeneity (both soil water content and soil texture were highly variable across short distances) and fragmentation of belowground growing space.     

Variation in sugar maple root respiration with root diameter and soil depth

Root respiration may account for as much as 60% of total soil respiration. Therefore, factors that regulate the metabolic activity of roots and associated microbes are an important component of terrestrial carbon budgets. Root systems are often sampled by diameter and depth classes to enable researchers to process samples in a systematic and timely fashion. We recently discovered that small, lateral roots at the distal end of the root system have much greater tissue N concentrations than larger roots, and this led to the hypothesis that the smallest roots have significantly higher rates of respiration than larger roots. This study was designed to determine if root respiration is related to root diameter or the location of roots in the soil profile. We examined relationships among root respiration rates and N concentration in four diameter classes from three soil depths in two sugar maple (Acer saccharum Marsh.) forests in Michigan. Root respiration declined as root diameter increased and was lower at deeper soil depths than at the soil surface. Surface roots (0-10 cm depth) respired at rates up to 40% greater than deeper roots, and respiration rates for roots < 0.5 mm in diameter were 2.4 to 3.4 times higher than those for roots in larger diameter classes. Root N concentration explained 70% of the observed variation in respiration across sites and size and depth classes. Differences in respiration among root diameter classes and soil depths appeared to be consistent with hypothesized effects of variation in root function on metabolic activity. Among roots, very fine roots in zones of high nutrient availability had the highest respiration rates. Large roots and roots from depths of low nutrient availability had low respiration rates consistent with structural and transport functions rather than with active nutrient uptake and assimilation. These results suggest that broadly defined root classes, e.g., fine roots are equivalent to all roots < 2.0 mm in diameter, do not accurately reflect the functional categories typically associated with fine roots. Tissue N concentration or N content (mass x concentration N) may be a better indicator of root function than root diameter.     

Field measurements of root respiration indicate little to no seasonal temperature acclimation for sugar maple and red pine

Increasing global temperatures could potentially cause large increases in root respiration and associated soil CO2 efflux. However, if root respiration acclimates to higher temperatures, increases in soil CO2 efflux from this source would be much less. Throughout the snow-free season, we measured fine root respiration in the field at ambient soil temperature in a sugar maple (Acer saccharum Marsh.) forest and a red pine (Pinus resinosa Ait.) plantation in Michigan. The objectives were to determine effects of soil temperature, soil water availability and experimental N additions on root respiration rates, and to test for temperature acclimation in response to seasonal changes in soil temperature. Soil temperature and soil water availability were important predictors of root respiration and together explained 76% of the variation in root respiration rates in the red pine plantation and 71% of the variation in the sugar maple forest. Root N concentration explained an additional 6% of the variation in the sugar maple trees. Experimental N additions did not affect root respiration rates at either site. From April to November, root respiration rates measured in the field increased exponentially with increasing soil temperature. For sugar maple, long-term Q10 values calculated from the field data were slightly, but not significantly, less than short-term Q10 values determined for instantaneous temperature series conducted in the laboratory (2.4 versus 2.62.7). For red pine, long-term and short-term Q10 values were similar (3.0 versus 3.0). Sugar maple root respiration rates at constant reference temperatures of 6, 18 and 24 degrees C were measured in the laboratory at various times during the year when field soil temperatures varied from 0.4 to 16.8 degrees C. No relationship existed between ambient soil temperature just before sampling and root respiration rates at 6 and 18 degrees C (P = 0.37 and 0.86, respectively), and only a very weak relationship was found between ambient soil temperature and root respiration at 24 degrees C (P = 0.08, slope = 0.09). We conclude that root respiration in these species undergoes little, if any, acclimation to seasonal changes in soil temperature.     

The lesions of toe tip necrosis in southern Alberta feedlot cattle provide insight into the pathogenesis of the disease

Gross and histologic postmortem studies were performed on the hind feet of feedlot cattle that had, or were free from, lesions of toe tip necrosis. The hind feet of feedlot cattle were collected by 3 veterinary feedlot practices in southern Alberta, Canada. Three studies of these feet were conducted: i) prediction of disease based on the presence or absence of apical white line separation, ii) gross assessment of the distribution and severity of lesions within affected claws, and iii) microscopic evaluation of the distal phalanx and surrounding soft tissues of affected claws. Prediction of toe tip necrosis based on the presence of apical white line separation was statistically significant (P < 0.0001). This, combined with a pattern of lesions indicative of an ascending infection of the distal phalanx and the absence of other lesions, suggests that the pathogenesis involves bacterial infection originating from the most distal aspect of the toe, at the apical white line.     

The increasing risk of Lyme disease in Canada

There is an increasing risk of Lyme disease in Canada due to range expansion of the tick vector, Ixodes scapularis. The objectives of this article are to i) raise public awareness with the help of veterinarians on the emerging and expanding risk of Lyme disease across Canada, ii) review the key clinical features of Lyme disease in dogs, and iii) provide recommendations for veterinarians on the management of Lyme disease in dogs.     

Effects of interferon-γ knockdown on vaccine-induced immunity against Marek’s disease in chickens

Interferon (IFN)-γ has been shown to be associated with immunity to Marek’s disease virus (MDV). The overall objective of this study was to investigate the causal relationship between IFN-γ and vaccine-conferred immunity against MDV in chickens. To this end, 3 small interfering RNAs (siRNAs) targeting chicken IFN-γ, which had previously been shown to reduce IFN-γ expression in vitro, and a control siRNA were selected to generate recombinant avian adeno-associated virus (rAAAV) expressing short-hairpin small interfering RNAs (shRNAs). An MDV challenge trial was then conducted: chickens were vaccinated with herpesvirus of turkey (HVT), administered the rAAAV expressing shRNA, and then challenged with MDV. Tumors were observed in 4 out of 10 birds that were vaccinated with HVT and challenged but did not receive any rAAAV, 5 out of 9 birds that were administered the rAAAV containing IFN-γ shRNA, and 2 out of 10 birds that were administered a control enhanced green fluorescent protein siRNA. There was no significant difference in MDV genome load in the feather follicle epithelium of the birds that were cotreated with the vaccine and the rAAAV compared with the vaccinated MDV-infected birds. These results suggest that AAAV-based vectors can be used for the delivery of shRNA into chicken cells. However, administration of the rAAAV expressing shRNA targeting chicken IFN-γ did not seem to fully abrogate vaccine-induced protection.