Testing prediction accuracy in short-term ecological studies

Applied ecology is based on an assumption that a management action will result in a predicted outcome. Testing the prediction accuracy of ecological models is the most powerful way of evaluating the knowledge implicit in this cause-effect relationship, however, the prevalence of predictive modeling and prediction testing are spreading slowly in ecology. The challenge of prediction testing is particularly acute for small-scale studies, because withholding data for prediction testing (e.g., via k-fold cross validation) can reduce model precision. However, by necessity small-scale studies are common. We use one such study that explored small mammal abundance along an elevational gradient to test prediction accuracy of models with varying degrees of information content. For each of three small mammal species, we conducted 5000 iterations of the following process: (1) randomly selected 75 % of the data to develop generalized linear models of species abundance that used detailed site measurements as covariates, (2) used an information theoretic approach to compare the top model with detailed covariates to habitat type-only and null models constructed with the same data, (3) tested those models’ ability to predict the 25 % of the randomly withheld data, and (4) evaluated prediction accuracy with a quadratic loss function. Detailed models fit the model-evaluation data best but had greater expected prediction error when predicting out-of-sample data relative to the habitat type models. Relationships between species and detailed site variables may be evident only within the framework of explicitly hierarchical analyses. We show that even with a small but relatively typical dataset (n = 28 sampling locations across 125 km over two years), researchers can effectively compare models with different information content and measure models’ predictive power, thus evaluating their own ecological understanding and defining the limits of their inferences. Identifying the appropriate scope of inference through prediction testing is ecologically valuable and is attainable even with small datasets.     

Mapping economic trait loci for somatic cell score in Holstein cattle using microsatellite markers and selective genotyping

Marker-assisted selection (MAS) uses genetic marker genotypes to predict an animal's production potential and will provide additional selection information for progeny testing. With the discovery of highly polymorphic microsatellite markers, the tools now exist to begin the search for economic trait loci (ETL), which is the first step toward MAS. The objective of this study was to identify ETL for somatic cell score in an existing Holstein population. Using the granddaughter design, sons from seven grandsire families were genotyped with 20 autosomal microsatellites from five chromosomes (4, 8, 13, 17, 23), with an emphasis on chromosome 23, which is the location of the bovine major histocompatibility complex (BoLA). Selective genotyping was used to reduce the number of genotypes required, in which the 10 highest and 10 lowest sons from the phenotypic distribution curve were tested (140 sons in seven families). One marker (513), located near BoLA, showed evidence of an ETL in three of five polymorphic families. Additional sons were genotyped from the five families to estimate the effect and to compare selective and ‘complete’ genotyping. Both methods detected an ETL at marker 513, but in different families. This study provides evidence of the usefulness of microsatellite markers and the granddaughter design in the detection of ETL; however, additional markers need to be evaluated to determine the usefulness of selective genotyping. Based on the results from the 20 studied markers, the most likely position of a somatic cell score ETL lies near marker 513, located on chromosome 23.     

Observations on the Staining of Corynebacterium Diphtheriae

Albert's method, of staining diphtheria cultures consists of staining a fixed smear for one minute (some laboratories stain for five minutes) with a solution containing toluidine blue and malachite (or methyl) green, washing with water, and applying Albert's iodine for one minute. This procedure is discussed and criticized, and in addition the mechanism of the stain is elucidated. Also, the procedure which involves staining a fixed smear for one minute with Loeffler's alkaline methylene blue solution is discussed and criticized.

To overcome the objections to the above staining methods, a different method is proposed. This consists of staining a fixed smear with an acid solution of toluidine blue, washing with water, applying Albert's iodine for one minute, washing with water, and finally applying a safranin solution for 15-20 seconds. The theoretical basis for this method is presented.     

Outlier detection methods are still effective even using virtual species created with the probabilistic approach

Liu et al. (Journal of Biogeography, 2018, 45 :164–176) presented an approach to detect outliers in species distribution data by developing virtual species created using the threshold approach. Meynard et al. (Journal of biogeography, 2019, 46 :2141–2144) raised concerns about this approach stating that ‘using a probabilistic approach … may significantly change results’. Here we provide a new series of simulations using the two approaches and demonstrate that the outlier detection approach based on pseudo species distribution models was still effective when using the probabilistic approach, although the detection rate was lower than when using the threshold approach.     

Differentiation of bermudagrass (Cynodon spp.) genotypes by AFLP analyses

 Bermudagrasses (Cynodon spp.) are major turfgrasses for home lawns, public parks, golf courses and sport fields, and are widely adapted to tropical and warmer temperate climates. Morphological and physiological characteristics are not sufficient to differentiate some bermudagrass genotypes because the differences between them are often subtle and subject to environmental influence. In this study, a DNA-typing technique, amplified fragment length polymorphism (AFLP), was used to differentiate bermudagrass genotypes and to explore their genetic relationships. Twenty seven bermudagrass cultivars and introductions, mostly from the Coastal Plain Experiment Station in Tifton, Ga., were assayed by the radioactive (³²P) and the fluorescence-labeled AFLP methods. The AFLP technique produced enough polymorphism to differentiate all 27 bermudagrass genotypes, even the closely related ones. An average of 48–74 bands in the 30–600-bp size range was detected by the ³²P-labeled AFLP method. The results indicated that most of the 14 primer combinations tested in this study could be used to distinguish bermudagrass genotypes, and that some single primer-pairs could differentiate all 27 of them. To test the reliability and reproducibility of the AFLP procedure, three DNA isolations (replications) of the 27 bermudagrass genotypes were assayed using five primer pairs. Only 0.6% of the bands were evaluated differently among the three replications. One replication of one genotype (which was most likely a planting contaminant) was grouped in an unexpected cluster using the Unweighted Pair Group Mean Average (UPGMA) method. A one- or two-band difference in scoring did not change the clustering of genotypes or the replications within genotypes. The 27 genotypes were grouped into three major clusters, many of which were in agreement with known pedigrees. Trees constructed with different primer combinations using ³²P- and fluorescence-labelling formed similar major groupings. The semi-automated fluorescence-based AFLP technique offered significant improvements on fragment sizing and data handling. It was also more accurate for detection and more efficient than the radioactive labelling method. This study shows that the AFLP technique is a reliable tool for differentiating bermudagrass genotypes and for determining genetic relationships among them. Received: 28 July 1998 / Accepted: 3 November 1998     

The importance of subjectivity in perceptual errors on the emergence of indirect reciprocity

Indirect reciprocity is one mechanism that allows for unilateral resource giving among n-persons. Using analytical methods and computer simulations, previous studies have examined a number of strategies that make indirect reciprocity possible. In particular, previous investigations have concentrated on whether differentiating between justified and unjustified not-giving is important. However, whether or not a given strategy is ESS depends on the type of perceptual errors that are assumed. When errors are objective, regarding those who do not give to "bad" as "good" is critical. When perceptual errors are subjective, however, regarding those who give to "bad" as "bad" is critical. Since we believe that there is no guarantee that perceptual errors are shared among all individuals in a society, we argue that the latter moral principle may play a more important role in human interactions.     

Modeling markers of disease progression by a hidden Markov process: application to characterizing CD4 cell decline

Multistate models have been increasingly used to model natural history of many diseases as well as to characterize the follow-up of patients under varied clinical protocols. This modeling allows describing disease evolution, estimating the transition rates, and evaluating the therapy effects on progression. In many cases, the staging is defined on the basis of a discretization of the values of continuous markers (CD4 cell count for HIV application) that are subject to great variability due mainly to short time-scale noise (intraindividual variability) and measurement errors. This led us to formulate a Bayesian hierarchical model where, at a first level, a disease process (Markov model on the true states, which are unobserved) is introduced and, at a second level, the measurement process making the link between the true states and the observed marker values is modeled. This hierarchical formulation allows joint estimation of the parameters of both processes. Estimation of the quantities of interest is performed via stochastic algorithms of the family of Markov chain Monte Carlo methods. The flexibility of this approach is illustrated by analyzing the CD4 data on HIV patients of the Concorde clinical trial.     

On generalized Simes critical constants

We consider the problem treated by Simes of testing the overall null hypothesis formed by the intersection of a set of elementary null hypotheses based on ordered p‐values of the associated test statistics. The Simes test uses critical constants that do not need tabulation. Cai and Sarkar gave a method to compute generalized Simes critical constants which improve upon the power of the Simes test when more than a few hypotheses are false. The Simes constants can be viewed as the first order (requiring solution of a linear equation) and the Cai‐Sarkar constants as the second order (requiring solution of a quadratic equation) constants. We extend the method to third order (requiring solution of a cubic equation) constants, and also offer an extension to an arbitrary kth order. We show by simulation that the third order constants are more powerful than the second order constants for testing the overall null hypothesis in most cases. However, there are some drawbacks associated with these higher order constants especially for , which limits their practical usefulness.     

Field test of an automated radio‐telemetry system: tracking local space use of aerial insectivores

Documenting local space use of birds that move rapidly, but are too small to carry GPS tags, such as swallows and swifts, can be challenging. For these species, tracking methods such as manual radio‐telemetry and visual observation are either inadequate or labor‐ and time‐intensive. Another option is use of an automated telemetry system, but equipment for such systems can be costly when many receivers are used. Our objective, therefore, was to determine if an automated radio‐telemetry system, consisting of just two receivers, could provide an alternative to manual tracking for gathering data on local space use of six individuals of three species of aerial insectivores, including one Cliff Swallow (Petrochelidon pyrrhonota), one Eastern Phoebe (Sayornis phoebe), and four Barn Swallows (Hirundo rustica). We established automated radio‐telemetry systems at three sites near the city of Peterborough in eastern Ontario, Canada, from May to August 2015. We evaluated the location error of our two‐receiver system using data from moving and stationary test transmitters at known locations, and used telemetry data from the aerial insectivores as a test of the system's ability to track rapidly moving birds under field conditions. Median location error was ~250 m for automated telemetry test locations after filtering. More than 90% of estimated locations had large location errors and were removed from analysis, including all locations > 1 km from receiver stations. Our automated telemetry receivers recorded 17,634 detections of the six radio‐tagged birds. However, filtering removed an average of 89% of bird location estimates, leaving only the Cliff Swallow with enough locations for analysis of space use. Our results demonstrate that a minimal automated radio‐telemetry system can be used to assess local space use by small, highly mobile birds, but the resolution of the data collected using only two receiver stations was coarse and had a limited range. To improve both location accuracy and increase the percentage of usable location estimates collected, we suggest that, in future studies, investigators use receivers that simultaneously record signals detected by all antennas, and use of a minimum of three receiver stations with more antennas at each station.