The relationship between species detection probability and local extinction probability

In community-level ecological studies, generally not all species present in sampled areas are detected. Many authors have proposed the use of estimation methods that allow detection probabilities that are <1 and that are heterogeneous among species. These methods can also be used to estimate community-dynamic parameters such as species local extinction probability and turnover rates (Nichols et al. Ecol Appl 8:1213–1225; Conserv Biol 12:1390–1398). Here, we present an ad hoc approach to estimating community-level vital rates in the presence of joint heterogeneity of detection probabilities and vital rates. The method consists of partitioning the number of species into two groups using the detection frequencies and then estimating vital rates (e.g., local extinction probabilities) for each group. Estimators from each group are combined in a weighted estimator of vital rates that accounts for the effect of heterogeneity. Using data from the North American Breeding Bird Survey, we computed such estimates and tested the hypothesis that detection probabilities and local extinction probabilities were negatively related. Our analyses support the hypothesis that species detection probability covaries negatively with local probability of extinction and turnover rates. A simulation study was conducted to assess the performance of vital parameter estimators as well as other estimators relevant to questions about heterogeneity, such as coefficient of variation of detection probabilities and proportion of species in each group. Both the weighted estimator suggested in this paper and the original unweighted estimator for local extinction probability performed fairly well and provided no basis for preferring one to the other.     

From heritability to probability

Can a heritability value tell us something about the weight of genetic versus environmental causes that have acted in the development of a particular individual? Two possible questions arise. Q1: what portion of the phenotype of X is due to its genes and what portion to its environment? Q2: what portion of X’s phenotypic deviation from the mean is a result of its genetic deviation and what portion a result of its environmental deviation? An answer to Q1 provides the full information about X’s development, while an answer to Q2 leaves out a large portion unexplained—that portion which corresponds to the phenotypic mean. Q1 is unanswerable, but I show it is nevertheless legitimate under certain quantitative genetics models. With regard to Q2, opinions in the philosophical and biological literature differ as to its legitimacy. I argue that not only is it legitimate, but in particular, under a few simplifying assumptions, it allows for a quantitative probabilistic answer: for normally distributed quantitative traits with no G-E correlation or statistical G × E interaction, we can assess the probability that X’s genes had a greater effect than its environment on its deviation from the mean population value. This probability is expressed as a function the heritability and the individual’s phenotypic value; we can also provide a quantitative probabilistic answer to Q2 for an arbitrary individual where the probability is a function only of heritability.     

Logical networks and probability

The logical network of a Pitts-McCulloch type is developed further with the specific addition of a motivational system which acts as a differential reinforcer to the network. A consideration is given to the various ways that probabilities enter the logical net transforming it into a probability net, and lastly some notes are added on the possibility of including the strategies of Theory of Games in the logical network.     

The probability of parallel evolution

Abstract How often will natural selection drive parallel evolution at the DNA sequence level? More precisely, what is the probability that selection will cause two populations that live in identical environments to substitute the same beneficial mutation? Here I show that, under fairly general conditions, the answer is simple: if a wild‐type sequence can mutate to n different beneficial mutations, replicate populations will on average fix the same mutation with probability P= 2/(n + 1). This probability, which is derived using extreme value theory, is independent of most biological details, including the length of the gene in question and the precise distribution of fitness effects among alleles. I conclude that the probability of parallel evolution under natural selection is nearly twice as large as that under neutrality.     

Predictive probability in clinical trials

In a recent paper, Choi and Pepple (1989, Biometrics 45, 317-323) consider the use of predictive probabilities in the monitoring of clinical trials. In particular, they characterise the predictive probability as a "useful conservative measure" for monitoring purposes. In this note the nature and source of this "conservatism" are investigated.     

Fractal probability measures of learning

Herein we introduce a technique for determining the fractal dimension of time series obtained from complex systems, in particular brain-wave data in which the fractal dimension is arguably a measure of awareness and learning. The technique is based on determining the probability distribution for the degree of irregularity in random time series and has been shown to be superior in terms of efficiency and reliability to more commonly used methods that rely on the correlation function. We speculate that the scaling behavior of the probability measure is an indicator of an allometric relation between learning and brain-wave activity.