Automatic detection of fish scale circuli using deep learning.

Teleost fish scales form distinct growth rings deposited in proportion to somatic growth in length, and are routinely used in fish ageing and growth analyses. Extraction of incremental growth data from scales is labour intensive. We present a fully automated method to retrieve this data from fish scale images using Convolutional Neural Networks (CNNs). Our pipeline of two CNNs automatically detects the centre of the scale and individual growth rings (circuli) along multiple radial transect emanating from the centre. The focus detector was trained on 725 scale images and achieved an average precision of 99%; the circuli detector was trained on 40 678 circuli annotations and achieved an average precision of 95.1%. Circuli detections were made with less confidence in the freshwater zone of the scale image where the growth bands are most narrowly spaced. However, the performance of the circuli detector was similar to that of another human labeller, highlighting the inherent ambiguity of the labelling process. The system predicts the location of scale growth rings rapidly and with high accuracy, enabling the calculation of spacings and thereby growth inferences from salmon scales. The success of our method suggests its potential for expansion to other species.

Information use and resource competition: an integrative framework.

Organisms may reduce uncertainty regarding how best to exploit their environment by collecting information about resource distribution. We develop a model to demonstrate how competition can facilitate or constrain an individual's ability to use information when acquiring resources. As resource distribution underpins both selection on information use and the strength and nature of competition between individuals, we demonstrate interdependencies between the two that should be common in nature. Individuals in our model can search for resources either personally or by using social information. We explore selection on social information use across a comprehensive range of ecological conditions, generalizing the producer-scrounger framework to a wide diversity of taxa and resources. We show that resource ecology--defined by scarcity, depletion rate and monopolizability--determines patterns of individual differences in social information use. These differences suggest coevolutionary processes linking dominance systems and social information use, with implications for the evolutionary demography of populations.