Paper No. 290-2
Presentation Time: 9:00 AM-6:30 PM
HIGH-THROUGHPUT SEMI-3D IMAGING OF MACROINVERTEBRATES: A TEST CASE USING NORTHEASTERN PACIFIC PATELLOGASTROPODS
For many questions related to body size and shape evolution, a single exemplar measurement is used to represent a species' average state. This approach is taken due to the difficulty and extremely time-intensive nature of measuring size and shape in assemblages using traditional methods, but necessarily ignores intraspecific variation in time and space. Here we showcase a new method, high-throughput imaging, that allows large scale measurement of entire assemblages, collected to reconsider latitudinal gradients in molluscan body size distributions. This type of imaging allows for the conversion of imaged specimens into data that can then be used for morphometric analysis. It is highly useful for projects requiring large numbers of specimens, as a single image can contain hundreds of individuals, and the total image processing time is much less than would be required for manual extraction of size data. Our automated image processing software (Automorph) was designed for use with microfossils; we tested its applicability to macroinvertebrates using recent patellogastropods from the collections of the University of California Museum of Paleontology (UCMP). We focused on the northeastern Pacific coast (Baja California to Alaska), as this region includes strong environmental and climatic gradients and has been subject to intensive study and collecting.
In six weeks we imaged over 1400 lots of specimens, ranging in size from a single specimen to several hundred, from ~430 locations. During processing, lots are segmented into images of individual specimens, and each is assigned locality data. Simple morphometric data such as length, width, height, perimeter, and volume are extracted algorithmically. Because specimen information is attached to each image, they can later be used for additional morphometric analyses. We will use data generated from this pilot study to assess both recent and Pleistocene size frequency distributions and size and shape gradients along the northeastern Pacific coast. All images will ultimately be uploaded to UCMP servers and made available for further study. High-throughput morphometric processing can vastly increase the amount of data that can feasibly be gathered, expanding the potential scope and resolution of investigations of body size and morphometric gradients in space and time.