Statistical modeling and machine learning present opportunities to address complex hypotheses that relate the physical environment to biological distributions of zooplankton. Application of relatively novel models (such as Zero Inflated Negative Binomial (ZINB), Hurdle, etc.) to the biological oceanography field can help to handle many of the traditional issues of patchy distributions, variable effort and confounding spatial and temporal components from sampling design. Automated identification and selection of photographic and video from either field or bench-top equipment presents a way to drastically increase the processing capability for zooplankton research projects.

I am primarily interested in the interplay between plankton and the physical environment that drives their spatial and temporal distributions. My goal is to use a combination of net sampling, in situ imaging, laboratory, satellite, and modeling to address integrative hypotheses about individuals through populations. My previous research has focused on how periodic atmospheric events contribute to recruitment of estuarine dependent species, age and growth, and small scale spatial movement and distribution of zooplankton. Currently my work is aimed at understanding larger patterns in zooplankton spawning and primary production, machine learning for automated lab identification, and spatial distributions of mid-water and benthic fauna. I am always interested in new opportunities for collaboration, and you can contact me

In situ imaging of zooplankton provides multiple advantages over traditional sampling systems. The number of different imaging systems developed and currently in development each have advantages for particular environments and conditions. However, the ability to image soft-bodied zooplankton without damaging them provides opportunities for estimation of feeding volumes and spatial distribution. Long series video measurements collected in situ can provide information on the patchy distribution, and when combined with remote sensing data may help to further elucidate the nature of the primary to secondary to tertiary steps in the ecological ocean model.

In microtidal areas or where periodic, high-energy events can exert strong physical forcing of waters either into or out of estuaries, mechanisms for larval recruitment may differ significantly from environments where there is more vertical structure in the water column. Moreover, a combination of both the physical and biological components may have variable and significant effects on retention for estuarine dependent larvae spawned offshore. In general, the effects of these periodic atmospheric events appear to be species specific, and have variable magnitudes of importance for recruitment and retention, depending on the life history strategy with respect to spawning.

Matthew J Kupchik

Office 2143 Energy, Coast, & Environment Building

Department of Oceanography

Louisiana State Univeristy

Baton Rouge, Louisiana 70803

The partnership between science and the oil/gas industry with SERPENT represents an ability to provide  research and data on the deep sea that would be either impossible due to time and platform constraints, or exceedingly too expensive to conduct with high frequency. This type of partnership also presents as model where industries with an interest in the marine environment and scientists can work together to address questions of both ecological and management importance.

Age and growth from otoliths is a long standing and fairly common practice. However, the use of larval otoliths for not only age and growth, combined with hydrodynamic models to estimate spawning locations can help management strategies to avoid affecting spawning zones and to avoid targeting spawning females. Machine learning and estimation using signal processing can also be used for rapid reading and to help alleviate a lot of the current backlog of NOAA otolith samples. Combined with the stable isotope analysis across the reading radii, remote sensing data products and estimated spawning foci can further investigate the hypothesis on spawning theory in an ecosystem based approach with respect to management.