Summary Report for Network Analysis Workshop (September 20, 2003)

Robert R. Christian and Alan P. Covich

May 4, 2004

Summary of workshop

This was the third workshop that has been held on the topic of the use of network analysis within the LTER network.  Previous workshops were organized to introduce the modeling system, collectively called “network analysis” (NA) to LTER scientists.  The long-term purpose of these workshops has been to build a community of scientists who have expertise in NA across the LTER program.  With satisfaction we can state that at least three sites have begun the process of using network analysis in their programs; Palmer Station, McMurdo, and Luquillo.  This is above the ongoing work at the Virginia Coast Reserve.  Thus, our goals have been at least partially met.  We now have trained scientists from several LTER sites and know of scientists at other LTER sites and elsewhere who use NA.  

The workshop was attended by approximately 20 individuals from a variety of LTER sites and researchers not directly associated with the LTER program.  The following LTER sites were represented: VCR, PAL, FCE, LUQ, and network office.

The workshop began with a series of presentations.

• Bob Christian, East Carolina University. 
• An overview and history of project, including a brief review of previous workshops .  Emphasis was placed on developing products from the activities.
• Alan Covich, University of Georgia. 
• Purpose of workshop and experiences.  Several opportunities were described.
• Todd Crowl, Utah State University. 
• Network concepts in analysis of stream food webs: structure and function of detrital-based food webs.  A series of foodweb networks have been constructed, and initial visualizations have been performed.
• Bob Daniels, College of William & Mary (student of Hugh Ducklow). 
• Inverse model analysis of plankton food webs in the Western Antarctic Peninsula.  The work included the use of inverse analysis, a statistical technique for network construction.
• Kerry McKenna, University of Toledo (student of Daryl Morehead)
• Models of microbial food webs in dry valley lakes.  Dynamic foodweb models were constructed on Stella.  The dynamic models were used to mass balance the system for future network analysis and further dynamic simulation.
• Ann Krause, Michigan State University (student of Doran Mason). 
• Organization of foodweb networks.  She discussed techniques to aid aggregation rules and interpretation of aggregation schemes.  Her work is on invasive species and their effects on food webs in Lake Michigan.  This is work outside the LTER network.
• Doran Mason, NOAA/GLERL. 
• Invasive species and food web disruption.  A summary was given of how network analysis is being used in assessing the effects of invasive species in the Great Lakes and neighboring systems.  A Windows based version of NETWRK4 is being developed that will aid future work.
• Ursula Scharler, University of Maryland. 
• She summarized work being done in Belize on mangroves, using network analysis.

Presentations were followed by discussion on manuscripts, proposals and subsequent workshops.  Network analysis was seen as an avenue for evaluating the role that primary producers play and how that role may be altered should an invasive primary producer enter a food web or should disturbance alter the community of producers.  This will be pursued in the future.

Evaluation

The workshop can be considered a success in several ways:  (1) It drew a respectable number of researchers from a variety of LTER sites.  (2) The research presented by several of the researchers can be traced directly to previous workshops.  (3) Several non-LTER scientists participated in the workshop and expressed interest in continued participation. (4) Specific goals were set for future activities and products.  (4) A proposal was generated by Ann Krause on analysis of the LTER network.  (5) Kerry McKenna, a student of Morehead, is scheduled to visit Christian during the summer to extend her analyses.  And (6) a group for the Virginia Coast Reserve is collaborating with an Italian group on evaluating N cycling in coastal lagoons through network analysis.

There was one glaring and interesting omission from the workshop; the participation of many terrestrial ecologists and social scientists.  There appears to be a definite divide between ecologists who study aquatic ecosystems and those who study terrestrial systems in the interest in and willingness to use network analysis.  The latter group has researchers who use network analysis in their studies, and yet none came to share knowledge.  These issues would be useful to explore by the network.