120 



The "matched" or "paired t-test" design is of particular value because even in situations of large scale variability,it 

 is still effective since comparisons are made within a transect originating from the individual structures. This 

 design may fall under criticism, however, because it looks in one direction only. On the other hand, any design 

 which would necessitate a more concentrated effort would result in a reduction in the number of platforms 

 examined; the classic concern of site specific verses regional studies. 



While the "matched" or "paired t-test" design was the strategy of choice, several of the workshop participants felt, 

 for one reason or another, that several platforms need to be examined. The sampling density will continue to 

 be an issue and the literature needs to be thoroughly reviewed to determine the most effective and economical 

 sampling density. From the depositional standpoint, some participants felt that the local current regime may 

 result in an irregular and/or heterogeneous deposition of material around a production facility. However, this 

 may only really apply to the short-term; over extended periods, the distribution may be more homogeneous. 



No matter which plan is used, some regional considerations will have to be considered. For example, eastern 

 areas are generally associated with oil production and thus more likely to have impacts associated with produced 

 waters, while areas to the west are generally gas producers typically not associated with produced waters. 



The issue of sampling/study design may be summarized by the distinction of two approaches: 



• multiple platforms with minimal number of sampling sites, or 



• minimal platforms with maximum sampling density (the "bulls eye" approach). 



Time Frames 



The actual times at which any studies will be conducted will be important because of such phenomena as: 

 seasonality of reproduction and recruitment (typically spring), late summer storms, hypoxic events, etc. 

 Generally, the best time to sample will be in July and August— just after intense periods of reproduction and just 

 prior to late summer, early fall storms. 



3.2.6 General Management Concerns, Concepts, and Conclusions 



During the workshop several points were discussed which fit under the category of "general concepts." It was 

 felt that these were of a nature so as to be listed again separately and to serve as conclusions for the session. 



• In order to maintain a consistent focus, the program should be based around a central core; however, 

 it must also be sufficiently flexible so that tcchniques/designs/parameters could be deleted and added 

 as required. 



Screening surveys must be incorporated into the overall plan and conducted before a definitive studies 

 plan is chosen. 



Innovative, state-of-the-art techniques must be emphasized so as to avoid being locked into "standard 

 methods" which may be incapable of providing the pertinent information. 



There must be close coordination of all data collection, particularly regarding stations and sampling 

 techniques. 



A data quality objective (QA/QC) must be defined at the onset of the program. 



The program must utilized the appropriate standards and allow for over sampling. It is far cheaper to 

 store samples than to resample. 



Before any real intensive sampling is conducted, the chosen sights must be groundtruthed. 



Coordinate all efforts in regards to data analysis and data storage. 



