Michael, Minhan and Paul:
    Having just completed, with 6 others, preparation of a proposal for the synthesis phase of the northeast Pacific GLOBEC program, I can say from first-hand experience that 'synthesis' means different things to different people and the scientific community is generally unsure of what 'synthesis' is. Here are some guidelines for our symposium.
    As George Wong has pointed out, our synthesis cannot be a cross-system comparison because the talks are not directly parallel. The science and data for such direct comparisons probably don’t exist anyway. We do not believe we should try to achieve a complete descriptive comparison of the properties and characteristics of our four systems. Perhaps one of our concluding points should be that some studies need to be designed to look specifically at comparing systems. Nevertheless, we do not believe we can do this for our four systems.
    Our synthesis should be focused around the theme of our symposium: “Coastal ecosystem responses to changing nutrient inputs from large rivers.” As much as possible, we need:
 1) Historical data on nutrient inputs How have inputs (concentration, nutrient ratios, etc.) changed over the decades. We could also try to trace why they changed over these several decades. In addition we need to examine as much as possible, long-term data sets of what the ecosystem was like in the beginning and then show how it has changed.
 2) Data on ecosystem responses Responses/impacts could be pelagic and benthic. Most data will probably be mainly chemical (nutrients, hypoxia, etc.) and biological (community structure), and there will be few data available for higher trophic levels such as fish. For speakers, please try and determine if such long-term data exist for your system and bring that information with you, even if you do not incorporate it into your presentation.
However, such long-term data may not exist and we should also consider a broader level of synthesis. We must draw on the presentations and everyone’s wide knowledge of other systems to answer some broader questions that are of a synthesis nature. Thus:
 3) Which factors/processes control the processing and fates of nutrients in large river-shelf systems? (from KK Liu's 1.3 question) Factors include: runoff, tides, wind mixing, coastal circulation, topography, sediment (light), nutrient loads, and top down effects. Are there others? Are the same factors of equal importance in all systems? Can we prioritize or rank these factors? Can we develop a universal list of factors/processes that control the way nutrients are utilized by coastal systems and how these change with changes in nutrient inputs?
    Sub-questions include:
    How do the processes and fates change with nutrient changes?
    How and why do the processes and fates vary among river-shelf systems?
    What are the crucial biological communities that control the material flow in the river plume-coastal water system?
    Broadly: can we identify general rules and is there enough information to make a generalized model or develop some universal hypotheses about nutrients and river-coastal ecosystems?
    For discussion leaders, please be aware of this perspective when thinking about your questions and synthesis. Bring such data sets with you to the symposium if you have them, or are aware of them. Also, your goal should be to put together an outline or framework for a synthesis paper before the symposium so please continue to circulate your thoughts among yourselves. KK – is there a chapter in the book you are preparing that might be especially relevant to this and, if so, would you be willing to circulate it to us as background material prior to the meeting?
    These are not necessarily the only issues we should address in our synthesis efforts but serve as examples of the approach we need to take. Please continue the dialog on this difficult issue.

Alan M. Shiller:
    Just to add in a couple of thoughts.........
    With regard to inputs: An interesting question is how big is the offshore input of nutrients to these systems. This, of course, may well vary by system and by season. However, this flux is very much relevant to the issue of whether a system needs a small or large increase in river inputs to change the nutrient dynamics. Additionally, in stratified coastal systems where there is a flux of salty water along the shelf bottom, what is the flux of regerated nutrients off-shelf with the bottom water and does this flux change with changes in bottom oxygen.
    With regard to ecosystem responses: How robust are the sediment-based proxies that have been utilized to gauge certain historical changes (e.g., BSi)? How does diagenesis (and changes in diagenesis due to changes in bottom oxygen) affect theses sediment-based proxies?
    With regard to factors/processes: It might be useful to ask not only what key factors/processes we don't understand, but also what new tools would be useful to bring to bear on these problems.

Michael Dagge:
    I am sure you have all noticed that it is common at meetings for speakers to use Discussion time for part of their Presentation. I urge you all to avoid this at the Xiamen symposium, and for each and all to keep your presentation to 30 min or less so that there can be a full 15 min question and discussion period following each presentation. Good luck with your preparations?

Brent McKee:
    I have enjoyed reading the abstracts and the comments that Minhan, George and KK have made-------and I've been thinking more about Synthesis. I like Mike's last guidelines and suggestions and propose that we adopt that as a starting point for all our discussion about synthesis. That is, the important point of the outline would be:
(I) Inputs
(II) Ecosystem Response
(III) Controls on Processing and Fates of Nutrients.
    George and KK, in particular, have provided more detailed structure which will help direct talks and discussions. I'd also like to suggest a few things that in some cases bring back specifics that were raised earlier and reapplies them to Mike's outline:
 (1) Our first section "Inputs" should include other parameters in addition to nutrients. Specifically, sediment and organic matter loading should be included. As many talk abstracts point out, the interaction between nutrients, organic matter and lithogenic particles make each system unique (yet perhaps controlled by similar mechanisms). I would ask speakers to consider the salient aspects of river systems that impact the character of the inputs. For example, lithology, size/elevation/slope of drainage basin, anthropogenic effects, etc.
 (2) George pointed out that the relative weighting among talks regarding the benthic and pelagic systems differs greatly. One thing that I would be interested as a discussion leader is "Is this because of the relative importance (pelagic vs. benthic) in each system, or does this reflect the disciplinary focus of those who have historically worked in each system?" As Mike points out, one important synthesis item may be that we have undersampled important aspects of some systems.
 (3) As an overlay on the outline above, I would like to add the element of time. As mentioned by others, historical time is important when framing the discussion around the idea of global change. I echo Mike's urging that we all bring to the symposium what we know about historical changes in nutrients (concentrations and ratios) and I would add parameters like sediment and water discharge, organic carbon trends etc. I would also add two other aspects of Time that may be important in a comparative sense. One is the concept of residence time within various compartments within a river system (drainage basin, floodplain, plume, seabed (prior to burial), etc.). Unique aspects of each river system means that the residence time of materials in each subsytem will vary from river to river and perhaps the processing in one compartment will govern the fate of materials in each system. The second temporal aspect is seasonal (or shorter) changes. Our understanding about some systems may be biased by when we have examined them. We should note any underrepresented/undersampled times of the year within each system. The concept of net autotrophy/heterotrophy is defined within a time frame. Do we know enough about all important sub-annual time periods (re: temperature, flow, loading etc) to make clear statements about NET metabolic status on these systems and how that may change? If not, what do we need to work on?
    This should be a very enjoyable symposium and discussion. I look forward to it!

Paul J. Harrison:
    I agree that we should focus on the title/theme of the symposium and therefore the discussion/synthesis should focus on these aspects as suggested in the title, which are:
 1) Changing nutrient inputs
    We need historical data here. How have inputs (concentration, nutrient ratios, etc.) changed over the decades. So this becomes like a multi-decadal experiment of changing nutrient inputs. We could also try to trace why they changed over these several decades.
 2) Coastal Ecosystem Responses
  - Again long term data sets would be ideal to demonstrate what the ecosystem was like in the beginning and then show how it has gradually changed.
  - The actual ecosystem responses will be mainly chemical (nutrients, hypoxia, etc.) and biological (community structure, but here will be few data available for higher trophic levels such as fish.).
  - Responses/impacts could be pelagic and benthic, however we have no benthic data for the Pearl River that I know of. So there will be some focus on the pelagic responses and food webs.
 3) Why are the responses different in different river systems?
  - This probably involves the physics of the estuary, topography, sediment load, etc.
    I expect that most speakers have not prepared their talks yet If we could give them suggestions of key aspects and key data to include in their talk, this would help our final comparisons.
    We should also ask them to bring data besides what they are going to present. This could amplify the discussion sections.
    Let's discuss the above aspects before you send out your 'guidance notes' for the speakers and synthesizers.

Michael Dagge:
    Below is a draft of a memo I would like to distribute to the discussion leaders, to give them some guidance. Can you look it over and let me know if you agree with it, or modify it if you don't. Thanks