From the Chair: Making a stronger LTER network

Network News Winter 2014, Vol. 27 No. 4

This has been both a challenging and exciting year for the Long Term Ecological Research (LTER) Network.  LTER science is stronger than ever, with numerous critically important findings at both the site and network scale.  On a more sobering note, this fall, the National Science Foundation (NSF) decided not to renew the Sevilleta LTER program.  Coincident with this decision, Scott Collins resigned as Chair of the LTER network and I moved from chair-elect to Chair.

Our immediate task as a network is to ensure that we don’t lose any more sites. To that end, we have started a series of meetings to discuss issues related to LTER proposals including conceptual models, long-term data, expectations from NSF, probation, etc.  The first of these meetings took place on November 10 at NSF in Washington and was attended by a large group of LTER scientists and NSF program officers and was extremely useful in laying the groundwork for a series of activities that will lead to a stronger LTER network.  Many thanks to the LTER scientists who travelled long distances to attend this meeting and to the NSF program officers who helped plan the meeting and then spent four hours with us.

As follow up to the November 10 discussion, we envision a larger meeting associated with the LTER mini-symposium in March 2015, a still larger meeting associated with the LTER Science Council meeting in May 2015, and a yet larger meeting at the LTER All Scientists Meeting  in September.   Our goal is to get a clear sense of what NSF expects from our sites, an understanding of the approaches that different sites have taken to meet these expectations, and a set of “best practices” that sites can use as they prepare proposals.  We hope to involve a large group of current, former and future LTER leaders in these meetings.

At the meeting on November 10, we discussed six key issues: 1) LTER expectations for sites and the network, 2) conceptual frameworks and quantitative models, 3) mid-term reviews, 4) site probation, 5) participation in network activities, and 6) transitions to new principal investigators and/or institutions.  Below I summarize our discussions and future plans for each of these topics.

1.  NSF expectations for individual LTER sites and the network.  Our discussions on November 10 illuminated several key points about this topic. First, the discussion on how the requests for proposals (RFP) for LTER sites have evolved over the decades made the important point that expectations for individual sites change over time in response to site age and the emergence of new tools and topics.  Second, we now understand that the inclusion of text about specific quantitative modeling techniques in the RFP such as “data model assimilation” are meant more as general guidelines than specific requirements.  We note that this was quite surprising to many of us as we read the RFPs very carefully and take their specifics quite seriously.  We look forward to further clarity about both the content and implications of text in the RFPs. Of particular interest is whether and how “guidelines” become more like “requirements” over time. We hope this can be an ongoing discussion so that changes in expectations and RFPs can be understood and addressed by the sites in a timely way. Many sites begin work on their proposals several years before they are due, so it is important for us to be “up to date” on any changes in expectations.

2.  Conceptual frameworks and quantitative models. There were two important aspects to this discussion.  First, the discussion about conceptual frameworks highlighted that there are many types of frameworks that have been successfully used within the LTER network. However, a consistent characteristic of the successful frameworks is that they provide a coherent and synthetic narrative of the key factors regulating ecosystem structure and function and how long-term data have been used to improve understanding of these factors. These conceptual models and frameworks, which do not necessarily have to be mathematical, need to integrate across different aspects of the research project.   And while sites do not need to completely change their conceptual models and frameworks with each proposal, they need to be clear about what they have learned and how ideas have evolved. All research components need to be carefully linked to the conceptual framework. We are looking forward to a series of activities within the network to discuss, refine, and improve our conceptual models and frameworks.

Second, our discussions were useful for distinguishing the importance of conceptual models and frameworks versus quantitative models.  We now have a better understanding of NSF interests in ensuring that the unique and powerful long-term data that we collect are used to address critically important questions in a quantitative way. We further understand that there are many ways that this can be done and that there is no specific requirement for particular approaches such as uncertainty analysis, data model assimilation, etc.  We are keen to have further communication with NSF and discussion within the network about how expectations for quantitative analyses will evolve and be implemented within the network.

3.  Mid-term reviews.  The discussion about the focus, process and follow-up to mid-term reviews was extremely useful, providing a better understanding of how these reviews now focus solely on progress made on the proposal submitted three years prior. We appreciate that this change was made to focus the attention of reviewers more on what is actually going on at the site and less on what they might think ought to be going on at the site. The discussion also highlighted how the mid-term reviews have become less important in the evaluation of subsequent renewal proposals than they used to be and the uncertainty about just how NSF and the LTER network can best make use of the mid-term review team reports and the response letter from NSF, especially given NSF’s new decision to not provide the review results to proposal panels. We look forward to further discussion of these topics over the next few months and clarification of how we can maximize the benefits of these very important reviews.

4. Site probation. Our discussions highlighted several challenges that NSF, individual sites, and the LTER network face with these unique situations. There is a clear need for a more specific “road map” for how sites can successfully emerge from probation. The network will begin an effort to develop a set of “best practices” that sites can follow to facilitate this emergence. These practices will likely involve assembly of an outside advisory team, several rounds of internal and external consultation and review, and exercises designed to facilitate “deep thinking” about conceptual frameworks and project participants.

5. Participation in network activities.  Our discussion highlighted two distinct classes of network activities: cross-site research and participation in non-research related network activities such as governance, information management, and outreach. It is clear that cross-site activities that produce research advances are quite valuable, but the importance of non-research related activities is not as clear and is harder to evaluate. The network will begin an effort to develop a set of “expectations” for participation in these activities, and criteria for evaluating how sites meet these expectations. We do not expect that this will be easy as there is quite a bit of variation within the LTER network in these activities, and concerns about “unfunded mandates”, etc.  Still, LTER is viewed as a network by the scientific community and there is a need for participation in and credit for network activities as a significant “broader impact.”

6. Transitions to new principal investigators and/or institutions.  The meeting was very useful for clarifying the procedures that must be followed for changing lead principal investigators (PIs)and/or lead institutions for LTER projects. The current LTER solicitations are open to renewal proposals only. This means that NSF can accept proposals only from current PIs and current awardee institutions. Thus any changes need to be made well before a renewal proposal is due. To change PIs or the awarded institution on an award requires approval from NSF. To transfer a current LTER award to a new institution requires a formal request to NSF that is made after the old and the new institution have decided on the amount to be transferred and the new institution has generated a budget that fits this amount. The new budget will take into consideration changes in indirect costs that the new institution may need or want to incur (e.g., on sub-awards). NSF considers carefully the ability of institutions and individuals to handle projects as large and complex as LTER projects are.  More fundamentally, we recognize that these transitions are complex and difficult and that there have been many different successful (and not so successful) transition models. The network will initiate a series of discussions of different models for site leadership and institutional transitions.

The challenges of this year remind us that we need to redouble our efforts to write strong proposals, rigorously prepare for site reviews and exploit the unique opportunities provided by the LTER program. On a more positive note, NSF has assured us that it remains firmly committed to a strong and full LTER network. Indeed, there is active discussion within the agency about a request for proposals to add a new site (hopefully arid land or grassland) to the network to replace the losses of Shortgrass Steppe and Sevilleta. Stay tuned as specific plans for follow-up activities develop over the next few months.