Global analysis of seagrass restoration

In this post Marieke van Katwijk discusses her recent paper ‘Global analysis of seagrass restoration: the importance of large‐scale planting

Have you seen a lot of your restorations fail? Not immediately… but in the end? Repeatedly? Could you not find a consistent environmental correlation to explain it, nor a technique to overcome it? It happened to me. I must admit I felt stupid. I am not sure even why I continued to study restoration ecology. It must have been a Don Quichot-like optimism, or an inclination to not take myself too seriously as a scientist (a cultural thing), or stubbornness. Either way, I had to face the complex reality. My struggle culminates in a global analysis that I now published with the help of 25 co-authors from 12 different nations for Journal of Applied Ecology. How did it get there?

First, what do plants, in my case seagrasses, need in order to grow? And how do we transplant them? In the 1990s, we started with basic ecology and basic development of techniques. While planting, testing and experimenting, I noticed that plants had feedbacks on their environment. They are ecosystem engineers. In 2001, the paper of Scheffer and co-workers about alternative stable states as a result of strong feedback greatly appealed to me, and helped me to accept our loss of plants. I was convinced that, once we restored positive, self-sustaining feedbacks, restoration would become more feasible. My manuscript on this ‘conviction’ in 2001 was encouraged in a friendly way but also rightfully considered naive and ended in a drawer.

Starting in the 2000s and up to the present day, our team substantiated several feedback mechanisms in seagrass ecosystems, related to among others turbidity, water and sediment dynamics and toxicity. Following observations of unexplained losses and increasing evidence of self-sustaining feedback, the Dutch environmental managers encouraged: (i) spreading of risks, allowing for replication and treatments, thus allowing for the generation of knowledge, as well as (ii) crossing a critical mass in restoration projects. For example, in 2007, a massive seagrass transplantation started as a mitigation for dike reinforcements in The Netherlands. We recently completed the project. We had success at 2 out of 6 tidal flats, we measured ‘everything’ and yet, the title of the paper starts with “Unpredictability in seagrass restoration” (Suykerbuyk et al. Journal of Applied Ecology).

Unpredictability is discomforting to the scientist. It is even more discomforting to environmental managers, usually having short policy horizons (2–4 years). Unpredictability had been haunting me for more than 20 years, and I needed more help. At the first European Seagrass Restoration Workshop in 2010, Nuria Marbà generously granted me the lead in a worldwide restoration review on the basis of the spreadsheet she had developed together with Alexandra Cunha. Anitra Thorhaug was prepared to share her life work of seagrass transplantations dating back to the seventies, as were co-authors of US, Korea, Japan, Indonesia, Canada, China, Australia and several European countries. Bob Orth, Carlos Duarte and Gary Kendrick contributed their huge seagrass expertise and two students helped analysing 215 sources. How was this huge dataset going to be of help? Firstly, I learnt that we, in The Netherlands, were certainly not the only ones that had poor results (estimated 63% trial loss worldwide). Secondly, I learnt that some restorations are very successful!

What did we learn for the future? Of course we extracted best practices, but still, these are largely local and did not give major surprises – although our novel finding that anchors should best be weighted appealed to me: it seemed so logical! Of course water dynamics are likely to dislodge the light anchors in the end! A nice anecdote for this comes from co-author Chris Pickerell of the United States. He told me how he had to leave a transplantation in a hurry because of an upcoming storm. He and his diving team hastily buried the remaining plants under whichever stones they could find on the sea bottom – and found these plants were the most successful of all…!

What could we learn from the global data at a more conceptual level? From my experience with the positive feedback and unpredictability, it would be logical to assume that a big scale would favour success, and I was delighted to see this hypothesis confirmed by the dataset! But how does it exactly work? I tried to explain my ideas to co-author Kate O’Brien who combines an attentive ear with the mind of a top engineer. She frantically drew my complex story into a simple scheme. It was like magic. There it was:

figure 1
Figure 1. Framework depicting the synergy to investing in spatial extent and planting density, and the trade-off, given a high but limited number of plants, to invest relatively more in either spatial extent or in planting density. A large investment in high numbers may be needed for best restoration practice in dynamic systems to capture windows of opportunity generated by spatial heterogeneity (horizontal axis: spreading of risks, or spatial extent of planting, m2) and to reach threshold required to initiate self-sustaining feedback (vertical axis: recovery of feedback, or planting density, m-2). Knowledge of the local environment is essential to choose the best planting strategy. (Picture courtesy clockwise: A. Meinesz, R.J. Orth, C. Durance, A. Bos).

Of course, the horizontal axis, spreading of risks, applies not only to spatial extent, but likewise to temporal extent related to windows of opportunity that need to be captured. From this figure, it was a simple step to figure 2, which I depict below. It bore a similarity to my first drawing in 2001; I must admit that it was very satisfactory to see my scientific quest had come full circle.

figure 2

None of us like unpredictability, and spreading of risks seems like a waste. Perhaps it comforts us to realise that nature itself also works with spreading of risks in many ways. We could perhaps even consider that as the basis of biodiversity. I hope that our framework will help to accept the inevitable unpredictability. Or, as my PhD-student Suykerbuyk and coworkers formulate it in a forthcoming paper (Suykerbuyk et al. Journal of Applied Ecology): “Environmental managers can improve transplantation success by restoring the positive feedback, reducing stress, but also via risk spreading by performing transplants over wider areas. They thereby accept the complexity of processes and unpredictable temporal and spatial variation in which transplantation sites turn out to be successful.”

3 thoughts on “Global analysis of seagrass restoration

  1. The important point of the review in my mind was to handle Global seagrass revegetation data objectively again after a quarter of century of biased reviews. This required a broad array of practitioners analyzing much global data statistically. For 25 years, many reviews have been biased on anecdotal not scientific statistical analysis of comprehensive data. Perhaps the reason was to push one or another governments’ point of view. This level of science review does little for the practitioners. Thus, seagrass restoration suffered terribly compared to mangroves or to marshes or to Coral , which other revegetations received huge funding and made fine progress. The reviewing is an important task which deserves detailed scientific attention. This same phenomena of having an ” out of favor” biological group has been observed in other reviews.

    Many of the important physiological criteria in this seagrass review were not quantified because the original matrix did not include them or that individual projects did not adequately record them. Exact depth, ambient light, and periods of turbidity being a few. Others were background of site ( always barren, scrape down, polluted and now barren) and species adaptability to the present anthropogenic circumstances. Far too little animal recolonization studies were done and we reported none, and little chemistry or sediment morphology or chemistry in monitoring even immediately within the study period. It was not clear , whether a neutral and trained third party minimally observed the monitorings or remonitored to confirm the data. This would have been useful. Personally, the long term monitoring is very useful for future studies and little excuse is possible not to carry out biannual monitoring at a macro-level of past revegetative plantings. Lack of funds are certainly not an excuse.

    However, with the large array of factors chosen, it was already difficult to do an analysis, so it was decided that it was much better to carry out a statistical analysis on what we had than none at all as the reviews found in Short and Coles and several others locations. The large array of data which was not included had many successful projects among them and might have tipped the balance since they tended to be larger restorations. Especially in the USA, England, Sweden, and a few other global areas. So that there is additional objective data.

    So the data could have been more complete, but time pressured us and we needed to make a statement for especially newly emerging governments to decide whether to require seagrass mitigation as they had promised under RAMSAR.
    One should keep in mind the large loss at 7% per year ( Waycott et al 2006) shows that despite the 40 km2 restored that globally the governments and scientists are not keeping up with the losses, so that more, much more, is required to reach seagrass sustainability.
    Thus, the four and a half decades struggle to have seagrass losses replaced continues. Let us hope the increased carbon within restored seagrass will assist in bringing more restoration for seagrass carbon sequestration values. ( Thorhaug et al, 2017)

    Anitra Thorhaug, Ph.D.
    Sustainable Forestry Physiological Laboratories
    School of Forestry and Environmental Studies
    Yale University


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s