Building for Nature: Preserving Threatened Bird Habitat in Port Design

Round 1

Reviewer 1 Report

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

This is an interesting contribution showing possible compromise solutions to the planned development of harbour infrastructure at Southern Yellow Sea coast which is linked to substantial habitat loss for migrating shorebirds, including species severely threatened and crucially dependent on the tidal mudflats to be developed. Such solutions, backed by reliable research, are badly needed if we are to successfully protect threatened shorebird populations migrating along East Asian-Australasian Flyway (EAAF). Authors use extensive hydrodynamic simulations to assess whether alternative variants of harbour development may result in habitat changes that might be less detrimental/more favourable to two focal bird species. This makes the MS a truly welcomed and timely contribution to the ongoing discussions on conservation of millions of shorebirds migrating along EAAF.

However, I am concerned by the way the analyses of bird preferences are presented, by restricted reporting of the whole habitat changes projected to occur and by equating predicted presence of potentially suitable habitat with its actual suitability for shorebirds. These points are discussed below.

(1) I don’t understand the idea to compare use vs availability of ecotopes only within 90% home-range of two focal bird species (chapter 5.1 and Fig. 9). What does it intend to show? Fig. 9 based on this set-up generally show us that – within this domain -  birds do not show ecotope preferences and use them in proportions roughly mirroring their availability. However, that is totally misleading picture – if we do not restrict availability to areas actually used by birds (90% home range), we can see very clearly that birds do avoid blue ecotopes – shallow sublittoral and low-range littoral, and prefer mid-range littoral and high-range littoral areas. Please explain very clearly the rationale behind applying use/availability ratios only to areas already used by birds. Or, preferably, give it up, i.e. do not restrict this comparison to 90% home-ranges and use the entire site as denominator in analysing habitat preferences. Additionally, due to the way it is calculated, a substantial part of the 90% home range is effectively empty of birds, particularly for Bar-tailed Godwit, making the situation still worse. I believe, the scale of this analysis is inappropriate and should instead focus on the level of the entire area, as indicated in line 243. Such an analysis would be also much more revealing.

(2) Calculations shown in chapter 5.2 relate only to projected changes in area of focal habitats (ecotopes). However, to have a full, meaningful picture, we need to know – for each habitat - what are the total areas at the start of simulations (reference point) and what are the areas to be reclaimed and lost to harbour infrastructure. It is also unclear if figures presented reflects net gains/loses for each habitat or raw gains, without accounting for losses occurring in other parts of the study area (due to direct loss to infrastructure or indirect loss due to temporal turnover of habitats). Obviously, it is difficult to assess the extent of habitat compensation occuring without knowledge, how raw gains relate to losses, and how net gains relate to the amount of habitat available initially. I suggest to provide a table with a full accounting of starting values, gains and losses for every habitat (ecotope), broken by each harbour variant. This may be supplementary on-line material if space is a problem.

(3) The main reasoning in the MS is that shorebirds prefer some sites because of a favourable combination of sediment morphology and hydrodynamic proporties (=ecotopes). Thus, providing new sites with appropriate combination of these abiotic factors (sediment morphology+hydrodynamic features) would safeguard their needs and enable protection of their populations. Yet, that is an oversimplification, which should be (and is not) discussed in the text. Shorebirds use mudflat sites with particular abiotic features mainly because they are good foraging sites, supporting high densities of preferred prey. Thus, ecotopes (defined on basis of sediment and hydrodynamic features) serve only as a proxy for bird habitat suitability. Importantly, prey densities in suitable habitat should be high enough to make foraging efficient. I.e. birds would leave some sites even despite presence of preferred prey, when prey densities are not high enough, yielding too low energy gain per unit time of foraging (so called “giving-up” prey densities). This is particularly important for birds refueling before making another long, non-stop migratory flight - if the refueling rate is not high enough, they will not reach Arctic breeding grounds in time. Moreover, bird's preferences for particular sites were shaped also by predation danger, competition with other species, favourable microclimate (e.g. shelter from the wind) and a number of other factors. Thus, looking only on abiotic habitat features, while perfectly understandable, is only half of the story and do not fully measure chances for real success, i.e. providing habitats that actually will be used by foraging birds.

Here, we don’t know if prey stocks will easily colonize these newly established sites deemed good for birds, how much time it will take, whether and when prey populations reach high densities here (above the giving-up threshold for birds), and what the birds are supposed to do during this transition time. In short, we do not know anything about prey response to the appearence of new patches of apparently preferred ecotopes, and time-lag for this reponse. We don't know if the newly established patches will provide safety from avian predators, or human disturbance. All these and similar caveats should be addressed in the discussion chapter, in reference to ample literature on foraging ecology (incl. optimal foraging).

Furthermore, I have several minor points listed below.

(4) The mudflats of Yellow Sea are truly crucial for the survival of many migratory shorebird populations. It might be good to make this point even more clear by providing the information that it is the main refuelling site for shorebirds that fly non-stop some 5,000-9,000 km from Australia or New Zealand to reach the Yellow Sea and then make the second, comparably long, non-stop flight to Alaska or Chukotka. This information is visual and clearly shows how important the site really is. Please consider including this information somewhere in the introductory text.

(5) Figure 1. I do not think the truly research paper is the right place for such figures. It shows neither methods nor results and its place is rather in PR materials or in presentation to the wide public. When I see pictures of this kind, I feel I am far from science and close to politics and propaganda. Please re-consider the necessity of including this figure in the paper, as, for me, it undermines scientific image of your paper.

(6) Table 1. I can't find explanation for MHWN acronym. MHW explained in L167 is not the same as MHWN.

(7) L229-231. How many individual birds (per species)? How many bird-days used? How many data points in total? Please provide the very basic data on the material used to calculate bird habitat preferences.

(8) Fig. 5. There is no correspondence between names of variables on the plot axes and in the figure legend – “Speed” on plots vs “velocity” in legend. Also “Current direction” in the legend should be replaced by “Modeled direction”

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

This is a very interesting and inter-disciplinary manuscript about a topic of ecosystem conservation relevance. I read it with interest. I am an animal ecologist with a background in bird ecology and conservation: consequently I can provide comments only on these topics. Other sections on engineering and climatic (meteo-marine, morphodynamics) arenas need to a further reading carry out by an other peer-reviewer.

Language is clear. Logic is correct. Figures are readable and complete (overall Figs. 5, 6 and 7!). Statistical analyses are strong. I have only minor comments to add. Therefore I think that this ms deserves to be published with MINOR REVISIONS. Here below some points that, I hope, could a bit improve the ms.

COMMENTS

Everywhere on the text you used the term ‘threatened’. Nevertheless I don’t read never the term ‘(anthropogenic) threat’ and its meaning. I suggest to add this concept in the Introduction and some references (e.g. Salafsky et al., 2008 on Conserv. Biol. and the Springer book ‘An introduction on disturbance ecology’: https://link.springer.com/book/10.1007/978-3-319-32476-0).

43. ‘.. is China, (net growth of 6.7% in 2016’. Delete comma.

77-79. Which is the rationale for this sentence? ‘We expand on this approach by combining hydro- and morphodynamic modelling and unique distributional data on two endangered shorebirds, the bar-tailed godwit Limosa lapponica and great knot Calidris tenuirostris.’ Why you used these two species? Are they indicators? Or they are only rare and endangered species? (A rare species could be also not an indicator). There is literature about it? In rows 189-195 you reported ecological traits of the species but I think that there are many others waders having similar ecological preferences (rows 197-199). Why you select these species?

526: the proposal ‘to create species-specific ‘ecotope’/ ‘species-tope’ is interesting. See also the literature about ecological suitability models (at least a reference should be provided here).

Add the role of anonymous reviewers (in acknowledgments).

Check for the Editorial style of the Journal.

Have a nice work.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

I am happy with the changes introduced to the revised version of the MS. In particular, I am glad to see extensions to the Discussion that take into account that providing suitable habitat defined in terms of abiotic elements is only a first step, that does not necessarily means that birds will use these newly established patches.  

However, I am still uneasy with the way the bird habitat preferences are calculated. Currently, the main analysis relies on comparing frequencies of different ecotopes for 90% home range (90HR) to the frequencies in the entire area. This is shown on Fig. 8. Although no formal analysis was made (i.e. standard errors or confidence intervals for percentages were not calculated, hence differences in frequencies were not tested), this comparison generally shows no big differences for percentages within 90HR vs percentages within entire area (except perhaps for “reclaimed” (both species), “mid-range low-dyn” (BTGO), “shallow sub high-dyn” (GRKN), “deep sub high-dyn” (BTGO)). Anyway, this analysis – albeit generally correct – apparently does not capture the fact that both bird species use areas close to the mainland and completely ignore vast areas in the NE part of the bed, further from the mainland. In other words, comparisons like yours, do ignore the fact that localisation of 90HR within the entire study area, reflects another, very basic and important, level of habitat choice. This issue is acknowledged in a standard textbook on resource selection (Manly et al. 2002, Resource Selection by Animals: Statistical Design and Analysis for Field Studies; 2nd edition; Kluwer Academic Publishers), where they describe your analytical approach as follows:

Individual animals are identified and their home range determined. The proportions of resource types in home ranges (as determined by sampling or the partitioning of maps) are compared to the proportions in the entire study area (…). This approach has been criticized by White and Garrott (1990, p. 201) on the grounds that the home range represents a prior selection of habitat.  

I am not sure, how to treat this level of habitat choice analytically. One possible solution is, perhaps, to divide all existing ecotopes into two categories: close and far from the mainland, based on some arbitrary threshold/buffer (say, 10 km) and repeat the current analysis (this time using formal tests for differences in percentages, please). This should better capture the avoidance of mudflats located far away from the mainland.

In any case, I feel the MS will benefit if this pattern (avoidance of ecotopes far from the mainland) will be more fully acknowledged in the discussion section, alongside with other factors already mentioned in the revised version. That is important, as it adds another layer of uncertainty to the predicted consequences of alternative port variants, since birds may avoid patches of newly created suitable ecotopes if they emerge in places that will deem them unattractive for some reasons other than already mentioned in the discussion. And, possibly, we can see this right now, with the information we have, but left not analysed.  

Minor points are listed below.

L48 “net growth” of what? Human population? I believe it is growth of GDP, and in any case it should be specified.

L149 “morphology” of what? Bird morphology? Or, perhaps geomorphology?

L249 remove one comma.

L311 “Section 0”??

L333 plural for “datum” is “data” rather than “datums”

L351 delete “from”

L535-536 “Therefore, a ratio >1 reflects that birds prefer to utilize those ecotopes but enough area is not available to them”. I am not really sure, if the used/available ratio >1 can be used to imply that area of the preferred habitat is “to small” for birds. It is about preferences not about sufficiency. So, I would suggest delete or change accordingly the latter half of the sentence.

L692 “our recommended port design would be the first step towards conservation” – consider adding word “only” after “would be”

L711 “how much preferable habitats of a species is loss/gained” – replace “loss” with “lost”

Hectares should be abbreviated as “ha” rather than “hec.”, I guess.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf