I agree with John’s admonition that we should conduct ourselves with courtesy and consideration, in a collaborative spirit. There is enough acrimony in politics today that we need not introduce any more into science.
I note that Hugh deals very dismissively with my comments on the AIMS monitoring program-which themselves are by now very old. Perhaps he might have wondered from where I got the information… My comments were almost a direct quote from one of the most senior scientists at AIMS, who knows a great deal about that program. He chose to remain anonymous because he feared the sort of reaction which in fact eventuated.
Mixed in with the various elements of this thread were comments from Hugh that we need to pay more attention to the geological literature [a noble goal], as well as his dismissing some of the evidence of transport of terrigenous material out onto the GBR shelf. In fact, almost 20 years today this subject was covered in a “geological” paper that showed long-range transport of material important to corals: “terrigenous influence on diet is measured out to the edge of the continental shelf, circa 110 km offshore” [Risk et al. 1994: MEPS 106: 121-130].]]>
“Hugh is wrong about a few things too. However well trained AIMS techs are, there simply is no way to ensure visual estimations of community structure do not change over time.”
If your aim is perfection then this is true but there are ways of minimising this. Training, SOPs, independant re-checking , consistency in leadership, goals and intent and so on. All these seem to be in place at AIMS. Human error is a feature of all science, how do you know a photo from 40 years ago isn’t mis-labelled? This is a question of applying the same stringent assessment to all the data.]]>
There’s an extraordinary amount of effort that’s obviously gone into this and I’m gratefully you make it publically available.
Look I’m in total agreement with you on a couple of things
1) Your paper is not about baseline. It’s really hard for me to fathom how John comes to that conclusion. It seems to revolve around this sentance. “We argue that the GBR is currently less degraded from its natural, resilient state than some published reports have asserted.”. I read it as a critical comment on the ability of previous meta-analysis to capture the degradation process rather than a statement on the natural state per se.
2) The repeated association of your work with the tactics and techniques of dis-informing skeptics/deniers is extremely unnecessary. I read it as you did as guilty by association. Apart from that it sheds absolutely no light on the nature of your scientific dis-agreement. John chastised you for unprofession original post but I think he should try a little introspection.
3) This is essentially a disagreement about how to handle data. The shift in mean and SD in Fig 1a coupled with the knowledge that this represents a shift in data source makes it pretty obvious this is all about data handling.
It would have been interesting to see the un-edited verion of Hugh’s post because on Skepticalscience I was most critical of the approach John had taken in his article. While there is a genunine scientific disagreement here I think John too readily strays into the wider climate change debate with his article. He seems to have one eye on the overall message, is concerned how The Australian newspaper is going to twist your results. Obviously this is well outside the science but I’m interested in whether you think the wider issues are imposing constraints on what are acceptable conclusions to draw?
One final thing John you write
“But if you now concede that “our paper is not very relevant to baselines for the GBR.” then I think Iv’e achieved more than I could have hoped with this post.”
This really sums up the problem I have with your reading of Hugh’s work. “now concede” – the whole point is that at no time was he proposing baselines. His point seems to be that previous meta-analysis (yours included) can’t tell us anything about baseline. Rather than reflect on a perceived victory what you really should be doing is critically assessing your own work in light of the critique. I realise you’ve been doing that and it seems your way forward is to drop the AIMS data from future work. It still seems the longest, most wide ranging, consistently driven assessment of the GBR that exists. I wonder whether it sensible to take that approach?]]>
Hugh is wrong about a few things too. However well trained AIMS techs are, there simply is no way to ensure visual estimations of community structure do not change over time. The estimations are based on perception which is obviously not a static element. Hugh points out that AIMS techs spend 100 days a year at sea. Awesome! I spend 365 days a year on land, but that doesn’t mean my perceptions about things – beauty, weather, food, language – don’t change. Hugh is also wrong in arguing that it is necessary to apply such consistency tracking when using quantitative measures of coral cover, eg, “Do PhD scientists doing unrelated and often short-term projects ever compare their estimates for consistency between observers or over time?” No they don’t, because there are not any “observers”, there is a value of measurement, which is not based on observed perceptions.
Hugh makes the point that the sample size of the pre-AIMS, pre-1986 survey data (~100 reefs/surveys) is too small to accurately describe the state of the GBR. Perhaps. Yet the whole thesis of Hugh’s paper is that the post-1986 surveys data somehow tell us about the GBR in earlier decades, ie, when the sample size is zero! I can accept as a reasonable opinion the argument that we don’t have enough data to reliably estimate GBR coral cover pre-1986. But to argue that data collected between 1986 and 2004 does tell us about reef state in the 1960s, 1970s and 1980s but data collected during those periods do not, seems dubious to me.
Yet all of this is, I think, largely irrelevant. What I care about – what I’m interested in- is, as the title of the post implied, coral reef baselines. What was natural and how much have reefs changed from that pre-human state.
Hugh thinks that I have misrepresented the arguments regarding the baseline in Sweatman et al 2011:
From his comments;
“I would urge people who are interested to read the paper for themselves rather than rely on John Bruno’s recent post as a summary. In particular, readers can see for themselves whether we state, or even imply, that the GBR was substantially in the same pristine state as it was prior to European settlement up until AIMS surveys started. I maintain we do not and our paper is not very relevant to baselines for the GBR.”
“JB: We argue that the GBR is currently less degraded from its natural, resilient state than some published reports have asserted”. Global climate change skeptics have frequently use a very similar approach: they rationalize cherry picking a favored data set and time interval in an attempt to show land and ocean temperatures haven’t increased, that sea ice hasn’t declined, etc.”
“Hey, hang on a minute! The sentence in boldface is a quote from the paper and I do believe that the GBR is currently less degraded from its natural, resilient state than some published reports (including one by John Bruno) have asserted, but that is not the same as stating that the GBR has changed little if at all due to human influences. There is no such statement anywhere in the paper by Sweatman et al. (while a number of anthropogenic changes that have affected the GBR are listed).”
Maybe. I will paste the relevant passages from Sweatman et al 2011 below and ask Hugh to explain what he means by “We argue that the GBR is currently less degraded from its natural, resilient state than some published reports have asserted.”
If the data and discussion in Sweatman et al is as Hugh says “not very relevant to baselines for the GBR” how did he come to this conclusion? What does he think was natural? How much has the GBR changed? There must be a secret AIMS estimate somewhere, otherwise how could Sweatman et al 2011 have determined that previous estimates are inflated? I’ve said what I think based on the evidence at hand. What does Hugh think?
“We argue that much of the apparent long-term decrease results from combining data from selective, sparse, small-scale studies before 1986 with data from both small-scale studies and large-scale monitoring surveys after that date. The GBR has clearly been changed by human activities and live coral cover has declined overall, but losses of coral in the past 40–50 years have probably been overestimated.”
“Subheadings in the Discussion: Long-term changes in coral cover on the GBR and How much is the GBR degraded?”
“Whatever levels of coral cover existed prior to 1986, the homogeneous data series from the long-term monitoring programme shows firstly that the average cover on the perimeters of reefs across the GBR declined from 28.1 to 21.7% between 1986 and 2004 and secondly that this has been mainly due to large declines in some subregions rather than a consistent, system-wide decline.”
“We argue that the GBR is currently less degraded from its natural, resilient state than some published reports have asserted.”]]>
In his post, John Bruno ultimately accepts that “… the drop in GBR coral cover in the mid-1980s could be due to a methodological bias of the manta tow technique.” I think this is a methodological issue but really concerns meta-analyses.
In his post, John Bruno refers more than once to the perception by Sweatman et al. that AIMS monitoring data are superior to those from other sources (which we think should be ignored). Neither of these points is true. I do not think it is the quality of the available data that is the main issue, it is the use of the data in some recent studies. Here is a “cherry picked” example: Bellwood et al. (2004) inexplicably did not give any details of their data sources or analyses, other than that the data are from published sources. Their Figure 1a (the first figure in John Bruno’s post) shows “coral cover on the GBR” plotted against time. The earliest data point in the plot appears to be from 1963. Reef ecology is a young science and among published studies, that data-point can only be from Joe Connell’s study at Heron Island. Joe Connell is a most distinguished ecologist, his studies at Heron Island Reef have been invaluable and most influential as long-term studies of the dynamics of coral communities. He started the study on a small scale and his initial plots that were surveyed in 1962 and 1963 amounted to 11 m2 of reef area (Connell et al. 1997). Small areas may be appropriate for careful studies of community dynamics, but one does not need a PhD to grasp that 11 m2 from just one reef in the far south of the Great Barrier Reef (GBR) province is unlikely to provide a reliable estimate of average coral cover on the nearly 3,000 reefs spread over 335,000 km2 of the GBR in that year. The relative emphasis given to that point in any analysis of changes in coral cover clearly needs to be adjusted accordingly.
In my view, the drop in coral cover over 50-odd years has not been overestimated because of an inadequacy of the manta tow technique, the overestimate came about because a large body of manta tow data (with its associated strengths and weaknesses) was mixed with earlier estimates of coral cover by other methods into over-simplistic meta-analyses. In the un-cut version of my comment, I referred to a quip from an editorial on meta-analyses in the medical literature (DeMaria 2008) which John Bruno deleted, along with this whole section. I found it amusing, but it may offend some; readers can form their own opinions. Dr De Maria was commenting on the rapidly increasing number of studies using meta-analyses (compilations of results from several separate studies to address a question) in that field, and ruminating on some strengths and weaknesses of such analyses (- just one of many such discussions in the literature). One of the commonly listed traps for the unwary in meta-analyses is mixing up of quite different types of data (especially if these are confounded in time). Graphs like Figure 1a of Bellwood et al. would make any alert person ask “What on earth happened to the coral on the GBR in 1986?” Such a graph should ring very loud alarm bells in the minds of meta-analysts (and peer-reviewers).
I am not contemptuous of meta-analyses, but I freely confess I am a sceptic, and the case in point is an example of why.]]>
“In particular, readers can see for themselves whether we state, or even imply, that the GBR was substantially in the same pristine state as it was prior to European settlement up until AIMS surveys started. I maintain we do not and our paper is not very relevant to baselines for the GBR.”
If that is now the case, I think we’ve got no disagreement. What led me to the belief that you were, was statements like this: “We argue that much of the apparent long-term decrease results from combining data from selective, sparse, small-scale studies before 1986 with data from both small-scale studies and large-scale monitoring surveys after that date. The GBR has clearly been changed by human activities and live coral cover has declined overall, but losses of coral in the past 40–50 years have probably been overestimated.”
But if you now concede that “our paper is not very relevant to baselines for the GBR.” then I think Iv’e achieved more than I could have hoped with this post.
Although I’m still curious about your views about the GBR baseline, eg, your agreeing with my point that; “Sweatman et al are effectively arguing that the GBR has naturally lower coral cover (averaging a mere 28%) than the Caribbean, which lacks plating species; even in dense thickets of branching Acroporids (as seen above), Caribbean coral cover rarely exceeds 70% whereas on the GBR, cover can easily reach 100% (see the photo below).”
And I’m puzzled by your new found contempt for meta-analysis; you seemed supportive when invited to coauthor papers using them, eg. Bruno et al 2007 and Bruno at al 2009.]]>