Damaging Science and Ecology: How The Australian Inflated Avian Impacts by 886%

Attacking science isn't easy. Simply burying your head in the sand isn't nearly enough - you need to push back against scientific research. Delete the evidence, and write your own narrative.

The best recipe for this is the production of a thin, barely-visible veneer of scientific credibility. Bury this in whatever scientific uncertainty you can find. Stir.

Yesterday, the environment editor for The Australian, Graham Lloyd, added another article to a vast catalogue of anti-wind pieces covering wind turbine syndrome, wind farm output, wind farm economics,  aesthetics and of course - the impact of wind turbines on bird life. 

Lloyd's 'exclusive' on avian impacts at AGL's Waterloo Wind Farm was obtained from a mechanical engineer named Hamish Cumming - the two have history. In September 2012, Lloyd published an article featuring an unfortunate but incredibly significant misunderstanding. Cumming, a trained engineer, assumed that electricity produced by wind farms is somehow 'spilt' into the aether during times of low demand and high wind, as coal-fired power stations in Victoria don't alter their output downwards.

It's a very, very, very big misunderstanding of the National Electricity Market, but all that was needed was the veneer of scientific credibility. So it is with Lloyd's freshest offering: 

"EAGLES, falcons and other raptors make up to a third of the estimated 1500 birds killed each year at Australia’s biggest wind farm. 

The finding of an independent report for Macarthur Wind Farm operator AGL follows 12 monthly searches of 48 turbines at the 140-turbine operation in Victoria that found 576 bird carcasses."

An anti-wind blog published a scanned copy of the report, covered in (presumably) Cumming's hand-written notes.

So did the scientists find 1,500 dead birds at the Macarthur wind farm? They must have - Lloyd claims that 1,500 birds were killed in the space of a year by the machines. The report states: 

"A total of 65 individual birds from 15 species and six bats from three species were found during carcass searches"
- Australian Ecological Research Services

Lloyd breathlessly declares that 1,500 birds are killed each year, and that 576 bird carcasses were found at the wind farm. So, what's going on? Why the discrepancies? First of all, Lloyd has misread a quote in the report: 

"A total of 576 carcass searches (12 searches of 48 turbines) were conducted over the subsequent 12 months from March 2013 to February 2014"
- Australian Ecological Research Services

Yep. That's right. 576 is the number of searches undertaken; not the total number of carcasses found during those searches (65). It's okay. That's only an inflation of 886%. Close enough.

But, what about the 1,500 figure declared in the byline of the piece? That's slightly more complex, and it betrays a distinct attitude towards scientific uncertainty and estimation.

The report relies primarily on 'correction factors' - they multiply the number of carcasses they find by a certain value, to adjust for the fact that carcasses may be removed by predators before the researchers had a chance to find them (they went out once a month to look for deceased birds): 

Australian Ecological Research Services (AERS) euthanised a collection of turkeys obtained from a turkey farm in Victoria, and used these as controls, to determine how long it takes for a carcass to be removed. As AERS state: 

"Estimates of bird and bat mortality are subject to several sources of bias which may result in inaccurate estimates. Such sources of bias include the use of correction factors for searcher efficiency and scavenging rates which are ineffective if no fatalities are found at a turbine due to prior removal from scavengers.  

However, the likeliest source of error in the current estimates of mortality at the Macarthur Wind Farm is the search interval between consecutive carcass searches Each turbine was searched approximately 30 days apart but as illustrated by the scavenger trials, most carcasses are removed by scavengers within one week This results in fewer carcasses being detected and of those that are detected, a very high scavenging correction factor is applied. As such, the estimate of mortality is likely to be inaccurate as it relies primarily on correction factors rather than actual fatalities"

The estimates used by Lloyd and Cumming to headline their piece are just that: "very high" factors to adjust for the fact that the data were acquired at a low resolution (once a month) rather than at a higher resolution (once a week).

They multiply the '10.19' per turbine figure with 140 turbines to get 1,426, and automagically add on another 74 to reach that wholesome round number of 1,500. Fits the narrative a bit better, if it's a bit bigger. 

The scientists do their best to generate approximations in the face of uncertainty, and they're open about the limitations of their estimates in the study. Do Lloyd and Cumming declare the same information in their article? Not really: 

"But an AGL spokesman said the report had “shown no significant impact on threatened ­species”. The company said overall ­estimates of bird and bat mortality “are subject to several sources of bias which may result in inaccurate estimates”" 

What? 'The company' said that? Those words come directly from the report. Lloyd is making out that the company is begrudgingly criticising the findings of the scientists, when really, they're simply quoting the uncertainty declared in the report, as Lloyd has spectacularly failed to do in his article. 

Another of Cumming's multitudinous angles of attack against wind energy is the survival of the brolga: 

"All of Australia’s GHG is now less than 1% of the world’s emissions, wind farms in Australia will not reduce the global GHG emissions by any measurable amount. Wind farms will however displace Brolga from their habitat, threatening their survival"

AGL commissioned a separate environmental firm, Biosis, to review the veracity of their Macarthur report: 

"Biosis found that the collision mortality rates at Macarthur wind farm are ‘not high relative to other wind farms’.” Key findings of the monitoring research found no brolga deaths and the protected species had successfully bred on the wind farm and continued to return to the site"

Significantly, Cumming's focus on the preservation of the Brolga isn't included in Lloyd's article. Presumably, it distracts from the narrative, somewhat. A science communication writer wrote on uncertainty at The Conversation recently: "Genuine researchers are those rare individuals who have come to terms with their uncertainty and confront it on a daily basis". This is true; but I'd append a further feature of uncertainty.Those who seek to erode research and damage science use uncertainty as a blank substrate for fear and anger.

As an avian ecologist pointed out on ABC Radio National's Ockham's Razor: 

"Wind farms are one of the few sources of impacts to birds and bats that are being systematically monitored. Most other forms of energy generation do not monitor impacts. Nor do we have widespread systematic monitoring of all the other human-related activities that we inflict on species – such as collisions with cars, powerlines, windows, poisoning, shooting, pollution etc."

By presenting estimates of bird mortality without any context, such as monitoring conducted at a coal-fired power station or a gas-fired power station, Lloyd and Cumming together work to damage efforts by scientists to directly address the issue of avian impact.

They're making it harder for scientists to publish work without it being co-opted and misquoted and tarred with the brush of anti-wind groups. Uncertainty and estimation ends up under the wheels of the narrative. 

There is already some data on comparisons between generation technologies and avian impacts: 

"Environmentalists and environmental scientists have criticized wind energy in various forums for its negative impacts on wildlife, especially birds. This article highlights that nuclear power and fossil-fuelled power systems have a host of environmental and wildlife costs as well, particularly for birds. Therefore, as a low-emission, low-pollution energy source, the wider use of wind energy can save wildlife and birds as it displaces these more harmful sources of electricity"

Source + You can read a response to this paper here

So why won't you see any articles about the impact of Australia's coal-fired power stations on bird life? Well, fossil fuel companies aren't required to do any monitoring or reporting, despite the likelihood their impact on wildlife is many times greater than that of wind energy.

This is one field of uncertainty very likely to remain untouched by The Australian, Graham Lloyd, and Hamish Cumming. Can't mess with the narrative, can we?

Update 24/09/2014

Found this report from the Australian Transport Safety Bureau - provides information about bird strikes from aircraft: