Sunday, 15 September 2019

Linking SABAP2 reporting rates to bird density estimates

Reporting rate is the most commonly used abundance measure derived from SABAP2 data and reflects how many times a species appears per pentad; or set of pentads. Intuitively, species with high reporting rates should be more abundant, i.e. have a higher density, measured as the number of individuals per unit area. But reporting rate might also be influenced by differences in the ease with which species are detected. Factors that might affect detection rate in addition to abundance include bird size, sentinel and vocal behaviour, as well as the habitat in which a species occurs.  

For example, density estimates of Cape Rockjumper are between 1–5 individuals per km2 across their range, Cape White-eye occur at a density of 40–50 individuals/km2 in Fynbos (Lee & Barnard 2017, Ostrich 88: 9-17), while Lark-like Bunting occurs at a density of 20-500 individuals/km2 (Lee et al. 2018, Ostrich 89: 363-372). By comparison, SABAP2 reporting rates are 5–15% for Cape Rockjumper, 50–60% for Cape White-eyes and 15-25% Lark-like Buntings. Larger, louder birds also tend to have higher reporting rates compared to their densities: Karoo Korhaan for instance has a reporting rate of 35–40%, despite occurring at a density of 1–2/km2. So, while reporting rates broadly reflect densities, these relationships are confounded by habitat, size and life-history traits. This means we can’t just compare reporting rates between species to say one species is more common. But what about within one species range? Does higher reporting rate in some locations mean the species is more common there?

Recently I explored the relationship between density and reporting rate in pentads for several species in the southern Karoo region. Our team calculated pentad specific density estimates for 49 species and compared these to reporting rates, finding a good match for 75% of these. That means for a given species, as a general rule higher reporting rate generally means the species is more common.
But what about the exceptions to the rule? The species for which there was no clear link between reporting rate and abundance were generally the most common species. This is because they have high reporting rates, and reporting rate has an upper threshold of 100%. However, density estimates have no upper limit: for Cape Sparrow a reporting rate of 100% can mean a density estimate of 2, 10 or 50. The only way to get around this is to have many, many, cards for a set of pentads (a hundred or more), where the required detail when then become apparent. A repercussion of this is that for common species with high reporting rates we may not be able to detect declines using SABAP2 data. Generally, though, the implications for this are we can have greater confidence in the information derived from the SABAP2 in terms of what reporting rates are telling us for a species. For instance, consistent declines in reporting rate over time are likely due to local declines in density: as long as we are consistent in our atlassing efforts.   

With SABAP2 rolling forward, it also means that nearly all information required to make decisions regarding a species conservation status using IUCN criteria can now be acquired from SABAP2 data: range sizes and population trends (Lee et al. 2017, Bird Conservation International 27:323-336), and for some species, population sizes (de Kock and Lee 2019, African Zoology). That is a great achievement, which anyone who has ever submitted a list to SABAP2 can be proud of, although there are certainly species for which field work will be required: Hottentot Buttonquail being a case in point. May the atlassing efforts continue long into the Future. 

SABAP2 website now displays reporting rate as a gradient, which allows one to get an idea of abundance within a range: in this case red indicates high abundance, while pink indicates low abundance. White areas had no data at the time this map was created.

This chart illustrates the theoretical relationship for a hypothetical species, where increasing reporting rate is correlated with increasing real abundance in pentads across the species range. However, the species reaches an upper density threshold where this relationship fails to provide more information on abundance: when reporting rates reach levels close to 100% then density estimates in this case could by ANY figure upwards of 20 birds per square kilometer. The blue line is a loess smoother to illustrate the trend in the points (density in pentads), but the points could also represent average density of a species together with their average range-wide reporting rate (i.e. each point could also represent a species).

Cyber-bully Klaas van Dijk vs the Omar F. Al-Sheikhly - a cautionary tale to anyone corresponding with Van Dijk about Basra Reed Warbler research

Cyber-bully Klaas van Dijk vs the Omar F. Al-Sheikhly - a cautionary tale to anyone corresponding with Van Dijk about Basra Reed Warbler research

I must admit have my flaws. I generally go through life believing in the best in people. So, as editor of Ostrich, Journal of African Ornithology, when I received an article titled “TnF facilitating research fraud” I was genuinely interested, and believed I was receiving an article from a rational, decent human being. Turns out not to be the case; after only a few weeks I was also being accused of being complicit in fraud, and if you dare disagree with Mr van Dijk, then you will be too. Possibly also your neighbour, even if they haven’t even heard of a Basra Reed Warbler and the journal called Zoology in the Middle East. Certainly, Klaas van Dijk appears to me to resemble a jihadist with the sole purpose in life of getting papers published by Al-Sheikhly retracted.

So why as editor of Ostrich was I interested in the van Dijk article? At the time, with no background other than the article written by Mr van Dijk, I took him at his word: the article had quotations of support from many big names in science, and I presumed the article was novel and would be interesting to readers of the journal: we could demonstrate that we are willing to consider all sorts of articles, and have the publisher respond to the allegation. A sort of click bait saga. As such, I informed Mr van Dijk that I was interested and would pursue the matter further with TnF.
However, it soon became clear that this had already been done: Mr van Dijk (VD for short) has a project on ResearchGate and a wiki page where he documents his correspondence with everyone. The article he finds offensive has been commented on, examined, cross examined to the point of boredom. TnF had already issued various statements on the matter.... So - not exactly the novel, controversial story it appeared to be after all, and certainly nothing pertaining to research in Africa, although the species does occur here to some extent, I have learnt. And by the way, a common theme in his documentation of communications is that people stop communicating with him. It was starting to seem a bit like Ostrich was being used to fulfil a personal agenda.  Meanwhile, considering this to be old news, I declined the story.
From here, things went downhill rather fast, and I suspect if you are a semi rational reader, you will begin to understand why the communication with VD soon stops. Firstly he writes this to me:
The views and the statements in your e-mail of 18 May 2019 reveal that you reject that Al-Sheikhly et al. (2013, 2015) is based on fraud.  
WHHAT!? No-way! In fact, I have no opinion on if the Al-Sheikly story is fraudulent or not, I am not interested in it, not qualified to comment, and in fact wrote to VD that I saw his point of view. It could be bad science, fraud, good science, I honestly do not know. I am being forced to form an opinion on a species I have never seen, written by people I have never met, in a country I have never been to. Should it have been retracted? Sure: it would have made everybody’s life easier, but at this stage while the article is advisedly poor science, actual fraud has yet to be proven. VD further wrote:
This implies that it is mandatory for you to ensure that I, Richard Porter, and all co-authors of Porter et al. (2015..... get full and unlimited access to the full set of raw research data of Al-Sheikhly et al. (2013), and before the end of week 21 (Sunday 26 May 2019).
WHHAT!? That is about as reasonable a request as "Mr VD, please go to the moon and live out the rest of your days there". I do not have the data, I do not know the authors nor have any authority to comply with this demand. 
Alternatively, I propose you start with working together with me, with Richard Porter, with all co-authors of Porter et al. (2015a&b), and with lots and lots and lots etc. of other conservationalists and ornithologists, to ensure that the fraudulent study on the breeding biology of the Basra Reed Warbler (Al-Sheikhly et al. 2013, 2015) gets retracted.
Ok - so I am offered a choice of the impossible, or to join ranks with the lynch mob. To me, this is a threat, and as we shall see, it was no idle threat. In fact, I wonder how many others have been either tricked or coerced into supporting his position. I indicated to VD that I found his email threatening and would not be corresponding further, and engaged a filter to archive all email from him. Damn, if only it had been that easy to end this saga.
VD then embarked on a personal slander attack directed at me, with this email titled about Alan Lee and research misconduct and issues at Ostrich to Peter Ryan, director of the Percy FitzPatrick Institute:
One week ago I have informed you and some others at the FitzPatrick Institute of African Ornithology that your subordinate Alan Lee is currently involved in covering up a clear case of scientific misconduct. You and others at the FitzPatrick Institute of African Ornithology have until now not informed me that these allegations are unfounded and/or incorrect.
Clarification: I am not a subordinate of anyone, I am an independent researcher, with an affiliation to the Fitz. And if I was trying to cover it up - oops, I am a bit late for that too by about 4 or 5 years.
You and others at the FitzPatrick Institute of African Ornithology have also not provided me comments from reviewers / experts (together with their names and e-mail accounts) on reports which are published at ....
I have therefore concluded that there are 0 (zero, nul, nada, niente) experts / reviewers at the FitzPatrick Institute of African Ornithology who have been able to refute / rebut any of the findings of the reports which have been published at ....
I have thus also concluded that these allegations against Alan Lee are founded and that I am allowed to make public statements that your subordinate Alan Lee is currently involved in covering up a clear case of scientific misconduct. This all implies that the Editor-in-Chief of the scientific journal Ostrich is embroiled in covering up a clear case of scientific misconduct. It goes without saying that such an acting by your subordinate Alan Lee is disasterous for the good name of Ostrich.
I recall that you and others the FitzPatrick Institute of African Ornithology and Alan Lee and others (etc.) have not responded on the >1,000,000,000,000,000,000 times repeated requests to provide me and Richard Porter and all co-autors of Porter et al. (2015 a&b), see and , full and unlimited access to the full set of raw research data of Al-Sheikhly et al. (2013).
This acting by you and by others at the FitzPatrick Institute of African Ornithology and by Alan Lee on this formal request implies that I am correct by stating that publisher Taylor & Francis has ordered you that it is not allowed to communicate with me about this request.
WWHAAAT!? Do I even need to get into the lunacy? Apparently, VD is someone who believes the end justifies the means. This is a clear case of malicious slander. Ostrich is co-published by TnF - but I actually had permission from them to publish the VD letter should I wish. I do not wish - it would be giving into bullying. VD is quick to wave a copy of the COPE publishing code of conduct at everyone, but does not seem to abide by it himself (in fact appears complicit in attacking it when it suits him  
Unfortunately for me, the email to Peter Ryan has not get me fired by BirdLife South Africa. Apparently according to Mark Anderson, I am doing a great job. I wish I could get fired. Then I could get back to life in the mountains. Anyone want the editor position? You only have to deal with VD type issues about once a year. No takers? Damn. When Peter Ryan did not give into VD demands (to have me fired), VD laid a complaint of misconduct against him with the university: to waste the time of our highly ranked scientists in this way is inexcusable, frivolous behaviour.
By the way, to keep this all digestible, I have left out an entire page of demands that cannot be met, by me or anyone, such as 5 reviews and reviewer details of the information on his wiki page. I am sure VD will publish that email soon enough, bore yourself on his research page in the meantime.
And of course, anyone associated with Ostrich is now taking bullets from VD. Here an extract of an email to the director of NISC: .....the Managing Director of NISC, agrees with us that this full set of raw research data does not exist. You have unilaterally decided not to communicate with me about this crucial topic. I am therefore forced to communicate with you about this topic through the concept of 'tacit approval within a fixed period of time'. To put it in other words, you formally agree with us that this full set of raw research data of Al-Sheikhly et al. (2013) does not exist when there is no response within 24 hours from now. OK?
From this I further must infer that all supposed statements of support for his cause are coerced, taken out of context, or simply people who no longer wanted to respond to his impossible emails. VD is also a lone agent: when I contacted Richard Porter (who originally pointed out the unlikely data presented in the Al-Sheikhly paper), he wanted nothing to do with VD.
VD also embarked on a mission to coerce statements from members of BirdLife South Africa, copying in most staff for whom he could get an email address. By this stage, we had received legal council to not respond to any emails. VD then subsequently published on his Researchgate site a statement saying that BirdLife South Africa agreed with his position, which is a lie, because no such position exists.   
So, if you arrived here as an editor doing background research on Klaas van Dijk - my commiserations, you are in for a tough ride. My stance - say NO to cyber bullying, use the correct procedure for requesting a retraction, and make better use of your time by not engaging with people like VD.   

Tuesday, 3 September 2019

atlasing or atlassing?

Atlasing or Atlassing?

The other day I’m wearing my binoculars and tapping into my smartphone. Charlie asks “Daddy, are you atlaBing?”

Ah cute; but readers of German may be aware that there is character in their alphabet that looks like a B (eszett or scharfes S) but is pronounced as an ‘ss’ and is often written that way in modern texts. Clearly my anglo-germanic progeny was asking me “… are you atlassing?”

Recently, the verb form of contributing to the atlas has been the cause for some consternation. Ernst Retief, SABAP2 coordinator, recently sent out an email to the SABAP2 steering committee:

I need some clarity. We have always, since 2007, used atlasing, with one “s”, and not atlassing, with two. Can you please let me know what we should use and maybe why?

Peter Ryan deferred the question to Eve Gracie, editor of African BirdLife magazine, who responded:
In SA most formal writing still follows so-called standard English and the OED style, which is to forgo the double ‘s’ in words like busing, focusing and atlasing etc. The double ss is more American English in style.

Now I’m of the opinion that our ex-colonial overlords don’t really need to have a say in the matter: they’re not exactly setting a great example to the world at the moment, and according to Microsoft Office, South Africa now has a chance to define its own vocabulary (I presume, since I can select ‘English (South Africa)’ from the list of available language options). Currently writing in Microsoft Word, under all versions (UK, United States, South Africa), both atlasing and atlassing bring up the red underline for unrecognized words. This means in effect we have the chance to define how we wish to spell the word, right here, right now.

I have for my most recent writing been using ‘atlassing’. This was after Google researching ‘atlasing’ provided a definition related to the computer graphics industry, and then learning from Google Scholar that ‘atlasing’ is used in the realm of computational medicinal sciences. On the other hand, ‘atlassing’ brings up SABAP2 as the second and 3rd links, with Birdlasser (note the double s) in 4th place in a standard Google search. The results from a Google Scholar search on ‘atlassing’ return all results related to citizen science activities. Certainly, the web of knowledge seems to think we are atlassing rather than atlasing when it comes to watching birds.

But what about precedence? Well, since ornithology is a realm of biology, where first name should be adhered to, can SABAP2 claim that we should be using ‘atlasing’? Unfortunately, also not: James Harrison, first author of “The Atlas of Southern African Birds” based that book on ‘atlassing’ which can be traced back to publications from the 80s:

Harrison, J. A. "Atlassing as a tool in conservation, with special reference to the Southern African Bird Atlas Project." Biotic diversity in southern Africa: concepts and conservation (1989): 157-169.      

As a family, we often have a range of cereals on the breakfast table, which occasionally feature ProNutro or FutureLife. Recently, Charlie asked “Why is FutureLife called FutureLife?”: a fair question since the boxes contain neither future nor life in their lists of ingredients. I had to explain that this was ‘marketing’: inventing names from positive sounding words to make us buy stuff that should better be described as “assortment of ground up cereal crops”. Contributing to the atlas project is a positive thing, and so is singing: so Atlassing is a ProNutro/FutureLife activity so to speak, and certainly my choice going forward.  

Related Posts Plugin for WordPress, Blogger...