Just completed seven days of bird surveys at four sites here in the Pilbara. As a form of "experiment", I alternated wearing a high visibility vest or just my kakhi shirt while undertaking the survey. Anecdotally there was an significant improvement in the bird counts and the activity of birds noted while NOT wearing the vest.

That's great Stephen. I'm a little hamstrung here in Perth living out of a suitcae while waiting to fly out to the next survey. Your reference and abstract will come in very handy in making my arguement in the field. Next weel I will be undertaking timed area counts of birds in the Pilbara and will be required to wear a fluoro yellow vest and white hard-hat - neither of which will help with the survey. I am hoping to make a convincing arguement as to why we should be allowed to go back to our Khaki's and broad-brimmed tan hats while undertaking the survey (everywhere else, wearing the PPE is not a major issue).

The title and abstract of one additional reference (Riffell & Rifell 2002) is shown below. The general finding seems to be that a bird species whose body coloration is in part the same colour as the safety vest will behave more visibly, whereas other species will behave less visibly. But the magnitude of this effect seems to vary seasonally. The authors provide possible explanations for the seasonal variation in detectability.

This paper deals only with point counts of birds. I would imagine that there would be different results if a bird census was conducted while an observer was moving slowly along a transect or using the area search method.

Journal of Field Ornithology 73(4):351-359. 2002

Can observer clothing color affect estimates of richness and abundance? An experiment with point counts

¹Corresponding author. riffells@msu.edu

Thanks Edward for these references. I look forward to reading your precis.

Hi all

Thank you for your anecdotes.While researching this issue a little more, I came across the following papers:

Bye, S.L., Robel, R.J. & Kemp, K.E. 2001. Effects of human presence on vocalizations of grassland birds in Kansas. Prairie Naturalist 33: 249 - 256.
Gutzwiller, K.J. & Marcum, H.A. 1997. Bird reactions to observer clothing colour: implications for distance-sampling techniques. J. Wildl. Manage. 61: 935 - 947.
McShea, W.J. & Rappole, J.H. 1997. Variable song rates in three species of passerines and implications for estimating bird populations. J. Field Ornith. 68: 367 - 375.

These three were cited in:

Bachler, E. & Liechti, F. 2007. On the importance of g(0) for estimating bird population densities with standard distance-sampling: implications from a telemetry study and a literature review. Ibis 149(4): 693 - 700.

Having undertaken density measures in the past, this is quite an interesting article.

I will chase them up once I return to Sydney from my current field trip and provide some sort of precise.

I recall Martin having this issue with mining companies as well.

Hi Edward,

I am not aware of any published (or unpublished) studies on this topic, but I think that most consultants would have their own anecdotes they can share.

The same situation applies when conducting roadside bird surveys. I can understand the road and construction safety angles of wearing PPE during surveys but, as you say, bright colours do affect the detectability of some bird species.

In addition to PPE, I was doing a roadside bird survey about 12 months ago when a road construction engineer drove up in his vehicle and asked me what I was doing. After my explanation, he said that I should have road traffic controllers slowly driving their vehicles with revolving flashing lights, one vehicle in front and one behind me while I was walking along the side of the road. I had a hard time convincing this engineer that this would disturb the birds and defeat the purpose of the bird surveys, but eventually he relented.

On another occasion, I was pulled up by another engineer for not wearing leather gloves (worn when handling construction material and machinery).  When I explained what I was doing, he handed me a pair of gloves and told me to clip them onto my belt. That was 3 years ago, and each time I conduct bird surveys on this site, I have these gloves clipped to my belt, but I've not worn them on my hands once!

How the detectability of birds is affected depends very much on the species, habitat type and size and the observation skills of the observer. In my experience, most bird species are alarmed when they see (a) person(s) wearing brightly-coloured safety vests. The more timid species will hide and keep quiet, or quietly move away. These are the more difficult species to detect. However, the majority of species sound alarm calls and often find safe vantage points from where they can keep an eye on the observer. Often the movements of birds to those vantage points make them more visible to the observer. A skilled and experienced observer should be able to identify the bird species from those alarm calls, subsequent contact calls between conspecifics (if it is a social bird species) or by sight if the bird is visible or while it is flying into the distance. Contact calls between conspecifics are often low in volume and/or subdued in nature, and probably serve to signal the location of an individual bird to its conspecifics, so an observer needs to have especially good hearing and excellent bird sound identification skills.

It is easier to detect alarmed birds within linear (narrow) habitats (e.g. in roadside reserves) because this habitat structure limits to some degree where a bird can move to when alarmed. However, in a larger, more widepread, habitat there is a greater opportunity for birds to move longer distances unnoticed.

And, obviously, it is more difficult to detect alarmed birds in closed habitats (e.g. heathland and closed forest) than in open habitats (e.g. open woodland).

On the positive side, common sense does prevail with some clients. I have been on some mine or oil field sites where I've been allowed to wear a wide-brimmed sun hat (instead of a hard hat) while working in bushland, away from works areas, and I've not had to wear a brightly-coloured safety vest.

Hi to all

One of the issues I have been grappling with over the last 5 years is how to convince mine managers, companies, etc on the impact of coloured safety clothing on the efficacy of bird surveys. We are now required to wear, at all times, a white hard-hat (and brim) and bright yellow/blue or orange shirts with fluorescent "night" strip (vests are not allowed anymore). The issue is not petulence on our part but we feel that the "loud" colours quiten down the bird life quite dramatically. We certainly don't mind wearing vests and the hard-hats, especially when in production or administration areas, but we have been warned not to be seen without the PPE at any time as it would constitute a breach and we would be banned from site. The nature of our work means that we are only in production areas on route to our sites (usually many km away).

We have only anecdotal evidence at the moment but during our spring surveys in open shrubland/woodland habitat, our bird counts are down about ~20% (we had one site noted for its birds down by ~35% - knee high salt-bush woodland surrounding densly vegetation waterhole). Due to the nature of our work it would be very difficult to implement a controlled experimental design to test the hypothesis.

What I would like to do is to prepare a Safety Work Practices paper to submit to the client that backs up our case to be limited to wearing such colourful PPE only in production and administration areas. Can anyone help with research here in Australia or from overseas on this subject?