19 September 2008    

Dear atlasers

We do apologise for the delay in getting this third newsletter out but most of the project team were involved in preparations for the 12th Pan-African Ornithological Congress (PAOC) that was held from 7-12 September at Goudini Spa near Worcester. This international congress, which is held every four years, brings together mainly African and European ornithologists in a one-week forum to discuss current research initiatives and conservation issues facing Africa's avifauna.

SABAP2 did feature at the congress - Doug Harebottle and Mike Brooks presented two posters while Les Underhill presented an oral on "An introduction to SABAP2". Mike's poster, which dealt with evaluating and verifying bird checklist data using integrated bio-demographic bird datasets (and based mainly on SABAP2), won first prize in the poster competition. Well done Mike! There was also a small contingent of atlasers who attended the congress and a meet-and-greet session was arranged during the week to chat about any pertinent issues. Each atlaser was encouraged to compile an atlas card for the 'Goudini pentad' during their stay at Goudini - this should turn the pentad to a dark blue once all the data has been received.

View to the north from the main conference facilities at Goudini - notice the snow on the peaks of the Badsberg mountains!

A warm welcome to all our new atlasers who have submitted their first card since our last newsletter. You represent a small but important part of a greater atlasing community striving to build a better information base for South Africa's avifauna. We hope that you will have a long and lasting relationship with the project. The most challenging part is getting into the field and then submitting your first checklist. We find that once this has been done, people usually feel more comfortable with the process and are then keen to get back out and do some more surveys.

In July we passed the 10% coverage and 250 000 records. Considering the relatively slow build up since the project took wings on 1 July 2007, we consider this to be a wonderful achievement. This has largely been due to the growing number of atlasers, from all over South Africa, who are registering to take part and submitting their data.

At the end of July, 265 atlasers had conducted surveys and submitted their cards to the project - this is compared with 181 active observers after the first six months, an increase of 45%.

By mid-September we were closing in on 13% coverage. Coverage is increasing at a rate of just over 1% per month, which translates to about 180 new pentads each month. We passed a third of a million records (333 333) on 16 September. At this rate, we will reach half a million records by the end of the year. Currently, we average almost 1900 records per day (about 35 cards) submitted to the project daily. Most of these arrive by email via the Data Management System, with the remainder being submitted manually (posted) to the ADU. By mid-September, the number of active participants had grown still further, to 332.

A broad look at coverage as at the middle of September 2008. The yellow squares represent new pentads that were surveyed since June 2008.

25 pentads are "in the red" - they have more than 25 checklists submitted. Consistent and regular atlasing in your home pentad achieves two things: a firm baseline against which future change can be measured and really valuable information on the timing of arrival and departure of migrants.

Thank you for all your contributions. Whether you have travelled far and wide to survey new pentads, or repeatedly surveyed your local pentads, you are collecting important data for monitoring changes in our bird distributions which will all go towards future bio-diversity assessments and re-prioritising avifaunal 'hotspots' (areas of conservation importance and concern) in South Africa.

2. Challenge results

The results of the last two challenges are as follows:

Late summer handicap challenge (20 March - 31 May)

John McAllister and Pieter la Grange scored over a thousand points taking the top two places respectively. Their efforts were rewarded for having covered new pentads in uncovered two-degree, one-degree and half-degree grid cells.

Deep and wide challenge (1 June - 31 July)

This challenge encouraged people to go both deep (repeated surveys) and wide (surveying new pentads) during the middle of winter, a time of year when we were expecting atlas surveys to subside.

Well, atlasers responded surprisingly well and we had over 100 observers taking part in this challenge. Geoff Lockwood (Delta Park, Jhb), one of our SteerCo members, and Eddie Du Plessis (Middelburg, Mpumalanga), topped the Deep Challenge with 389 and 388 points respectively - Arnold vd Westhuizen (Aliwal North) and Jaco Janse van Vuuren (Beaufort West) also made significant contributions scoring 358 and 307 points respectively. Just goes to show that by doing repeated surveys of your local home patch you can still make a significant contribution to the project.

For the Wide Challenge more points were scored for surveying pentads in un-atlassed one-degree, half-degree and quarter-degree grid cells, while pentads blank on 1 June scored the least amount of points. The Wide leaderboard was topped by Johan vd Westhuizen from Moreesburg (365 points) with John McAllister (Wakkerstroom), Stefan Theron (Beaufort West) and Andre Marx (Pretoria) all scoring over 150 points. Following this challenge, all one-degree grid cells had at least one pentad surveyed. Well done to the above 'travellers' for making a real effort to get to some far and wide pentads.

Our grateful thanks to all the atlasers who submitted cards as part of these challenges. You have made a significant contribution to the project database, but most of all we hope that you had fun taking part.

We would like to shed some more light on out-of-range records as there seems to be some confusion as to how the process works. As SABAP2 records are submitted they are automatically checked (vetted) against existing bird datasets, notably SABAP1, and if they do not meet certain criteria (see below) they are flagged for verification. This enables us to keep the database as accurate as possible.

The vetting process works at various levels as the data arrives and gets processed:

• Level 1: records checked against verified SABAP2 data. Records are compared with existing data for the submitted pentad and in the surrounding eight pentads. If there are any records within this 9 pentad grid that have been verified, the record is accepted. If no verified records are found, the record is passed onto the next level of the vetting process.


• Level 2: This level takes the record and compares it to the SABAP1 data. The software currently looks at the Quarter-Degree Grid Cell (QDGC) that the pentad forms part of. If the species was atlased in the QDGC during SABAP1, then the species is accepted. If the species was not recorded in the QDGC, then the record is passed to the next vetting level. Note however that if the reporting rate is less than 5% for a QDGC the species is automatically marked as a Marginal out-of-range record due to its apparent 'rarity' in that grid-cell during SABAP1.


• Level 3: The record is then checked against vetted (cleaned) Birds in Reserves Project (BIRP) data. Again, if the record is in a QDGC where a BIRP record for the species has been verified, the record is accepted. Should the record fail this level as well, an Out of Range Form (ORF) is sent to the observer.


• Level 4: This level was introduced to try and reduce the number of marginally out-of-range records that were generated through levels 1-3. Here average SABAP1 reporting rates for the eight neighbouring QDGCs are used to determine the appropriate action to be taken for the record. This is done through a vetting matrix where records are accepted as the average reporting rate increases together with the no. of neighbouring quarter-degree grid cells in which the species was recorded. Only records that fail the first three levels of vetting will be subjected to this fourth level.


• Note: Regional and/or National Rarities, are automatically flagged as being 'out-of-range' and do not go through any of the levels described above.

When receiving ORFs please remember to do the following:

(1) Check the pentad code and species are correct before forwarding it on to the relevant RAC chairperson (contact details are on the Species Verification Sheet, the Word document). If the pentad code and/or species are incorrect (due possibly to typo errors during data capture) please send the corrections to Ray Sebola quoting the card number as a reference. If the species and pentad code are correct do one of the following:

(2) If it is a Marginal Out-of-Range record you just need to forward the Information Sheet (pdf document) to the relevant Regional Atlas Committee Chairperson for verification.

(3) If it is a Full Out-of-Range record you will need to complete the Species Verification Sheet (SVS) before forwarding the record to the Regional Atlas Committee Chairperson. You must please submit both the Species Verification Sheet and the Information Shhet (pdf) to the RAC. Please try and complete the SVS as thoroughlly as possible as the more information you can supply the easier it will be for the Regional Atlas Committees (RACs) to adjudicate the record otherwise the committee will have to base their decision on the information that is supplied. If you have photographed the bird, please attach the photo(s) when submitting your ORF. Photographs usually provide the best clues as to the species identity.

The RACs then assess these records and either Accept, Hold over or Reject the record. Note, however, that 'Held over' and 'Rejected' records are not deleted from the database but simply marked as unusable. This allows us to re-visit (and change) these records should future surveys in these pentads confirm the birds occurrence as being valid.

We hope that this breakdown will help you the observer understand the vetting process and why ORFs are generated and why these records need to be assessed. Remember that an out-of-range record does not necessarily mean that you have made a mistake (unless the species was incorrectly captured), but rather that you have probably seen something out of the ordinary and we need your assistance in confirming the record. As new areas are surveyed, particularly rural pentads, fairly large numbers of ORFs may be produced. This is due simply to the fact that these areas were poorly covered during SABAP1 (i.e. low number of checklists received) and with more intensive surveys being carried out for SABAP2, species are now being reported that were missed or possibly absent during the first atlas. The golden rule here is that if these ORFs are attended to quickly and are accepted they will not be flagged as being out-of-range the next time the pentad (or neighbouring pentad) is surveyed.

This is another exciting initiative that we have recently launched. Inspired by Ernst Retief, the Gauteng Regional Atlas Committee chairperson, this simply strives to encourage more atlasers to get out and conduct surveys during a single week each month. The first CAW was arranged for August and proved to be a great success.

Each month a CAW will be planned and a challenge put out to see if last month's targets (e.g. no. of new pentads covered, no. of observers taking part, no. of checklists submitted) can be bettered. The November CAW will form part of BLSAs Birding Big Day atlas category. More information will be available on the website shortly.

If you have any general queries about the project please contact Doug Harebottle or Les Underhill

If you are experiencing problems with the Data Management System (DMS) please consult the Frequently Asked Question page on the website, and if this does not solve your query please contact Michael Brooks, Tel. 021 650 4998

Happy atlasing!

Regards
The SABAP2 Project Team