|
A PRELIMINARY ANALYSIS OF DISTRIBUTION RECORDS FOR THE SOUTHERN GROUND HORNBILL FROM THE REPUBLIC OF SOUTH AFRICA
Alan Kemp Ground Hornbill Research and Conservation Project, Mabula Game Reserve, Bela-Bela, Limpopo Province, South Africa. April 2007.
THE ANALYSIS
This analysis for the Southern Ground Hornbill (SGH) is based on 1 860 records assembled, and in part collected, by the Mabula Project, and then plotted on a map using GIS software. The analysis attempts to provide an objective basis for future discussions on: 1. The extent of the historical range of the species in South Africa. 2. The extent of the current range. 3. Measurement of any rate of change in range and density.
DATA
The data have been assembled from any possible sources of information. Each record was entered into an Access database, with as comprehensive information as was available, especially for date and location. The database is available from the Mabula Project, by arrangement, and is being continually edited, improved and expanded as new information becomes available. As far as possible, these data are visualized on a GIS map that is available on and can be downloaded from the Mabula Project website, and the map will also be updated at intervals. This analysis uses the initial map that was presented in April 2007.
At this stage, the data presented on the map have been divided into sets covering intervals of 25 years (or multiples thereof). Twenty-five years is taken as the minimum time estimated for a SGH generation in a stable population, based on 0-6 years attaining maturity, 6-10 years attaining an alpha breeding position within a group, 10-25 years to produce four fledglings (one every 3-4 years), of which half are expected to attain adulthood and so replace their parents. It is expected that the real generation time for wild populations in undisturbed conservation areas, once recorded, will be much longer, of the order of 40-50 years, and so might extend across two of the time intervals used in this analysis.
1) Prior to 1900, 81 records have been located (plotted as yellow points, comprising 4.4% of all records). These came from a variety of published sources, starting with Wahlberg’s specimens from Gauteng in 1842 (Craig & Hummel 1994). 2) 1901-1950, only 22 records were located (plotted as orange points, comprising 1.2% of the total). Most were from a variety of published, anecdotal and specimen records. With so few records, they have been combined within a 50-year period. 3) 1951-1975, only eight records were located (plotted as red points, comprising 0.4% of the total). Most were from a variety of published, anecdotal and specimen records. 4) 1976-2000, 287 records were located (plotted as purple points or squares, comprising 15% of all records). This was the period of bird atlassing in southern Africa, so most records were only available to the resolution of a quarter-degree square (QDS). Records were extracted mainly from the atlases of the Transvaal (Kemp et al. 1985; Tarboton et al. 1987), Swaziland (Parker 1994), Natal (Cyrus & Robson 1980) and southern Africa (Harrison et al. 1997). There were also specific studies on SGHs that provided comprehensive data from the Eastern Cape (Vernon 1986) and KwaZulu-Natal (Knight 1990), and also a variety of published, anecdotal and specimen records. In future, it may be possible to expand this data set from records with a spatial accuracy of only a QDS to the spot records that comprised the original records. 5) 2000-present, 1 543 records were located (plotted as blue points, comprising 79% of all records). The Mabula Project was started in 1999, and soon after its inception an intensive programme began to solicit and collect records of SGHs.from throughout South and southern Africa. Most records came from sightings reported by the public (of which 78% were from the Kruger National Park, where they were recorded by visitors in books placed at reservation offices throughout the Park).
RESULTS
Records have been plotted on the map in two ways. Those only accurate to a QDS have been plotted as a smaller square centered within the relevant QDS. Other records have been plotted as points, either from exact coordinates, taken off maps or as a GPS reading, or as the centre point of a property, such as a farm or game reserve. Both levels of spatial accuracy are considered adequate for understanding SGH distribution, given that the estimated density of one group/100 km2 extends beyond the limits of most properties, or occupies a significant portion of a QDS (roughly 25 x 27 km, or 675 km2).
All records prior to 1976 were too few to offer any indication of the rates of change in distribution of SGHs in South Africa, but some records did indicate that the range of the species was considerably wider than after 1976, especially in Limpopo and Gauteng Provinces.
The various bird atlassing and research projects during 1976-2000 provided the most systematic data for each of their respective regions and, together, for the whole of South Africa. Systematic data collection for these studies started in 1976 (Kemp et al. 1985; Cyrus & Robson 1980) and ended in 1992 (Harrison et al. 1997), but only a further 14 records of SGHs were located from then until the end of the 25-year survey period, and the 20th Century.
The records assembled during the 21st Century, after the establishment of the Mabula Project, while more numerous than during any previous period, did not provide systematic spatial coverage.
These differences between the data available for each 25-year period make comparisons between periods difficult. The following is the best initial summary of these data, by province, that seems possible at this stage. Taking the data collected systematically during 1976-2000 as a baseline, it is possible to add up the number of QDSs in each province or region as a measure of the range of the SGH at that time. It is then possible to compare this with the distribution of records since 2000, noting the number of QDSs where SGHs have been recorded a) only during the 1976-2000 period, b) during both the 1976-2000 and 2000-present periods, and c) only after 2000. In addition, for records prior to 1976 that fall outside of the QDSs where SGHs were recorded after that date, one can connect these records and count the number of QDSs they enclose in which SGHs have not been recorded after 1976. This count provides a first estimate of the historical range of the species. These figures are presented in the table below:
Number of quarter-degree squares in which Southern Ground Hornbills have been recorded in South Africa, and which might have been included within their historical range.
DISCUSSION
Despite the systematic collection of data by QDS during 1976-2000, SGHs were not reported in all the same QDSs after 2000. SGHs were reported in some QDSs only during 1976-2000 and not subsequently (overall, 144/261, or 55%), and in other QDSs only after 2000 (overall, 28/261, or 11%). The differences in reporting during each time period varied between provinces/regions, with the greatest incongruity in Mpumalanga/Swaziland.
Because the records were only collected systematically during 1976-2000, no firm conclusions can be drawn from these differences between 25-year periods, overall or by province/region. However, the data do identify areas where additional research might yield numerically sound conclusions. If those QDSs where SGHs were recorded during 1976-2000, but not subsequently, were revisited and found lacking in SGHs, then this would provide evidence for their decline within such areas. If those QDSs where SGHs were recorded only after 2000 were revisited, it might be possible to trace anecdotal data as to whether they occurred there previously or not, showing either that they were missed during the 1976-2000 surveys, or that there has been a recent extension of their range.
Assuming that the spread of pre-1976 records encloses the historical range of the SGH, the natural distribution of the species would have covered 494 QDSs (about 331 968 km2), 40.3% in Limpopo/Gauteng (82+117/494), 7.7% in Mpumalanga/ Swaziland, 26.3% in KwaZulu-Natal/Free State and 20.0% in Eastern Cape. Then, assuming that there has been no change in distribution since 1976, the present QDS range of SGHs within South Africa would cover 261+28 QDSs (about 195 075 km2, or 59% of the historical range), 31.4% in Limpopo (82/261), 5.7% in Mpumalanga/Swaziland, 39.8% in KwaZulu-Natal/Free State and 23.0% in Eastern Cape. The differences between these figures suggest a retraction in range by SGHs over the last 134 years (1842-1976, a maximum of about 5 generations) of 41% overall (494-(261+28)/494), or of 56% in Limpopo/Gauteng, 61% in Mpumalanga/Swaziland, 20% in KwaZulu-Natal/Free State and 34% in Eastern Cape.
These analyses can be immediately improved in two ways. First, as mentioned above, visits can be made to QDSs where SGHs were found during 1976-2000 but not post-2000, and vice versa, to ground-truth the accuracy of survey results for these two periods. Second, the historical range of the SGH can be determined more accurately by GIS analysis of the habitats encompassed by historical and recent records, and by measurement of the exact extent of those habitats preferred most by the species.
CONCLUSIONS
It would be premature, with the data available, to make any objective conclusions about the extent and the rate of change in the distribution and density of the SGH. However, they do indicate trends in SGH decline that are of concern, and that could be measured. Validation of these trends requires more accurate and extensive data, 1) from a further search for published and anecdotal records prior to 1976, 2) from conversion of the 1976-2000 QDS data into the more accurate point data that comprised the original records, 3) from ground-truthing of those QDSs where 1976-2000 and post-2000 records are not congruent, and 4) from more accurate habitat-based estimation of the historical range of the SGH.
The main trends that are indicated by this analysis are:
A. There has been a decline in range of the SGH in South Africa of about 40% overall, and this decline has been greater in the northern areas of Limpopo/Gauteng and Mpumalanga/ Swaziland (56-61%) than in the southern areas of KwaZulu-Natal and Eastern Cape (20- 34%). Assuming that the estimation of historical range is accurate, these declines might be even greater if QDSs only found to contain SGHs during 1976-2000 are no longer occupied. B. The large decline apparent in Mpumalanga/Swaziland is of particular concern, since it suggests a break in the South African distribution and population of the species, and this has important consequences for conservation and management. That this ‘Swaziland Gap’ appears to be real is supported by the virtual lack of recent, post-2000 records in these areas, and in adjacent northern KwaZulu-Natal. C. The main decline in range has been in the northern savanna areas, especially outside of the extensive Lowveld conservation areas of the Greater Kruger region. This emphasizes the importance of the Greater Kruger region as the core of the current stable population of the species, and indicates that conservation management should be directed to the non- conservation areas of its range. The relative extent of the current range of the species in different provinces/regions, as opposed to the historical range, is also important in separating priorities for conservation of the current populations, from different priorities for re-populating the historical range.
Every effort should be made to extend these analyses, and all possible sources of new and more accurate records should be explored.
REFERENCESCraig, A. & Hummel, C. (eds). 1994. Johan August Wahlberg: Travel journals (and some letters). South Africa and Namibia/Botswana 1838-1856. Van Riebeeck Scoiety, Cape Town. Cyrus, D. & Robson, N. 1980. Birds of Natal. University of Natal Press, Pietermaritzburg. Harrison, J. A., Allan, D. G., Underhill, L. G., Herremans, M., Tree, A. J., Parker, V. & Brown, C. J. (eds). 1997. The atlas of southern African birds, 2 vols. BirdLife South Africa, Johannesburg. Kemp, M. I., Kemp, A. C. & Tarboton, W. R. 1985. A catalogue of the birds of the Transvaal. Transvaal Museum – Transvaal Nature Conservation Division, Pretoria. Knight, G. M. 1987. Status, distribution and forgaing ecology of the Southern Ground Hornbill (Bucorvus cafer) in Natal. Unpublished M.Sc. Thesis, University of Natal, Durban. Parker, V. 1994. Swaziland bird atlas 1985-1991. Websters, Mbabane. Tarboton, W. R., Kemp, M. I. & Kemp, A. C. 1987. Birds of the Transvaal. Transvaal Museum, Pretoria. Vernon, C. J. 1986. The Ground Hornbill at the southern extremity of its range. Ostrich 57: 16-24. |
|||||||||||||||||||||||||||||||||||||||||||||||||||||
