The Swaziland Bird Atlas is unique, I believe, in three ways: the small grid squares (compared with other African bird atlases), the depth and consistency of coverage and, not least, because almost all of the fieldwork was carried out by the author himself. The atlas plots the distribution of a remarkable 489 species recorded between 1985 and 1991 including such local species as Blue Swallow and Pink-throated Twinspot.
Swaziland is a small country of only 17,360 square km. The size of the country permitted the use of a one-eighth degree square grid (approximately 12.5 x 12.5 km), rather than the one quarter degree square employed in the southern African and Kenyan atlases. In the introduction the author demonstrates the value of the finer grid by plotting the distribution of the Fiscal Shrike and the Long-tailed Shrike on both scales. The coarser scale obscures the difference between the Knobthorn-Marula savanna of the eastern lowveld and the more mixed savanna of the western, giving the impression that the two species overlap in much of their range in Swaziland which, in fact, is not the case. A similar comparison for the Water Dikkop shows how closely its distribution is tied to major rivers, a fact that the quarter degree square fails to demonstrate.
Swaziland conveniently divides into 100 one-eighth degree squares. One of the remarkable achievements of the author was to visit each square at least once for a minimum of five hours in each month of the year. This gives the atlas a rare conformity of coverage suggesting that gaps in a species plotted distribution are real, rather than quirks of reporting, and provides a comprehensive picture of seasonal movements for migrant species.
An introduction section of 17 pages discusses the methods employed and the results. Maps are provided showing the major towns and rivers, topography and vegetation and rainfall. A chart details the number of species recorded in each square for each month and in total, the highest score of 300 species going to Big Bend on the Usuthu River in the southern lowveld. An interesting analysis shows that, while Swaziland's political borders seem at first like an eccentric colonial carve-up, they in fact coincide with the range boundaries of a surprising number of species.
There follows the main bulk of the work comprising the account and distribution map for each of the 489 species recorded, two species being treated on an A4 page. The taxonomy follows Roberts' Birds of Southern Africa (1995). For each species the author provides one of two types of distribution map (discussed below), its Roberts' number, English and scientific name, the number of times it was recorded out of 2,263 grid visit cards, an estimate of the number of breeding individuals in Swaziland and a brief description of its status and habitat preferences. If I have one criticism it is the brevity of the habitat preference section. Typically this runs to less than ten words and an opportunity to describe each species' habitat requirements in more detail has perhaps been lost. This information has considerable conservation value and could help provide clues to future changes in Swazi bird distribution. The population estimate seems somewhat arbitrary, but the attempt to put an exact figure on a population rather than place it in a broad band can only be admired. With the exception of a few censused populations, such as the Blue Swallow, the population figure is calculated by multiplying a rough estimate of the species density by the estimated area of suitable habitat. The author comments that he has erred on the side of caution and that most populations are probably larger than given.
Two types of map are used in the species accounts. For migrant species each grid square contains a circle divided into 12 segments, for the months of the year, the months in which the species was recorded being blacked out. This treatment clearly demonstrates the seasonal movements of Palearctic non-breeding and intra-African migrants. For example the map for Common Buzzard, a very common non-breeding visitor, shows the earliest arrivals reaching Swaziland in October with the majority of highveld squares occupied by November. Most squares are deserted by February. The map clearly demonstrates the species's preference for higher altitude, with a much briefer and more patchy distribution in the lowveld, especially in the north.
A resident species' presence in a given square is represented by dots of varying size, proportional to the recording rate within that square. The recording rate represents the percentage of months in which the species was recorded, not the commonly used proportion of times recorded in the square. This interpretation is used because 'many species showed significantly lower recording rates for those grid units for which there were a large number of field cards because many such field cards had relatively low species totals'. The author argues that this approach eliminates this bias, but readers accustomed to interpreting abundance of a species by size of dot need to be wary.
This atlas provides an accurate picture of Swazi bird distribution correcting many errors and false assumptions (perhaps in a few cases reflecting true change) in such standard works as Roberts, though perhaps it is not fair to compare a national atlas with a regional handbook, whose remit is much broader. It also provides valuable information on the conservation status of 54 endangered, vulnerable, rare or regionally rare species, of which Cape Vulture, Kori Bustard and Pel's Fishing Owl were found to be extinct as breeders in Swaziland.
I highly recommend the book as a very valuable contribution to African bird distribution. Needless to say it is essential reading for anyone living in or visiting this small but attractive and ornithologically rich country. I congratulate the author for his monumental task and the Conservation Trust of Swaziland for their vision in sponsoring the work.