The forecasts in different colours on the maps that the model produces are based solely on estimated rainfall. However, Red-billed Quelea will only breed below 1200m above sea level in suitable vegetation and where there is access to water for drinking. In terrestrial habitats the birds usually breed in vegetation such as Acacia tortilis, A. mellifera or Dichrostachys cinerea thickets. On the occasions when they breed in aquatic habitats they occur in reed-beds, usually amongst Typha, Phragmites or Papyrus.

The type of vegetation in an area may be related to the underlying soil type and its development rate may also be dependent on the soil. Rain falling on sandy soil will drain away faster than on clay soil, so vegetation on sandy soil is likely to need more rain and take longer to develop than vegetation on clays.

All these factors (altitude, vegetation, soil and presence of sources of drinking water) will affect whether the quelea birds can actually breed in a particular zone that the forecasts specify as having received sufficient rainfall. In order to help users to interpret the meaning of the red colour of a particular ½ x ½ degree square we have added maps that can be used to help refine the forecasts. In the future we hope to incorporate some of these parameters into more detailed model algorithms.

Until this has been done if you are interested in refining a particular forecast and want to find out if a red model square does mean a real threat of quelea breeding, we suggest that you follow the 'help map' based question and answer method detailed below.

Refined forecast method - Ask yourselves the following questions:

Q1.    Have Red-billed Quelea ever occurred or bred in this ½ x ½ degree square?

ANSWER    Use HELP MAP 1a: This information is displayed on the "Distribution of the Red-billed Quelea in southern Africa (all records)" map - which gives data for both occurrences and breeding at a 1º x 1º resolution. For more detailed data on confirmed breeding look next at Help Map 1b.

ALSO..        Use HELP MAP 1b: This is a higher resolution distribution map of the breeding data. This shows data for all colonies for which we know the dates and locations of breeding to an accuracy such that they can be plotted on a ½ x ½ degree scale. Note that such information is not available for all the colonies plotted on the lower resolution map, so, for instance, the high resolution map lacks any data points for colonies in Angola. The high resolution map presents information on frequencies of occurrences (up to and including 2003/2004 data), thereby identifying areas that are hotspots."

Q2.    Is the vegetation in this ½ x ½ degree square suitable?

ANSWER    Use HELP MAP 2a: This information can be found on the "White's Vegetation" map. This is based on White's vegetation map of Africa (see White, F. (1983) The vegetation of Africa: a descriptive memoir to accompany the Unesco AETFAT / UNSO Vegetation map of Africa. UNESCO, Paris; and/or White's vegetation map. Global Resource Information Database (GRID) of UNEP (http:// www.grid.unep.ch/ data/ data.php ? category=biosphere). If the square is not in a zone that includes woodland, bushland, thickets, grasslands or scrublands it is unlikely that quelea will be able to breed there.

ALSO..        Use HELP MAP 2b: If the known quelea breeding sites (see Help Map 1) are plotted onto White's vegetation map, and the frequencies of colonies in different vegetation zones are recorded, an index of the likelihood of a zone being suitable for breeding can be estimated. Help Map 2b shows the results of such an analysis and grades the areas. Note that because more colonies have been recorded from South Africa, especially in the Springbok Flats area, the results are biased in favour of that type of vegetation (woodland).

Q3.    Is the soil suitable?

ANSWER    Use HELP MAP 3: A similar analysis to that for Help Map 2b was conducted on soil using the FAO soil map ( FAO, 1991 Soil Map of the World, 1:5.000.000., Rome, Italy) and Help Map 3 provides the rankings for the suitability of the underlying soil in different places.

Q4.    Could a waterbody allow breeding in unsuitable vegetation zones?

ANSWER    Use HELP MAP 4: Quelea sometimes breed close to waterbodies in areas where the vegetation maps would suggest that it is most unlikely. For instance, there are many records of colonies along the Orange River. Help Map 4 shows the distribution of colonies on a background of the zone where suitable vegetation exists with USGS - NASA information from LDAAC on waterbodies and major rivers.

If a potential site is along such a watercourse but outside the green vegetation zone, it is still possible that quelea may breed there. Provided that there is adequate seed for the birds to feed on, the presence of water is by itself an indicator as quelea need to drink at least once daily and most colonies are within 10km of standing or flowing water, although occasionally they may be up to 30km from water.

Q5.    Is the altitude suitable?

ANSWER    We are hoping in the future to use a digital elevation model using USGS - NASA information from LDAAC to produce a map showing suitability according to altitude but, at this stage, please check the altitude of your zone from a map and if it is above 1200m, then it is very unlikely that quelea will breed there. Most quelea colonies are found between 100 and 1000m above sea level.

Q6.    How can I combine all this information?

ANSWER    At present the best way is to look at all the maps in the sequence that they have been presented here and tick yes or no for your answers to each of questions 1-3. A negative answer to any one of (a) "Have quelea bred here before?" (b) "Is the vegetation suitable?" or (c) "Is the soil suitable?" will mean that quelea are most unlikely to breed in the relevant square.

[Note, for instance, that the vegetation map suggests that nearly all of Zimbabwe is suitable for quelea breeding but that the soil map suggests that substantial areas in that country are unsuitable and this is supported by the breeding occurrences given in Help Map 1. ]