G.F. Maywald and R.W. Sutherst
CSIRO Division of Entomology,
Long Pocket Laboratories,
Indooroopilly, Brisbane, QLD.
Many organisms have not yet reached their potential distribution around the globe. As international trade and tourism increase, the risk of transferring pests and diseases demands ever greater vigilance from quarantine authorities. Artificial introductions of beneficial species are also being used more widely in the biological control of weeds and other pest species. It is possible to predict new areas which a species could colonise by matching the climates of its current distribution with potential areas at risk. In addition, the predicted 'Greenhouse effect' could change the distribution of insects and other organisms.
Early responses to the entry of pest organisms require rapid access to information on their likely ability to survive. After a pest has been found to have been introduced, quarantine measures to prevent its establishment and spread must be put into place rapidly. However, these measures can be expensive and inconvenient. If it is known with some certainty that the introduced organism is unable to survive, or propagate economically damaging populations in the affected location, these expenses can be avoided.
In an earlier paper , we described a computer program, CLIMEX, that produces rapid assessments of climatic suitability of given locations for any nominated species of plant or poikilothermic animal. The program uses a climatic database, and parameters describing the response of individual species to temperature and moisture, to generate an index describing the suitability of any location for permanent colonization by that species. Information in this form is readily interpretable by decision makers.
Since CLIMEX was first described, it has been considerably enhanced in several areas. A major recent enhancement was the addition of a facility to allow adjustments to the meteorological data to facilitate the exploration of possible changes in distributions due to various scenarios arising from the conjectured 'Greenhouse effect'. In addition, two further functions have been added. These are firstly, a function to search the CLIMEX database for locations with climates similar to that of a nominated location's climate, and secondly, the ability to perform simple database management functions on the climatic data files. A version of CLIMEX is now available that can run on small IBM or compatible personal computers, operated by an attractive 'Menu-driven' interface.
Presentation of CLIMEX results has been enhanced by an option that produces a report, with some interpretation of indices, and integrated seasonal graphs and maps. As well as this, the indices can be produced in a format suitable for interfacing with ExperTik, an expert system for application to cattle tick (Boophilus microplus) problems . Some of these improvements are summarised below and a full description is provided in .
CLIMEX is based on the concept that biological populations in most environments experience two seasons in the course of a year. One of these is favourable to the growth of the population, while the other is unfavourable and may jeopardize its survival.
A 'Growth Index', similar to that developed by Fitzpatrick and Nix , is used to describe the potential of the population to increase during the favourable season. This index, calculated weekly, is a product of temperature (TI), moisture (MI) and daylength (DI) indices, all of which are scaled between 0 and 1. Parameter values are optimised for each variable.
The probability of survival through the unfavourable season is described by four annual 'Stress Indices', which model the response of the organism to cold (CS), hot (HS), dry (DS) and wet (WS) conditions. An additional 4 parameters allow modelling of species, whose response is dependent on interactions between extremes of temperature and moisture. For example, an animal may not tolerate very wet and cold conditions, while similar levels of either wet or cold on their own may have much less effect on survival. The effect of these interactions is added to the corresponding stress indices.
The weekly growth indices are summed and scaled between 0 and 100 to give an annual Growth Index (GI). This index is then combined with each of the stress indices as shown below to obtain the 'Ecoclimatic Index', EI.
EI = GI x (1-CS) x (1-HS) x (1-DS) x (1-WS)
PThe EI is thus a single number that describes the climatic suitability of each location for that species.
Over the past few years, CLIMEX has been applied to various quarantine and biological control problems, both in Australia and overseas. Quarantine applications have covered such diverse groups as ticks [1, 5, 6, 7, 8], biting and myiasis flies [1, 9], fruit flies , aphids, Colorado potato beetle  and the European wasp. CLIMEX has also been used to determine suitable areas for introducing dung beetles into Australia. The way that CLIMEX is used is illustrated below with one quarantine-related example and another showing its application as a research tool in biogeography.
Screw-worm flies are a serious potential threat to Australia's livestock industries in the tropics and sub-tropics. Gravid females lay eggs in wounds or body orifices of their warmblooded host, and the larvae burrow into the host's tissues. This produces serious lesions, and a resulting loss in productivity, and even death, of the host. The Old World species, Chrysomya bezziana, is distributed throughout central Africa and southern Asia, including New Guinea. In recent years, adults of C. bezziana have been found on livestock vessels returning from various Asian ports to Perth, Portland and Darwin, and on passenger aircraft arriving in Sydney from India. The fly thus presents a real and serious risk to Australia .
In order to determine which areas of Australia are at risk, CLIMEX parameter values were fitted using the known distribution of the fly. This is an iterative procedure and is the most common method used to fix species parameter values. CLIMEX was run to generate predictions using initial estimates for the parameter values, and the resulting predictions were compared with the known distribution. The parameters were then adjusted to take these deviations into account, and the procedure repeated until the CLIMEX predictions agreed satisfactorily with the known distribution (Fig. 1)
Once the parameters have been fixed, CLIMEX can be run with meteorological data from different areas to predict the potential distribution of the species. The predictions for C. bezziana in Australia are shown in Fig. 2a, indicating that the fly could establish itself permanently in the wet eastern and northern areas of the continent. CLIMEX produces graphs of the weekly growth and temperature indices (Fig. 3). Such seasonal graphs indicate the chances of establishment of new species arriving at different times of the year.
As an illustration of the use of the 'Greenhouse' option of CLIMEX, Fig. 2b shows the projected potential distribution of C. bezziana assuming average temperatures 3deg. Centigrade above present averages, and increases in summer rainfall of 20%. It is evident that the fly could establish much further south than would be expected under average historical conditions. At the same time, some presently favourable areas in the north would become less suitable for the fly.
The cattle tick is an introduced pest in northern Australia. Its CLIMEX parameter values were fitted using the Australian distribution, which has reached its climatic potential except in New South Wales, where quarantine restrictions prevent the spread of the tick. This tick has also been introduced into southern Africa, where it has had sufficient time to reach a stable distribution. When CLIMEX was run for southern Africa, it was found that the cattle tick was absent from a large area predicted to be suitable. A local tick, Boophilus decoloratus, occurred in those areas.
Although the two species could interbreed, their progeny were sterile . Further investigation  revealed that hybrid sterility was the most likely mechanism that restricted the spread of B. microplus, rather than unsuitable climates as previously thought. This example has raised questions concerning the geographical distribution of many other species of ticks around the world. CLIMEX has thus provided a means for identifying such situations by independently determining the climatic requirements of each species.
CLIMEX operations can be grouped into four main functions described below. A major recent development of CLIMEX has been the addition of a simplified and attractive user interface. Users select the required functions and options from a series of graphic menus. The main menu allows selection of one of the four functions ('Compare Locations', 'Compare Years', 'Match Climates', and 'Manage Meteorological Data'), and each function then has its own menu screen to allow specification of file names and input and output requirements.
Comparing the favourableness of the climates of different locations is the CLIMEX function that was used in the examples presented earlier in this paper. Indices are derived that describe the relative favourableness of different locations for a species. Longterm average meteorological data from monthly temperatures, rainfall and relative humidity is used. Output is produced in the form of tables of indices, graphs of seasonal changes in the temperature and growth indices, and maps of the areas of interest.
The tables of results can be produced in several levels of detail, with the least detailed giving only annual indices, while the most detailed tables produce listings of all the weekly indices. One format of these results has been designed to be used as input to the cattle tick expert system, ExperTik.
Weekly graphs of the meteorological data and the temperature and growth indices can be produced on the screen, and selected graphs sent to a printer or plotter.
Maps of indices are drawn by a suitable mapping program, such as MAPROJ , from a coordinates file produced by CLIMEX. The indices are drawn as circles, whose areas are proportional to the value of the indices. Known distributions can also be drawn on this map as shaded or coloured areas.
CLIMEX can produce its results in report format. The parameter values are embedded in a text that provides some interpretation of these values in terms of the driving climatic variables. Tables, figures and maps required can be specified in a 'Report Directives' file. All of these are produced in correct format and labelled in a single run of the program. The text of the report can be edited using any editor or word processor, with any required extra information being added at this stage.
Comparing the favourableness of different years at the same location is similar to the function described above, but here the relative favourableness of different years at the same location is compared. Hence the meteorological data required is monthly or weekly data from consecutive years. This function is useful in determining if, for example, the abundance of a species in a particular year can be explained by climatic conditions, or some other factor.
The facility to match climates without reference to a particular species can be used to search for locations with climates similar to that of a given location. A 'match' index is generated using a least-squares technique that takes into consideration the closeness of match of maximum temperature, minimum temperature, rainfall and evaporation. A set of weighting parameters can be used to give emphasis to any of these variables. The output consists of a list of locations that best match the given location, with an index describing the goodness of match for each location. Graphs of temperatures and rainfall can be produced for each of these locations as overlays on a graph of the target location to enable a visual assessment of the closeness of the match to be made. The match indices can also be plotted on maps.
To make effective use of CLIMEX, it is necessary to maintain a large database of meteorological data. CLIMEX provides facilities for the addition, extraction or deletion of data, as well as for the conversion of daily meteorological data (which is the usual form in which meteorological data is obtained) to the weekly or monthly averages.
We plan to develop CLIMEX further in several ways. Firstly, we hope to incorporate the ability to handle non-climatic factors such as soil types, or the distribution of plant or animal hosts for parasites. The former would be especially useful for extending application of the program to plants. Each of these factors would constitute an additional dimension in the database, with the climatic suitability indices being modified by the presence or absence of other factors. Secondly, the parameter-fitting process is currently the most laborious step in the use of CLIMEX on a new species. It could be aided, especially for people with minimal computing skills, by extending the interactivity of CLIMEX. Much more use could be made of graphics, for example, to present such information as the changes in the predicted abundance from small changes in parameter values. Steps towards the other end of the task, interpretation of parameter values and indices, have already been taken with the report generator which will be enhanced progressively. Thirdly, we hope to expand the meteorological database to include measures of variability, and the geographic database to allow regional Greenhouse scenarios to be generated in order to handle expected regional variation in climate change.
 Sutherst, R.W. & Maywald, G.F. (1985). A computerised system for matching climates in ecology. Agric. Ecosystems Environ., 13, 281-99.
 Sutherst, R.W. & Bottomley, W. (1989). ExperTik: Experiences in the evaluation of expert systems for pest management. This Volume.
 Maywald, G.F. & Sutherst, R.W. (1989). User's Guide to CLIMEX, a computer program for comparing climates in ecology. 2nd Edition. CSIRO Aust. Div. Entomol. Rep. No. 35, pp. 1-35.
 Fitzpatrick, E.A. & Nix, H.A. (1970). The climatic factor in Australian grassland ecology. In: R.M. Moore (Editor), Australian Grasslands. Australian National University Press, Canberra, pp. 3-26.
 Besier, R.B. & Wroth,R. (1985). Discovery of the tick Haemaphysalis longicornis in Western Australia. Aust. vet. J., 62, 205-6.
 Maywald, G.F. & Sutherst, R.W. (1987). Ecological Models. 1. Assessing climatic favourability with CLIMEX In Sutherst, R.W. (ed). Ecology and Epidemiology of Ticks and Tick-Borne Diseases. ACIAR Proc. No. 17, ACIAR, Canberra, pp. 68-71.
 Sutherst, R.W. & Maywald, G.F. (1987). Modelling tick management. 2. The role of CLIMEX. In Sutherst, R.W. (ed). Ecology and Epidemiology of Ticks and Tick-Bourne Diseases. ACIAR Proc. No. 17, ACIAR, Canberra, pp. 137-40.
 Sutherst, R.W. (1987a). The role of models in tick control. In Hughes, K.L. (ed). Proc. Int. Conf. Vet. Preventiue Med Anim. Prod., Australian Veterinary Association, Melbourne, pp. 32-37.
 Sutherst, R.W., Spradbery, J.P. & Maywald, G.F. (1989). The potential geographical distribution of the Old World screw-worm fly, Chrysomya bezziana. Medical and Veterinary Entomology, 3, (in press)
 Worner, S.P. (1988). Ecoclimatic assessment of potential establishment of exotic pests. J. Econ. Entomol., 81, 973-83.
 Spickett, A.M. & Malan, J.R. (1978). Genetic incompatibility between Boophilus decoloratus (Koch, 1844) and Boophilus microplus (Canestnni, 1888) and hybrid sterility of Australian and South African Boophilus microplus (Acarina: Ixodidae). Ondestepoort J. vet. Res., 45, 149-53.
 Sutherst, R.W. (1987b). The dynamics of hybrid zones between tick (Acari) species. Int. J. Parasitol., 17, 921-26.
 Hutchinson, M.F. (1981). MAPROJ--A computer map projection system. CSIRO Aust. Div. Land Use Res. Tech. Pap. No. 39, 1-38.