a geographic information system on the potential

A geographic information system on the potential

distribution and abundance of Fasciola hepatica

and F. gigantica in east Africa based on Food

and Agriculture Organization databases

J.B. Malonea,*, R. Gommesb, J. Hansenb, J.M. Yilmac,J. Slingenbergb, F. Snijdersb, F. Nachtergaeleb, E. Atamanb

a School of Veterinary Medicine, Louisiana State University, Baton Rouge, LA 70803, USAb Food and Agriculture Organization, Viale della Terme di Caracalla, Rome 00100, Italy

c Faculty of Veterinary Medicine, Addis Ababa University, P.O. Box 34, Debre Zeit, Ethiopia

Received 23 April 1997; accepted 23 February 1998

Abstract

An adaptation of a previously developed climate forecast computer model and digital

agroecologic database resources available from FAO for developing countries were used to

develop a geographic information system risk assessment model for fasciolosis in East Africa, a

region where both F. hepatica and F. gigantica occur as a cause of major economic losses in

livestock. Regional F. hepatica and F. gigantica forecast index maps were created. Results were

compared to environmental data parameters, known life cycle micro-environment requirements and

to available Fasciola prevalence survey data and distribution patterns reported in the literature for

each species (F. hepatica above 1200 m elevation, F. gigantica below 1800 m, both at 1200±

1800 m). The greatest risk, for both species, occurred in areas of extended high annual rainfall

associated with high soil moisture and surplus water, with risk diminishing in areas of shorter wet

season and/or lower temperatures. Arid areas were generally unsuitable (except where irrigation,

water bodies or floods occur) due to soil moisture deficit and/or, in the case of F. hepatica, high

average annual mean temperature >238C. Regions in the highlands of Ethiopia and Kenya were

identified as unsuitable for F. gigantica due to inadequate thermal regime, below the 600 growing

degree days required for completion of the life cycle in a single year. The combined forecast index

(F. hepatica�F. gigantica) was significantly correlated to prevalence data available for 260 of the

1220 agroecologic crop production system zones (CPSZ) and to average monthly normalized

difference vegetation index (NDVI) values derived from the advanced very high resolution

Veterinary Parasitology 78 (1998) 87±101

* Corresponding author. Tel.: 001 504 346 3232; fax: 001 504 346 5715; e-mail: [email protected]

0304-4017/98/$19.00 # 1998 Elsevier Science B.V. All rights reserved

PII S 0 3 0 4 - 4 0 1 7 ( 9 8 ) 0 0 1 3 7 - X

radiometer (AVHRR) sensor on board the NOAA polar-orbiting satellites. For use in Fasciola

control programs, results indicate that monthly forecast parameters, developed in a GIS with digital

agroecologic zone databases and monthly climate databases, can be used to define the distribution

range of the two Fasciola species, regional variations in intensity and seasonal transmission patterns

at different sites. Results further indicate that many of the methods used for crop productivity

models can also be used to define the potential distribution and abundance of parasites. # 1998

Elsevier Science B.V.

Keywords: Geographic information systems; Climate; Distribution; Fasciola hepatica; Fasciola gigantica; East

Africa; Food and Agriculture Organization

1. Introduction

Geographic information systems (GIS) provide a way to use computers to create,

archive and analyze traditional map data on the epidemiology of disease and to combine

these with global environmental data derived from sensors on board earth observing

satellites. A GIS is created by linking standard computer database `attributes' to map

features represented as layered point, line or polygon `vector' data or to map features

represented as digital `raster' image data pixels (picture elements). Using a new

generation of commercially available digital geographic databases and hand held

geographic positioning systems (GPS) linked to a constellation of satellites, precise point

location and maps can be created for disease applications. Common software/hardware

and commercial data sources for personal computers (PCs) are available to provide

laboratory GIS capability suited for disease investigations at costs affordable by most

academic and government disease control units. These capabilities make it possible to use

computer mapping, remote sensing and GIS to create `health maps' that can be used in

routine disease control programs.

Literature reports on GIS medical and veterinary applications include malaria (Beck et

al., 1994), Rift Valley fever (Linthicum et al., 1987), filariosis (Thompson et al., 1996),

African trypanosomosis (Rogers and Randolph, 1993), theileriosis (Lessard et al., 1990),

Onchocerca (Richards, 1993), Leishmania (Cross et al., 1996), Amblyomma variegatum

(Hugh-Jones et al., 1992), Lyme disease (Kitron et al., 1992), Fasciola (Zukowski et al.,

1991; Malone et al., 1992; Malone and Zukowski, 1992; Malone, 1997), and Schistosoma

(Cross and Bailey, 1984; Malone et al., 1994, 1997).

According to the concepts of landscape epidemiology and the doctrine of nidality

(Pavlovsky, 1966), diseases have natural habitats in the same way as a species: they are

found in focal areas where the spatial distribution of the parasite, host, vector and

required environmental conditions coincide. Inside the distribution range, there are

favorable zones where a high level of abundance is maintained. Near the limits of

distribution there is often a patchy marginal zone of low abundance and epidemiologic

instability. The boundaries of distributions are not strictly fixed and may fluctuate with

climate and other components of the environment. The latter zones would be expected to

change most with relatively small changes in the environment, such as annual weather

variation or long term changes projected by recent studies on global warming.

88 J.B. Malone et al. / Veterinary Parasitology 78 (1998) 87±101

Crop models are well known that describe the suitability of different environments for

cultivation and recommended times for planting, fertilizer and pest control (Van

Velthuizen et al., 1995). It is proposed that similar methods can be developed, with GIS

agroclimatic databases and satellite sensor technology, to describe the distribution range

and suitability of different areas for propagation and transmission of disease, and that this

information can then be used to design control programs and optimize interventions. In

the current study, a previously developed climate forecast system (Malone et al., 1987)

and digital agroecologic database resources available from FAO for developing countries

were used, in a GIS, to develop a control model for fasciolosis in East Africa, a region

where both F. hepatica and F. gigantica occur as a cause of major economic losses in

livestock.

2. Methods and materials

2.1. GIS data resources

The Climate Based Parasite Forecast System developed at Louisiana State University

for mapping the geographic distribution, relative risk of economic losses and annual

variation in transmission intensity of Fasciola hepatica in the Southcentral USA (Malone

et al., 1987) was further developed for use in East Africa using Atlas GIS software (ESRI,

Redlands, CA) and a modification of the forecast system that utilized monthly rather than

daily climate values. The original forecast utilizes daily minimum and maximum

temperature and rainfall data for a given site to calculate F. hepatica risk indices based on

the local Thornthwaite water budget and the growing degree day concept. Growing

degree days (GDD�degrees over 108C minimum for development) are accumulated if

water is present in the top 2.5 cm of soil or multiplied times surplus water (cm) if soil

moisture exceeds field capacity. Thirty-year-average data can be used to calculate the

`Normal' risk for a site; annual variation in risk is calculated using current climate data

and compared to 30-year-average values or prior years (Malone and Zukowski, 1994). A

Fasciola GIS forecast system was developed for the Intergovernmental Authority on

Drought and Development (IGADD) sub-region of East Africa using the forecast and the

following Food and Agriculture Organization (FAO) resources.

2.1.1. CVIEW: IGADD crop production system zones (CPSZ)

CVIEW is an FAO software product used to map over 500 database variables for 1220

agroecologic/administrative zones in Ethiopia, Eritrea, Sudan, Somalia, Kenya, Uganda

and Djibouti. It is designed to view, select and export relevant environmental and crop

production datasets to commercial GIS software systems for specific applications (Van

Velthuizen et al., 1995). Atlas GIS (ESRI, Redlands, Ca) software was selected for use

because of its import compatibility with CPSZ databases.

2.1.2. FAOCLIM world climate databases

FOACLIM is available on a CDROM with worldwide monthly climate data and FAO

agroclimatic data analysis software programs (FAO, 1995).

J.B. Malone et al. / Veterinary Parasitology 78 (1998) 87±101 89

2.1.3. FAO-ARTEMIS/NASA-GSFC satellite image archive

Normalized difference vegetation index (NDVI) derived from the AVHRR of the

national oceanic and atmospheric administration (NOAA) environmental satellite series

for 1981±1991 is accessible on a CDROM. NDVI and estimates of rainfall based on

cold cloud duration (CCD) are used for real-time famine early warning and other

applications using a roof-top data receiver at the Rome headquarters of FAO with

Addapix (Griguolo, 1996) and image display and analysis (IDA) software (Hoefsloot,

1996) developed by FAO. Mean monthly NDVI values derived from the entire 1981±

1991 database (based on 7 km2 pixel size spatial resolution at earth surface) were

interpolated to provide a mean value for the area covered by each CPSZ in the CVIEW

database.

2.1.4. FAO soil type databases (1:1,000,000 scale) for the IGADD sub-region

(Nachtergaele, 1996)

Map file were extracted from this database on potentially waterlogged soils of neutral

or alkaline pH and acid soils of <5.5 pH.

2.2. Prevalence data

Prevalence data for Fasciola was entered for corresponding CPSZ mapunits from

available literature reports for Kenya, Sudan and Ethiopia (Gemechu and Mamo, 1979;

Graber et al., 1978; Karib, 1962). Reported prevalence was assigned to all CPSZ included

in the survey area described by authors.

2.3. Creation of atlas GIS project files

Excel files containing CPSZ data, results of climate forecast calculations and

prevalence data were exported as DBase IV files (.dbf), each column of which was used

by Atlas GIS as values for each of the 1220 CPSZ mapunits for display on the computer

monitor as separate layers (or as products of analysis of several layers/values). CPSZ

boundaries for the 1220 mapunits (polygon files), river/water bodies, railroads, roads

(line files) and capitals±towns±villages (point files) were exported as longitude and

latitude, comma delimited (.bna) files, then imported into Atlas GIS as (.agf) map files.

Map files were linked to database (.dbf) attribute data according to name (identical to

names of the 1220 mapunits) and then analyzed within Atlas GIS. An IGAAD soil

database (.bna) file on potentially waterlogged soils of neutral or alkaline pH and soils

with pH <5.5 were also imported into Atlas GIS as a separate (.agf) file. Fasciola

prevalence data was entered as a column in Excel and exported, with data from the CPSZ

database and forecast calculations, to Atlas GIS.

Regional F. hepatica and F. gigantica forecast index maps were created. Results were

compared to environmental data parameters and to available prevalence survey data and

distribution patterns reported in the literature (Bergeron and Laurent, 1970) for each

species (ie. F. hepatica above 1200 m elevation, F. gigantica below 1800 m, both species

at 1200±1800 m). Statistical analysis was done using the Spearman rank-order correlation

test and linear regression analysis (Bruning and Klintz, 1977).


3. Results

3.1. Monthly forecast adaptation

Initially, the FAOCLIM database was used to create a `typical year' for 27 representa-

tive sites using monthly 30-year-average data on minimum and maximum temperature

and number and intensity of rainfall events per month (Landsberg, 1972) using the

original daily forecast system. This proved too time intensive for estimating daily

baseline values for comparison to daily forecast indices. The climate forecast formula

was, therefore, adapted for large scale regional use using monthly climate data and

average annual mean temperature for each of the 1220 mapunits in the CPSZ database.

�GDD�days in month; IF �Rÿ �PET� 0:8� > 0��GDD� 6��Rÿ PET�=25��; IF Rÿ PET > 0� � Index

where R�rainfall, PET�potential evapotranspiration (calculated by the modified Penman

method) and GDD�average annual mean temperature ± base development temperature

for F. hepatica or F. gigantica.

In the first part of the formula, subtracting the factor (PET*0.8) from rainfall is

equivalent to counting monthly GDD if moisture storage is present in the top 2.5 cm layer

of a soil water budget model. The second part counts GDD if monthly surplus water is

present due to rainfall events; GDD is multiplied by 6 based on monthly records that

indicate that 6 rainydays/month with over 1 mm of rain are typical in areas that have a

rainy season that results in surplus water (i.e. most areas of Ethiopia). For use at specific

sites in the IGADD zone or elsewhere, this value can be adjusted based on 30-year rain

pattern data.

Climate data was selected from the CPSZ database and exported to an Excel 5.0

spreadsheet for monthly rainfall and potential evapotranspiration (PET), average annual

mean temperature, NDVI (mean monthly value based on the 1982±1991 database),

growing season (length, duration, beginning, end), altitude and irrigation status. Excel 5.0

was then used to calculate forecast values and to perform initial statistical analysis.

F. hepatica risk index (a temperate zone species) was calculated using a GDD base

temperature of 108C. For the tropical species F. gigantica, risk index was calculated

similarly but to the reported base temperature of 168C (Dinnik and Dinnik, 1963). The

average annual mean temperature was used since mean temperatures are relatively

constant year-round at IGADD sub-region latitudes; this value is successfully used in

regional crop forecast models in lieu of daily or monthly temperature data (Van

Velthuizen et al., 1995). The forecast revealed indices of over 6000 in Ethiopian highland

areas that receive major rainfall, a value that reflects the relative severity of the fasciolosis

problem in that country; indices in fluke zones or the southern gulf coast of the USA

exceed indices of 3000 only in very high risk years (Malone and Zukowski, 1994).

3.2. GIS model output

Using separate models for F. gigantica and F. hepatica (Figs. 1 and 2), fasciolosis risk

gradients were identified in the IGADD sub-region on the basis of 30-year-average


monthly climate data. For both models, the greatest risk occurred in areas of extended

high annual rainfall associated with high soil moisture and surplus water, with risk

diminishing in areas of shorter `wet' season and/or lower temperatures. Arid areas

were generally unsuitable (except where irrigation, water bodies or floods occur) due to

soil moisture deficit and/or, in the case of F. hepatica, high average annual mean

temperature.

For F. gigantica, regions in the highlands of Ethiopia and Kenya were identified as

unsuitable due to inadequate thermal regime (i.e. below the 600 growing degree days

required for completion of life cycle in a single year); this result is consistent with

literature reports of a dearth of F. gigantica over 1800 m elevation in Ethiopia (Bergeron

and Laurent, 1970).

F. hepatica endemic areas in the highlands of Ethiopia and Kenya could be identified

and characterized as to relative risk and geographic distribution. In initial forecast output,

Fig. 1. Fasciola gigantica potential distribution and abundance in the IGADD sub-region based on a GIS

constructed from FAO CVIEW agroecologic zone map files, 30-year-average monthly climate databases, a

modification of the LSU climate based parasite forecast system, a base life cycle development temperature of

168C and known irrigation zones. Irrigated areas and flood zones, also suitable for F. gigantica, were not

included in the climate forecast analysis.


`false positive' areas were seen in the lowlands of Sudan and Uganda. Further iterative

GIS analysis suggested that soil acidity regime in Southwestern Sudan and Uganda (<5.5

pH) may be unsuitable for lymnaeid snail hosts of F. hepatica and/or that tropical thermal

regimes with average annual mean temperature above 238C are unsuited to this temperate

species. Average annual mean temperature of 238C or above were found to correspond to

areas below the reported 1200 m elevation limit of F. hepatica in Ethiopia.

Statistical analysis revealed the sum of forecast indices (F. hepatica index�F. gigantica

index) was correlated (p<0.05) by the Spearman rank-order correlation test with available

prevalence survey data from 260 CPSZ in Ethiopia, Sudan and Kenya (Gemechu and

Mamo, 1979; Graber et al., 1978; Karib, 1962) and by linear regression analysis with

average monthly 1981±1991 NDVI values from the FAO ARTEMIS satellite data archive

Fig. 2. GIS map of F. hepatica potential distribution and abundance gradients for the IGADD sub-region. A base

temperature of 108C for life cycle progression was used. Zones with average annual mean temperatures of

>238C were excluded (shown as areas with horizontal dotted lines). This temperature corresponds to the reported

1200 m elevation lower limit of F. hepatica distribution in Ethiopia. Additional areas in Uganda, Kenya and

Southern Sudan may be unsuitable for Lymnaea truncatula snail hosts based on the presence of acid ferralsol

soils of <5.5 pH (Nachtergaele, 1996).


(Figs. 3±5). The latter suggests that NDVI threshhold criteria might be further developed

as an additional parameter or surrogate for climate values in the forecast.

4. Discussion

Results indicate that the Climate Based Parasite Forecast System, modified for use in a

GIS with CVIEW map viewer databases and the FAOCLIM worldwide monthly climate

database, can be used to define: (1) the potential distribution range of F. hepatica and

F. gigantica, (2) regional differences in intensity of Fasciola spp. transmission, and (3)

seasonal transmission patterns at divergent sites. This information is essential for design

of control programs and to determine the most cost-effective time(s) for treatment.

Current monthly climate values may also be used to generate forecasts of annual

variation in disease risk, although previous experience with FAO crop models shows that

dekadal (10-day-average) data is more suitable for forecast purposes (FAO, 1995). Where

both Fasciola species overlap, the forecast index for each may be summed. Where

Fig. 3. Average monthly normalized vegetation index (NDVI) values�1000 for CPSZ mapunits interpolated

from the FAO-ARTEMIS NDVI Image Bank, Africa, 1981±1991 decadal time series (FAO, 1993). While

NDVI and Fasciola forecast values are statistically related, probably due to common dependence on thermal-

moisture regime, the spatial relationships of the two Fasciola species and average monthly NDVI are not

the same.


available, daily data may be used in the original forecast model on a site-specific basis

where greater precision is required than is possible with monthly or dekadal climate data.

Results indicate the need for more accurate data on the environmental requirements

and limits of tolerance of Fasciola spp., including minimum, maximum, optimum

temperatures, moisture regime and soil characteristics suited for life cycle propagation.

Such information has in the past been generated by laboratory or field microenvironment

studies that define these factors under varying conditions (Appleton, 1978).

In the future, GIS may provide an alternative way to define the range and optimum

conditions for a parasite species by mapping survey data and then iteratively fitting them

to associated climatic and edaphic conditions. A possible upper limit of 238C for F.

Fig. 4. Relationship between reported fasciolosis prevalence (%) and combined annual forecast indices (F.

hepatica � F. gigantica) in 260 of 1220 agroclimatic crop production system zones in the IGADD sub-region of

East Africa, ranked by forecast indices. The Spearman rank-order correlation test showed a significant direct

relationship (p<0.05). Low prevalence rates (1±20%) were seen in the first 20 CPSZ, where no Fasciola risk was

forecast, were located in warm (>258C), arid to semiarid areas; most of them were associated with rivers or

major±minor irrigation which were not considered in the forecast model but which are expected to be important

enzootic sites for Fasciola.


Fig

.5

.L

inea

rre

gre

ssio

nan

aly

sis

show

edsi

gn

ific

ant

corr

elat

ion

(p<

0.0

5)

bet

wee

nco

mbin

edan

nual

fore

cast

indic

es(F

.hep

ati

ca�F

.gig

anti

ca)

wit

hav

erag

em

onth

ly

no

rmal

ized

dif

fere

nce

veg

etat

ion

ind

ex(N

DV

I)b

ased

on

the

FA

OA

RT

EM

ISsa

tell

ite

imag

edat

abas

efo

rA

fric

a,1981±1991,

sugges

ting

that

ND

VI

thre

shhold

crit

eria

cou

ldb

ed

evel

oped

for

reg

ion

alfo

reca

sts

of

fasc

iolo

sis.


hepatica was suggested by the correlation found in the current study with elevations less

than 1200 m in the IGADD zone. While the life cycle may progress with higher

temperatures, continuous or long term temperatures of above 238C over several years

may provide critical limits. This result is consistent with the reported 188C optimum for

this temperate species (Armour, 1975) and suggests the need for further studies on the

limiting effect of higher temperatures on the distribution range of F. hepatica. Similar

conclusions can be derived from the apparent inadequate thermal regime for the tropical

species F. gigantica at cool, high elevation sites. While complete development of this

species requiring two seasons has been reported at high elevations in Kenya where mean

maximum temperatures of 20±308C occur (Preston and Castelino, 1977), such sites are at

the edge of the reported distribution range of F. gigantica; forecast index values, and

prevalence survey data indicate optimum thermal conditions are found at lower elevation,

tropical sites. Dinnik and Dinnik, 1963 reported minimum development temperatures of

168C for F. gigantica in Kenya and found that inhibited cercariae development in infected

snail hosts occurred when mean maximum temperature was below 208C. Schillhorn van

Veen, 1980, however, reported cercariae shedding in the Nigerian savannah in the cool,

dry season (13±188C) and attached less importance to this factor in the total

epidemiology of F. gigantica.

Limited Fasciola prevalence data was available for model development in the present

study. To provide adequate validation data, cooperating scientists in each IGADD country

should be identified for further development and implementation of the Fasciola forecast

GIS and design of integrated parasite control programs. To accomplish this, there is a

need to

� compile published and unpublished prevalence data (eg. thesis, abattoir records, local

meeting reports).

� generate forecast indices to (1) describe a 30-year-average `typical year' in

representative agroecological zones and (2) the range of interannual variation, using

local climate data.

� evaluate efficacy of recommended control programs at selected sites based on seasonal

transmission pattern and current year climate forecasts.

� disseminate model control recommendations to veterinarians and producers, including

early warning of high risk years.

If successful, this application of GIS may serve as a model for developing strategic

control recommendations in other areas of the world where F. hepatica and/or F.

gigantica occur and 30-year-average climate databases are available. Further, the model

may be adapted to develop similar models for other environmentally sensitive diseases,

including tick-borne diseases, trypanosomosis, and other helminth infections, based on

the requirements for propagation and transmission of respective life cycles.

Work reported here may be applicable to strategies for future development of GIS in

animal production and health by FAO and other agencies and for applications in human

health. Broader implementation of GIS technology in control programs for disease will

likely be most effective if centered on PC-based models that allow interaction at user,

developer and central resource levels. Useful medical GIS applications are possible with

relatively simple software packages that are commercially available for use at the PC


level. Many disease control workers could become interactive users of GIS in routine

control programs if it were possible to invest minimal energy learning to access database

viewers, internet web sites and products of GIS models using standard desktop software

resources. Disease control program officers could contribute to shared GIS/database

resources as applications developers in their specific field of expertise and by interactive

database development with field collaborators and GIS support staff at central resource

facilities. Such a three-tier strategy, already operational in FAO crop programs, can be

applied in development of GIS disease control models.

4.1. GIS user level

Viewer software such as CVIEW produced by FAO can be provided inexpensively to

disease control officers, field collaborators and the public on a single disk for PC

operation, allowing users to view, select and export digital map data on a wide range of

environmental variables, crops±crop diseases-forages and livestock±livestock diseases.

Relevant data can be selected and exported to spreadsheets (eg. Excel) for statistical

analysis or linked to exported map files in a GIS for spatial analysis. Enhancements to

CVIEW software under development at FAO (Verelst, 1997; Gommes, pers. comm.) will

allow map display of FAO AGROSTAT-PC databases, development of new applications

and improved data import/export compatibility with most commercial GIS software

packages, speadsheets and presentation graphics. Databases similar to that for the

IGADD Sub-Region are being developed by FAO for Southern Africa and West Africa so

that similar datasets will be available for all of sub-Saharan Africa. PC viewer software is

suitable for users (medical officers, field workers, and the public) with minimal training

(<2 weeks). Such database viewer data for administrative-agroclimatic zones may

provide the basis of improved animal disease and productivity reporting systems, using

`mirror' software at FAO and agriculture ministries in developing countries.

4.2. GIS developer level

Depending on program needs and interest, medical officers who complete entry level

training on commercial GIS packages (1±4 months) may serve as developers for

interactive creation by field specialists and headquarters staff of early warning-

intervention strategies, agricultural statistics analysis, model development and other

applications. Specific GIS applications can be included in database viewers or used in

PC-based GIS analysis packages, spreadsheets and presentation software. Many

commercial PC GIS packages can be linked to FAO database map viewers, to satellite

imagery analysis software products (eg. ERDAS, Atlanta, GA; Idrisi, Newton, MA; IDA,

Rome, Italy) and to advanced resources available at central GIS facilities operational in

many countries and universities.

4.3. Central resource level

To provide higher level UNIX-based and/or PC GIS central resource software/

hardware, data archives (eg. NDVI Image bank, soils, specialized analysis products),


expert assistance, training assistance and project collaboration to support developers

and users. Digital map databases and basemaps are increasingly available worldwide

for purchase (eg. ArcData, Digital Chart of the World) or access over the internet.

Central resource personnel and GIS support staff are needed to create, geographically

transform and link spatial data derived from different GIS software packages.

These advanced systems and applications require large inputs of time and specialized

training.

GIS studies reported here suggest that many of the methods used for crop productivity

models can also be used to define the distribution and abundance of parasites. The

expanding availability of digital databases such as CVIEW, AGROSTAT, FAOCLIM,

internet data sources, improved hardware/software systems for analyzing climate, digital

thematic maps of selected environmental features, and sensor data from earth observing

satellites provide new opportunites for use of GIS models as decision support systems for

disease control programs.

Acknowledgements

This work was partly supported by the FAO Programme of Cooperation with Academic

and Research Institutions during a 4-month sabbatical leave by J.B. Malone at the Animal

Production and Health Division, FAO Headquarters, Rome, Italy. Dr Yilma's

contributions during a 9-month stay at Louisiana State University were supported by

the Fulbright Senior Research Scholar Program of the Council for the International

Exchange of Scholars, United States Information Agency, Department of State,

Washington, D.C.

References

Appleton, C.C., 1978. Review of the literature on abiotic factors influencing the distribution and life cycles of

bilharziasis intermediate snail hosts. Malacological Review 11, 1±25.

Armour, J., 1975. The epidemiology and control of bovine fascioliasis. Vet. Rec. 96, 198±201.

Beck, L.R., Rodrigues, M.H., Dister, S.W., Rodriguez, A.D., Washino, R.K., Roberts, D.R., Spanner, M.A.,

1994. Remote sensing as a landscape epidemiologic tool to identify villages at high risk for malaria

transmission. Am. J. Trop. Med. Hyg. 51, 271±280.

Bergeron, P., Laurent, M., 1970. Differences entre la morphologie testiculaire de Fasciola hepatica et F.

gigantica. Rev. Elev. Med. Vet. Pays Trop. 23, 223±227.

Bruning, J.L., Klintz, B.C., 1977. Computational handbook of statistics. Glenview, IL: Scott, Foresman and

Company.

Cross, E.R., Bailey, C.R., 1984. Prediction of areas of endemic schistosomiasis through use of discriminate

analysis of environmental data. Mil. Med. 149, 542±544.

Cross, E.R., Newcombe, W.W., Tucker, C.J., 1996. Use of weather data and remote sensing to predict the

geographic and seasonal distribution of Phlebotomus papatasi in Southwest Asia. Am. J. Trop. Med. Hyg.

54, 530±536.

Dinnik, J.A., Dinnik, N.N., 1963. Effects of seasonal variations of temperature on development of Fasciola

gigantica in the snail host in the Kenyan highlands. Bull. Epizoot. Dis. Afr. 11, 197±207.

FAO, 1993. FAO-ARTEMIS NOAA AVHRR NDVI Image Bank, Africa, 1981±91. FAO, Rome, Italy. 1 CD-

ROM.


FAO, 1995. FAOCLIM 1.2. A CD-ROM with world-wide agroclimatic data. User's Manual. Agrometerorology

working paper series No. 11, FAO, Rome, Italy, �1 CD-ROM, 68 pp.

Gemechu, B., Mamo, E., 1979. A preliminary survey of bovine fascioliasis in Ethiopia. Eth. J. Agr. Sci. 1(1), 5±

12.

Graber, M., Delavenay, R.P., Tesfamariam, G., 1978. Inventaire parasitologique de l'Ethiopie: Helminthes des

zebus adultes de la region de Kofele (Arussi). Rev. Elev. Med. Vet. Pays Trop. 31(3), 341±352.

Griguolo, S., 1996. Addapix. Pixel by pixel classification for zoning and monitoring. FAO, Rome. SD:GCP/INT/

578/NET. Technical Report. p. 66.

Hoefsloot, P., 1996. IDA for DOS v. 4.2. Image Display and Analysis. User Manual. FAO, Rome. SD:GCP/INT/

578/NET. Technical Report. p. 109.

Hugh-Jones, M.E., Barre, N., Nelson, G., Wheyness, C., Warner, J., Garris, G., Hubbert, W., 1992. Landsat-TM

identification of Amblyomma variegatum (Acari:Ixodidae) habitats in Guadeloupe. Remote Sensing

Environ. 40, 43±55.

Karib, E.A., 1962. Fascioliasis in cattle and sheep in the Sudan. Bull. Int. Epizoot. 85, 337±346.

Kitron, U., Jones, C.J., Bouseman, J.K., Nelson, J.K., Baumgartner, D.L., 1992. Spatial analysis of the

distribution of Ixodex dammini (Acari:Ixodidae) on white-tailed deer in Ogle County. Illinois, J. Med.

Entomol. 29, 259±266.

Landsberg, H.E., 1972. World Survey of Climatology. Climates of Africa. vol. 10, Elsevier, Amsterdam.

Lessard, P., L'Eplattenier, R.L., Norval, R.A.I., Kundert, K., Nolan, T.T., Croze, H., Walker, B., Irvin, A.D.,

Perry, B.D., 1990. Geographical information systems for studying the epidemiology of cattle diseases

caused by Theileria parva. Vet. Rec. 89, 255±262.

Linthicum, K.J., Bailey, C.L., Davies, G., Tucker, C.J., 1987. Detection of rift valley fever viral activity in Kenya

by satellite remote sensing imagery. Science 235, 1656±1659.

Malone, J.B., Zukowski, S.H., 1992. Geographic models and control of cattle liver flukes in the Southern USA.

Parasitol. Today 8, 266±270.

Malone, J.B., Huh, O.K., Fehler, D.P., Wilson, P.A., Wilensky, D.E., Holmes, R.A., Elmagdoub, A.I., 1994.

Temperature data from satellite imagery and the distribution of schistosomiasis in Egypt. Am. J. Trop. Med.

Hyg. 50, 714±722.

Malone, J.B., Williams, T.E., Loyacano, A.F., Muller, R.A., 1987. Fascioliasis in cattle in Louisiana:

Development of a system to predict disease risk by climate, using the Thornthwaite water budget. Am. J.

Vet. Res. 48, 1167±1170.

Malone, J.B., Fehler, D.P., Loyacano, A.F., Zukowski, S.H., 1992. Use of Landsat MSS imagery and soil type in

a geographical information system to assess site specific risk of Fasciola on Red river basin farms in

Louisiana. Ann. NY. Acad. Sci. 653, 389±397.

Malone, J.B., Zukowski, S.H., 1994. Geographic models and control of cattle liver flukes in the southern USA.

Parasitol. Today 8, 266±270.

Malone, J.B., 1997. The landscape epidemiology of fasciolosis: Geographic determinants of disease risk. In: J.C.

Boray (Ed.) Immunology, Pathobiology and Control of Fasciolosis. Proc. ICOPA VIII Roundtable on

fasciolosis, IZMIR, 1994. MSD Ag. Vet. 65±81.

Malone, J.B., Abdel-Rahman, M.S., El-Bahy, M.M., Huh, O.K., Shafik, M., Bavia, M., 1997. Geographic

information systems and the distribution of Schistosoma mansoni in the Nile delta. Parasitol. Today 13(3),

112±119.

Nachtergaele, F., 1996. Soils of the IGADD Sub-region. 1:1,000,000 scale, FAO, Rome. Pers. Comm.

Pavlovsky, E.N., 1966. The Natural Nidality of Transmissible Disease. University of Illinois Press, Urbana, 261

pp.

Preston, J.M., Castelino, J.B., 1977. A study of the epidemiology of bovine fascioliasis in Kenya and its control

using N-tritylmorpholine. Br. Vet. J. 133, 600±608.

Richards, F.O., 1993. Use of geographic information systems in control programs for onchocerciasis in

Guatemala. Bull. Pan Am. Hlth. Org. 27, 52±55.

Rogers, D.J., Randolph, S.E., 1993. Distribution of tsetse and ticks in Africa: past, present and future. Parasitol.

Today 9, 266±271.

Schillhorn van Veen, T.W., 1980. Dynamics of Lymnaea natalensis populations in the Zaria area (Nigeria) and

the relation to Fasciola gigantica infections. Acta Tropica 37, 183±194.


Thompson, D.F., Malone, J.B., Harb, H., Faris, R., Huh, O.K., Buck, A.A., Cline, B.L., 1996. Bancroftian

filariasis distribution and diurnal temperature differences in the southern Nile delta. Emerg. Inf. Dis. 2,

234±235.

Zukowski, S.H., Hill, J.M., Jones, F.W., Malone, J.B., 1991. Development and validation of a soil-based

geographic information system model of habitat of Fossaria bulimoides, a snail intermediate host of

Fasciola hepatica. Prev. Vet. Med. 11, 221±227.

Van Velthuizen, H., Verelst, L., Santacroce, P., 1995. Crop Production System Zones (CPSZ) of the IGADD Sub-

Region. Agrometerorology working paper series No. 10., FAO, Rome, Italy, �1 diskette�maps, GIS, 89 pp.

Verelst, L., 1997. Database Map Viewer. Users manual. FAO, Rome, Italy.


a geographic information system on the potential

Documents