- Email: [email protected]
Health and production measures for smallholder pig production in Kikuyu Division, central Kenya
Preventive Veterinary Medicine 63 (2004) 197–210
Health and production measures for smallholder pig production in Kikuyu Division, central Kenya J.K. Wabacha a,∗ , J.M. Maribei a , C.M. Mulei a , M.N. Kyule b,c , K.H. Zessin c , W. Oluoch-Kosura d a
d
Department of Clinical Studies, University of Nairobi, P.O. Box 29053, Nairobi, Kenya b Department of Public Health, Pharmacology and Toxicology, University of Nairobi, P.O. Box 29053, Nairobi, Kenya c Postgraduate Studies in International Animal Health, Freie Universität Berlin, Luisenstrasse 56, D-10117 Berlin, Germany Department of Agricultural Economics, University of Nairobi, P.O. Box 29053, Nairobi, Kenya
Received 21 August 2002; received in revised form 21 January 2004; accepted 6 February 2004
Abstract A longitudinal study was carried out in Kikuyu Division (a peri-urban area in central Kenyan highlands) between January 1999 and December 1999 to estimate the baseline parameters on reproductive performance of the sow, as well as health and productivity of grower and preweaning pigs of smallholder herds. Data were collected on 155 breeding pigs, 795 grower pigs and 801 preweaning piglets in 74, 50 and 40 smallholder herds, respectively, using record cards that were updated during monthly visits. The sow-level medians were: weaning-to-service interval 3 months; interfarrowing interval 6.4 months; number of live-born piglets 9.0; and number of piglets weaned per litter 7.5. The piglet crude morbidity incidence risk was 29%. The cause-specific incidence risks for the important health problems encountered in preweaned piglets were diarrhea (4.3%), pruritus (17.1%), and skin necrosis (4.2%). The estimated crude mortality incidence risk to 8 weeks of age was 18.7%. The cause-specific mortality incidence risks to 8 weeks of age for the important causes of mortality were overlying (9.9%), savaging (2.4%), unviable piglets (2.0%) and unknown (1.9%). Overall, 78.8% of the total live-born piglet mortality occurred during the first week postpartum with 69% of these deaths being caused by overlying. The grower-pig crude morbidity incidence risk was 20% and the cause-specific incidence risks of the important health problems encountered were gut edema (1.3%), pruritus (21.1%), and unknown (2.3%). The crude mortality incidence risk was 3.8% and the important causes were gut edema and unknown causes (cause-specific mortality incidence risks of 1.3 and 1.6%, respectively).
∗
Corresponding author. Tel.: +254-2-630-451; fax: +254-2-632-059. E-mail address: [email protected] (J.K. Wabacha). 0167-5877/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.prevetmed.2004.02.006
198
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
The median weight:age ratio and average daily weight gain for the grower pigs were 5.1 kg/month of age and 0.13 kg/day, respectively. For preweaning pigs, the median average daily weight gain was 0.13 g/day. © 2004 Elsevier B.V. All rights reserved. Keywords: Smallholder systems; Pigs; Health; Production; Kenya
1. Introduction In livestock production, suitable parameters for assessing health and productivity are those that measure production and also act as economic indicators (Martin et al., 1987). In breeding herds, the number of pigs weaned per sow per year is (from economic point of view) the most important measure of productivity (Radostits et al., 1994). This production parameter depends on litters/sow/year, farrowing rate, culling rate, weaning-to-service interval, number of live-born piglets and preweaning piglet mortality (Dial et al., 1992; Polson, 1996). For grower pigs, average daily weight gain is considered an important productivity measure (Radostits et al., 1994). Studies conducted in commercial piggeries in tropical areas indicate that sow reproductive performance is lower than in subtropical or temperate areas (Wongnarkpet et al., 1994; Kunavongkrit and Heard, 2000; Tantasuparuk, 2000). This poor performance has been attributed mainly to hot climate (leading to lower number of piglets at birth) (Kunavongkrit and Heard, 2000; Tantasuparuk, 2000). Studies on performance of the sow in smallholder herds in tropical regions are few (de Fredrick and Osborne, 1977; Wilkins and Martinez, 1983; Gatenby and Chemjong, 1992; Lanada et al., 1999). These studies demonstrated low performance of the sows as a result of pig diseases, substandard management, difficult in obtaining quality feeds and inappropriate service–delivery systems. Preweaning piglet mortality is an important cause of reduced sow performance in tropical smallholder herds (Wilkins and Martinez, 1983; Gatenby and Chemjong, 1992; Lanada et al., 1999; Taveros and More, 2001). However, information on the causes of preweaning piglet mortality is scanty. Piglet morbidity which contributes directly or indirectly to mortality (Cutler et al., 1997) also is not well studied in tropical smallholder pig herds and the available information is limited to cross-sectional studies that mainly focussed on parasitic diseases (de Fredrick, 1977b; Manuel et al., 1989; Kambarage et al., 1990; Esrony et al., 1997). The grower-pig performance in the smallholder herds in the tropics is also lower as compared to commercial production systems of the subtropics and tropical areas (de Fredrick and Osborne, 1977; More et al., 1999). The poor performance is attributed to inadequate and poor-quality feeds, poor genetics and diseases (de Fredrick and Osborne, 1977; More et al., 1999). Among the diseases, internal and external parasites are common as a result of climatic and management factors (de Fredrick, 1977b; Manuel et al., 1989; Kambarage et al., 1990; Esrony et al., 1997). In Kenya, few studies have been done on sow reproductive performance in smallholder herds (Masembe, 1985; Munyua et al., 1991). In those studies, passively derived data and convenience-sampling methods were used and high preweaning mortality was identified as
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
199
the main factor influencing sow performance. However, the methods used make us question data quality (Cameron, 2000). Studies on grower and preweaning pig health and productivity have not been undertaken in Kenya. To assess the economic importance of pig diseases and to design appropriate disease-control programs, information on productivity, disease incidence and mortality is necessary. We undertook a prospective longitudinal study on smallholder pig herds in a high-potential farming and peri-urban area in central Kenyan highlands. Our main objectives were to determine the reproductive performance of the sow, as well as the health and productivity of grower and preweaning pigs of smallholder herds in Kikuyu Division, central Kenya.
2. Materials and methods 2.1. Study area This study was undertaken in the high-potential farming and peri-urban area in Kikuyu Division of Kiambu District, central Kenya. The study area and the smallholder pig farms and their management practices were reported previously by Wabacha et al. (in press). 2.2. Selection of study farms The selection of the study farms was carried out as described previously (Wabacha et al., in press). 2.3. Data collection The selected herds were visited once-a-month by the first author and two recruited enumerators between January 1999 and December 1999. Data on sow reproductive performance were collected using sow-litter and dry-sow record cards. For the dry sow, the data collected included service date, farrowing date, date weaned and type and amount of feed provided (recorded in an open format). For the lactating sow, the data collected included farrowing date, number of live-born piglets, number of piglets born dead or mummified, date weaned, number weaned, piglet mortality, age at weaning and type and amount of feed provided. The data were recorded on the cards either from the recall of events that occurred between the visits or from records scribbled on the pen walls by some farmers. During the monthly visits, a detailed physical examination of the piglets and grower pigs in those herds with piglets and grower pigs was made and morbidity data recorded. The farmers provided additional information on morbidity and mortality occurring between the visits. Farmers had been motivated to keep close watch on their herds by being offered free treatment for any sick pig reported. Additionally, whenever possible, the first author also attended sick pigs at the farm at the request of the farmers. Clinical diagnosis of sick or dying pigs mainly was based on anti-mortem clinical signs observed during the monthly visits or from clinical history presented by the farmers. The small herd sizes allowed the farmers to follow the health events in their herds in detail. To avoid considering a case more than once, innovative ways of marking the upcoming cases (that included scraping off
200
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
of the hairs) were used—but in most cases, farmers were able to identify individual pigs. Data also were collected on type and amount of feed provided and preventive-medicine practices. During the monthly visits, unweaned piglets and grower pigs (up to a maximum of 40 kg) were weighed to the nearest 0.01 kg using a suspended weighing scale (Salter® , Salbrave, Kenya Limited). The data were recorded on prescribed record cards that were updated during the monthly visits. Rectal fecal samples and ear-wax scrapings from randomly selected preweaning and grower pigs from each of the selected herds were sampled once for fecal egg counts and mite identification, respectively. Fecal samples were taken from 83 preweaning and 204 grower pigs and analyzed using the modified McMaster technique (Anon., 1986). Ear-wax scrapings were taken from 117 preweaning and 154 grower pigs by scraping the inner aspects of the ear until traces of blood could be seen. The materials were examined under the light microscope for the presence of mites after digestion with 10% potassium hydroxide. 2.4. Measures of sow reproductive performance The measures of sow performance were preweaning piglet mortality (PWMT), weaningto-service interval (WTSI), interfarrowing interval (IFI), number of live-born piglets (NLBP), number of piglets weaned per farrowing (NWF), litters per sow per year (L/S/Y) and the number of piglets weaned per sow per year (PW/S/Y) (Dial et al., 1992) (Table 1). 2.5. Measures of health in preweaning and grower pigs Morbidity risk and mortality risk (Martin et al., 1987) were used as measures of health in preweaning and grower pigs. For preweaning pigs, the biologic period of risk was the period from farrowing to weaning while for grower pigs it was the period from weaning to the time the pigs left the herd through sale or mortality. The amount of disease (morbidity) was determined by recording any disease/condition with a recognizable clinical sign(s) while mortality was determined by recording all cases of death in pigs. 2.6. Measures of productivity in preweaning and grower pigs The measure of productivity used for the pigs was average daily weight gain (ADWG) in kg/day; additionally, the ratio of the measured weight and the reported age (weight:age ratio) was used for the grower pigs (Radostits et al., 1994; More et al., 1999). 2.7. Data management Data were entered in the data collection cards at the farm during the visits. Data files were created in Dbase (DBASE IV Plus, Ashton Tate, Torrance, California, USA) and Excel® 1997 (Microsoft Corporation, USA). The files were screened for any errors that might have occurred during the data entry and the errors were corrected by rechecking against the original data collection cards.
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
201
Table 1 Health and productivity measures for pigs in smallholder pig herds in Kikuyu Division, Kiambu District, central Kenya (January 1999–December 1999) Animal and measurement
Sows Weaning-to-service interval (month)
N
Minimum
Percentile
Maximum
25th
50th
75th
33a 22b
0.2 0.2
1.0 1.3
3.0 3.1
4.1 5.1
10.5 10.5
Interfarrowing interval (month)
29a 17b
5.0 5.7
5.9 6.0
6.4 6.9
7.5 7.8
13.0 13.0
Number of live-born piglets
95a 40b
0.0 4.5
7.0 7.6
9.0 9.0
12.0 10.4
20.0 17.0
Number of piglets weaned per farrowing
95a
0.0
5.0
7.5
9.0
14.0
40b
1.5
6.0
7.0
9.0
14.0
95a 40b
1.0 1.0
1.0 1.0
1.0 1.0
1.0 1.2
2.0 2.0
76a 40b
0.0 1.5
6.0 7.0
8.0 8.5
11.0 10.0
23.0 17.0
86c 40d 86c 40d 53c 28d
0 0 0 0 0.02 0.05
0 0 0 2.1 0.07 0.07
0 16 10 12 0.13 0.14
70 78 29 26 0.19 0.2
100 100 100 100 0.4 0.4
116c 50d 116c 50d 39c 17d 396e 26d
0 0 0 0 0.01 0.08 1.0 1.5
0 10 0 0 0.11 0.12 3.8 4.1
21 29 0 0 0.13 0.15 5.1 5.1
33 33 0 2.1 0.21 0.19 6.2 6.0
Litters/sow/year Piglets weaned/sow/year Preweaned piglets Crude morbidity incidence Crude morbidity incidence Crude cumulative mortality Crude cumulative mortality Average daily weight gain (kg/day) Grower pigs Crude morbidity incidence Crude morbidity incidence Crude cumulative mortality Crude cumulative mortality Average daily weight gain (kg/day) Weight:age (kg/month)
87 60 60 48 0.36 0.33 11.9 10.6
a
Sows. Herds. c Litters. d Herds. e Growers. b
2.8. Data analysis Data analyses were performed using Statistix® for Windows (Analytical Software, Tallahassee, FL, USA), Excel® 1997 (Microsoft Corporation, USA) and Minitab Statistical Software, version 13 for windows (Minitab Inc., USA).
202
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
2.8.1. Measures of sow reproductive performance A total of 155 breeding pigs were kept in 74 of the 76 herds that were enrolled for the longitudinal study. A number of parameters were only available for some sows and herds (Table 1) because sows in some herds were withdrawn from the study through death or sale while others did not farrow during the study period. The values of the parameters (section 2.4) for each sow in a herd were used as the sow-level measures of performance while the median values of the parameters for all the sows in the herd were used as herd-level measures of sow performance. 2.8.2. Measures of health in preweaning and grower pigs The cause-specific, age-specific and crude morbidity and mortality cumulative incidences for each litter in a herd were used as the litter-level measures of health. The median values of crude morbidity and mortality cumulative incidences for all the litters in the herd were used as herd-level measures of health. For crude morbidity, any pig that had at least one disease/condition was taken as a case. For the cause-specific morbidity, only new cases arising during the monitoring period were considered. The morbidity and mortality cumulative incidences for the piglets were calculated for the preweaning period of 8 weeks. Data on piglet health were collected in 40 out of the 74 herds with breeding pigs (Tables 1–3) because of the same reasons as given in Section 2.8.1. Grower pigs were kept in 50 out of the 76 herds that were enrolled for the study and the health parameters were determined in all the herds (Tables 1, 4 and 5). For the individual litters of grower pigs, the monitoring period lasted between 2 and 5 months. To take care of the different time-at-risk for the different litters of grower pigs, incidence rates were used to determine the grower-pig morbidities and mortality per grower-pig month-at-risk. Estimated cumulative incidence (CI(t) ) then was calculated using the formula: CI(t) = 1 − e−IR×t , where e is the base of the natural logarithm, and t indicates the time-unit of concern as described in Noordhuizen et al. (1997). 2.8.3. Measures of productivity in preweaning and grower pigs Piglet productivity was determined in only 28 herds (out of 76 with breeding pigs) because in addition to the reasons given earlier, piglets in some herds were weaned before two consecutive weights could be taken. For grower pigs, determination of average daily weight gain (kg/day) and weight:age ratio (kg/month of age) was only possible in 17 and 26 herds, respectively, due to censoring when the grower pigs were sold or when the study ended; others became too heavy to weigh with the suspended weighing scale during the subsequent visits. The average daily weight gain for the pigs in a litter was used as the measure of productivity. The ADWG was calculated as the difference in total litter weight between two consecutive monthly visits divided by the average number of pigs in the litters during the first and second visit and the number of days between the two visits. The ratio of the measured weight and the reported age of each grower-pig (weight:age) in kg/month of age was used as a measure of productivity to increase the number of herds where productivity of grower pigs could be determined, because the number of herds where ADWG could be determined was small. For each of the productivity measures, ADWG and weight:age ratio, the median of the parameters were used as herd-level measures of productivity.
Disease/condition
Callus on carpus Deformitya Diarrhea Herniaa Loss of claw Pneumonia Pruritis Skin necrosis Unthriftiness Unknown Age-specific Incidence risk (%) a
Age (weeks) 0 to <1
1 to <2
2 to <3
3 to <4
4 to <5
5 to <6
6 to <7
7 to <8
0 1 6 1 0 0 0 16 4 2
0 0 9 0 0 0 3 3 0 2
8 0 9 0 0 0 0 6 0 0
9 0 0 0 0 0 28 1 0 2
0 0 0 0 0 0 33 0 1 1
0 0 4 0 0 4 27 1 0 0
0 0 8 0 0 2 29 8 0 1
0 0 0 0 1 2 14 0 0 0
3
2
3
4
5
6
7
3
The cumulative incidence was not calculated because the conditions were present at birth.
Litter-level incidence risk (%) 2.4 – 4.3 – 0.14 1.3 17.1 4.2 1.3 0.77
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
Table 2 The distribution of morbidity cases by clinical manifestations, age in weeks, cause-specific and age-specific cumulative incidence for 801 preweaning piglets in 40 out of 76 smallholder pig herds in Kikuyu Division, Kiambu District, central Kenya (January 1999–December 1999)
203
204
Disease/condition
Overlying Pneumonia Predation Savaginga Scours Starvation Unviable piglets Unknown Age-specific Incidence risk (%) a
Age (weeks)
Litter-level crude cumulative mortality (%)
0 to <1
1 to <2
2 to <3
3 to <4
4 to <5
5 to <6
6 to <7
7 to <8
87 0 0 20 0 4 13 2
0 0 8 2 0 3 0 2
0 0 0 0 0 0 0 2
0 0 0 2 0 0 0 1
0 0 0 0 0 0 0 1
0 2 0 0 1 0 0 1
0 0 0 0 0 0 0 5
0 0 0 0 0 0 0 4
9.9 0.2 0.8 2.4 0.1 1.2 2.0 1.9
14.3
2.5
0.2
0.6
0.1
0.6
0.7
0.6
–
The 2 piglets in the fourth week were just found missing and the cause could not be ascertained.
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
Table 3 The distribution of mortality cases by clinical manifestations, age in weeks, cause-specific and age-specific cumulative mortality for 801 preweaning piglets in 40 out of 76 smallholder pig herds in Kikuyu Division, Kiambu District, central Kenya (January 1999–December 1999)
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
205
Table 4 The distribution of morbidity cases by clinical manifestations, age in months, cause-specific incidences and crude cumulative incidences for 795 grower pigs in 50 out of 76 smallholder pig herds in Kikuyu Division, Kiambu District, central Kenya (January 1999–December 1999) Disease/condition
Abscess Diarrhea Ear necrosis Gut edemac Pneumonia Pruritis Unthriftiness Unknown
Age (months) 2 to <3
3 to <4
4 to <5
5 to <6
6 to <7
7 to <8
0 4 0 14 12 133 1 19
0 1 0 0 0 95 2 2
0 0 1 0 0 31 0 3
1 10 0 0 0 17 0 0
0 0 0 0 0 24 0 0
0 0 0 0 3 18 0 1
IRa
CIb
0.1 0.22 0.001 1.3 0.4 23.6 0.2 2.1
0.1 0.22 0.001 1.3 0.4 21.1 0.2 2.3
a
Litter-level cause-specific morbidity incidence (%) per pig-month-at-risk. Estimated litter-level cause-specific morbidity incidence risk (%). c Neurological disorder, characterized by hoarse squeal, ataxia, swelling of the eyelids drooping ears, paresis and paralysis was tentatively diagnosed as gut edema. b
Table 5 The distribution of cases by clinical manifestations, age in months, cause-specific and cumulative mortalities for 795 grower pigs in 50 out of 76 smallholder pig herds in Kikuyu Division, Kiambu District, central Kenya (January 1999–December 1999) Disease/condition
Gut edemac Diarrhea Pneumonia PSSd Unknown
Age (months) 2 to <3
3 to <4
4 to <5
5 to <6
6 to <7
7 to <8
13 0 6 0 3
0 4 6 0 10
0 1 0 0 2
0 0 0 0 4
0 0 0 1 0
0 0 0 0 1
IRa
CIb
1.3 0.19 0.3 0.1 1.7
1.3 0.19 0.3 0.1 1.6
a
Litter-level cause-specific mortality (%) per pig-month-at-risk. Estimated litter-level cause-specific mortality (%). c Neurological disorder, characterized by hoarse squeal, ataxia, swelling of the eyelids, drooping ears, paresis and paralysis was tentatively diagnosed as gut edema. d Porcine stress syndrome, tentative diagnosis based on the history of death after exhaustion. b
3. Results 3.1. Measure of sow reproductive performance The measures of sow reproductive performance are shown in Table 1. At least 25% of the sows and herds achieved a WTSI of not more than 1 and 1.3 months, respectively. Twenty-five percent of the sows and herds achieved an IFI of not more than 5.9 and 6 months, respectively.
206
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
3.2. Diseases occurrence The diseases encountered in preweaning piglets were diarrhea, callus on carpus, loss of claws, pneumonia, pruritis, skin necrosis and unthriftiness. The prevalence for mange on mite detection was 10.3% while that for helminthosis was 27.7%. For grower pigs, the diseases encountered were abscesses, diarrhea, ear necrosis, gut edema, pneumonia, pruritis, unthriftiness, and other diseases whose tentative diagnosis could not be made. The prevalence for mange on mite detection was 19.5% while that for helminthosis was 50.5%. 3.3. Measures of health in preweaning and grower pigs The litter- and herd-level median crude morbidity cumulative incidences for the preweaning pigs were 0 and 16%, respectively (Table 1). The litter-level cause-specific cumulative incidence from birth up to 4 weeks was highest for skin necrosis (3%) and diarrhea (2.5%). From the fourth week postpartum to weaning, the cause-specific morbidity risk was highest for pruritis (17.1%) (Table 2). The litter- and herd-level median crude cumulative mortality of preweaning piglets was 10 and 12%, respectively (Table 1). At least 25% of the herds had a preweaning piglet mortality of not more than 2.1%, although mortality in some herds was as high as 100%. Preweaning piglet mortality was highest during the first week postpartum. Among the piglets born alive, overlying was the most frequent cause of death (9.9%) followed by savaging (3%) (Table 3). Overall, 78.8% of the total live-born piglet mortality occurred during the first week (with 69% of these deaths being caused by overlying). For the grower pigs, the litter- and herd-level median crude morbidity cumulative incidences during the period of the study were 21 and 28.5%, respectively (Table 1). The diseases with the highest incidence risk were pruritis (21.1%), gut edema (1.3%) and diseases of unknown causes (2.3%) (Table 4). At least 50% of litters and herds had crude cumulative mortalities of less than zero (Table 1). Among the causes of grower-pig mortality, gut edema (1.3%) and unknown causes (1.6) had the highest incidence (Table 5). 3.4. Measures of productivity in preweaning and grower pigs The measures of productivity in preweaning and grower pigs are shown in Table 1. The litter- and the herd-level median ADWG for the preweaning pigs were 0.13 and 0.14 kg/day, respectively. For grower pigs the median animal-level and herd-level weight:age was 5.1 kg/ month of age. The litter- and herd-level median ADWG was 0.13 and 0.15 kg/day, respectively.
4. Discussion The reproductive performance of the sow was low as compared to commercial piggeries in tropical countries including Kenya (Kabare, 1991; Wongnarkpet et al., 1994; Kunavongkrit and Heard, 2000; Tantasuparuk, 2000). However, the number of live-born piglets observed
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
207
was similar to the one observed in commercial piggeries in Kenya and other tropical regions (Kabare, 1991; Lanada et al., 1999; Kunavongkrit and Heard, 2000). The climatic stress due to high temperatures (especially from weaning until early gestation) can cause low litter size in tropical regions (Tantasuparuk, 2000). The NWF was lower than in commercial piggeries in tropical areas (Kabare, 1991) and this could be attributed to the high preweaning mortality observed. Similar observations have been made in other smallholder herds (de Fredrick, 1977a). The WTSI (median 3.0 months) observed in this study was much longer than observed in commercial tropical piggeries (Kunavongkrit and Heard, 2000; Tantasuparuk, 2000) and this could have contributed to the prolonged IFI observed in this study. The prolonged IFI (median 6.4 months), was in agreement with other observations in smallholder herds (Wilkins and Martinez, 1983; Gatenby and Chemjong, 1992; Lanada et al., 1999; de Fredrick and Osborne, 1977). The IFI is influenced by the lactation length, number of non-productive days and the gestation length (Dial et al., 1992; King et al., 1998). The non-productive days could be prolonged by poor nutrition offered to the sows and non-availability of the boar. The prolonged IFI could contribute to low number of litters weaned per sow per year. For the preweaning pigs, the cause-specific morbidity due to facial-skin necrosis, pruritis and scours were high and this was consistent with previous findings (Straw et al., 1997; Cameron, 1997; Kambarage et al., 1990). Facial necrosis commonly is observed during the first week postpartum and is as a result of infection by Staphylococcus hyicus and Fusobacterium necrophorum of wounds inflicted by piglets on each other during feeding (Cameron, 1997). Several factors cause pruritis in pigs (Davies, 1995; Hollanders et al., 1995; Cargill and Davis, 1997). Sampling of randomly selected pigs revealed high prevalence with sarcoptic mange with no other ectoparasites being detected; it thus appears that the pruritis in the studied herds was caused by sarcoptic mange. The preweaning piglet mortality (median 10%) observed in this study differs from that reported by Munyua et al. (1991) and Masembe (1985). They reported mean preweaning piglet mortalities of live-born piglets of 15.8 and 17.1%, respectively. The difference between the studies done in Kenya could be due to the sampling strategies used. We used a random sample of smallholder herds while the previous studies were based on convenient samples that also included both medium and smallholder herds. The high preweaning piglet mortality we observed was consistent with findings in other tropical smallholder herds (Lanada et al., 1999; Wilkins and Martinez, 1983; Gatenby and Chemjong, 1992; Taveros and More, 2001). Pig-management practices in smallholder herds are low (Lanada et al., 1999) and this could explain the high preweaning mortality observed in the tropical smallholder herds. Over three quarters of the overall mortality (78.8%) occurred during the first week of life and more than half of these deaths (69%) were caused by “overlying”. Although we classified this as a category on its own, “overlying” is interrelated with other factors—for example, a piglet which is getting little milk is likely to become weak and be “overlaid” by the sow (Vaillancourt and Tubbs, 1992; Cutler et al., 1997). The pattern of the causes of preweaning piglet mortality observed in this study differs from that observed in a previous study in Kenya (Munyua et al., 1991) in which gastrointestinal syndrome, starvation and pneumonias were the most important causes of preweaning piglet mortality in small and medium scale pig
208
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
herds. Those authors used convenient samples and relied more on secondary data than we did. The classification of the causes of preweaning mortality in this study was mainly based on the pig owners recorded causes of mortality between the monthly visits. These causes might need validation by performing post-mortem examinations (Vaillancourt et al., 1990; Vrbanac et al., 1995; Cutler et al., 1997). Due to the nature of the study design, this could not be accomplished. The observed crude morbidity risk (median 21%) of grower pigs seemed high to us. High morbidity risk would compromise the performance of grower pigs by contributing directly or indirectly to mortality or by reducing feed-conversion efficiency (Radostits et al., 1994). The high cause-specific morbidity due to pruritis (sarcoptic mange) was in agreement with findings in other tropical smallholder pig herds (Kambarage et al., 1990). Pruritis in grower pigs, due to sarcoptic mange, is a common problem especially in pigs kept in unhygienic conditions and where mange control measures are inadequate (Cargill and Davis, 1997). In this study, many cases of a neurological disorder (suspected to be gut edema from clinical presentation) were observed. Gut edema is a common problem in grower pigs, especially immediately after weaning (Radostits et al., 1999). The “unknown causes” contributed important morbidity of the grower pigs and they deserve further research to identify them. The other diseases recorded (for example, ear necrosis, unthriftiness and abscesses) are common disease/conditions in pigs (Straw et al., 1997). In the current study, the cause-specific morbidity for pneumonia’s and diarrhea were low and this was in agreement with findings in other tropical smallholder pig-production systems (de Fredrick, 1977b). The grower-pig crude mortality risk observed was high. The “unknown diseases” contributed the highest mortality risk and future recommendations to specifically delineate them are indicated. The high mortality risk of grower pigs would reduce grower-pig productivity in the smallholder herds. The median weight:age ratio observed (5.1 kg/month of age) was in agreement with observations in tropical smallholder herds in the Philippines (More et al., 1999). Poor growth of pigs in smallholder herds has been attributed to poor feeding (both in quality and quantity) (de Fredrick and Osborne, 1977; More et al., 1999). The data collected on preweaning and grower pigs morbidity and mortality mainly depended on farmer-reported morbidity and mortality. To ensure that the information was correct, the data were validated by gathering a detailed clinical history of pig morbidity and mortality at the farm during the monthly visits. However, it is still possible that there may have been some under reporting (especially with morbidity). Information based on farmer-reported morbidity and mortality is destined to have some errors (Losinger et al., 1998). Nevertheless, we believe that the data collected were representative of the diseases currently experienced on the smallholder pig herds in the study area. In conclusion, the reproductive performance of the sow (weaning-to-service interval, interfarrowing interval and number of piglets weaned per farrowing) was poor. The health and productivity of pigs in the studied smallholder herds were also suboptimal with the major health problems being mange and helminthosis. The variation in performance of the various herds indicates that an opportunity exists to raise productivity in the studied herds.
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
209
Acknowledgements We thank the smallholder pig farmers in Kikuyu Division, Kiambu District, for their enthusiasm throughout the study. The effort of the livestock extension officers (Ministry of Agriculture and Rural Development, Kikuyu Division) who facilitated the link with the farmers during farm recruitment and the subsequent visits is highly appreciated. This study was supported by Agricultural Research Fund (ARF) contract no. ARF/LSKP/1001004/1 and Akademischer Austausch Dienst (DAAD).
References Anon., 1986. Manual of Veterinary Parasitological Laboratory Techniques. Reference Book No. 418, 3rd ed. Ministry of Agriculture, Fisheries and Food. Helminthology, Her Majestys Stationery Office, London, pp. 1–68. Cameron, R.D.A., 1997. Diseases of the skin. In: Straw, B.E., D’Allaire, S., Mengeling, W.L., Taylor, D.J. (Eds.), Diseases of Swine, 8th ed. Blackwell Science, Osney Mead, Oxford, pp. 941–958. Cameron, A., 2000. Strategies for epidemiological surveys in developing countries. In: Proceedings of the International Workshop on Classical Swine Fever and Emerging Diseases, vol. 94, Vientiane, Lao, September 19–22, 1999, pp. 100–105. Cargill, C., Davis, P.R., 1997. External parasites. In: Straw, B.E., D’Allaire, S., Mengeling, W.L., Taylor, D.J. (Eds.), Diseases of Swine, 8th ed. Blackwell Science, Osney Mead, Oxford, pp. 669–683. Cutler, R.S., Fahy, V.A., Spicer, E.M., Cronin, G.M., 1997. Preweaning mortality. In: Straw, B.E., D’Allaire, S., Mengeling, W.L., Taylor, D.J. (Eds.), Diseases of Swine, 8th ed. Blackwell Science, Osney Mead, Oxford, pp. 985–1001. Davies, P.R., 1995. Sarcoptic mange and production performance of swine: a review of the literature and studies of association between mite infestation. Vet. Parasitol. 60, 249–264. de Fredrick, D.F., 1977a. Pig production in the Solomon Islands. I. Village pig production. Trop. Anim. Health Prod. 9, 113–123. de Fredrick, D.F., 1977b. Pig production in the Solomon Islands. II. Diseases and parasites. Trop. Anim. Health Prod. 9, 135–139. de Fredrick, D.F., Osborne, H.G., 1977. Pig production in the Solomon Islands. III. The influence of breed. Trop. Anim. Health Prod. 9, 203–210. Dial, G.D., Marsh, W.E., Polson, D.D., Vaillancourt, J.P., 1992. Reproductive failure: differential diagnosis. In: Leman, A.L., Straw, B.E., Mengeling, W.L., D’Allaire, S., Taylor, D.J. (Eds.), Diseases of Swine, 7th ed. Iowa State University Press, Ames, IA, pp. 88–137. Esrony, K., Kambarage, D.M., Mtambo, M.M.A., Muhairwa, A.P., Kusiluka, L.J.M., 1997. Helminthosis in local and cross-bred pigs in the Morogoro region of Tanzania. Prev. Vet. Med. 32, 41–46. Gatenby, R.M., Chemjong, P.B., 1992. Reproduction of pigs in the hills of eastern Nepal. Trop. Anim. Health Prod. 24, 135–142. Hollanders, W., Harbers, A.H.M., Huige, J.C.M., Monster, P., Rambags, P.G.M., Hendrikx, W.M.L., 1995. Control of Sarcoptes scabiei var suis with ivermectin: influence on scratching behaviour of fattening pigs and occurrence of dermatitis at slaughter. Vet. Parasitol. 58, 117–127. Kabare, N., 1991. Genetic and environmental aspect of reproductive performance and preweaning growth of three pig breeds in large-scale herds in Kenya. M.Sc. Thesis. University of Nairobi, Kenya. Kambarage, D.M., Msolla, P., Falmer-Hansen, J., 1990. Epidemiological studies of Sarcoptic mange in Tanzanian pig herds. Trop. Anim. Health Prod. 22, 226–230. King, V.L., Koketsu, Y., Reeves, D., Xue, J., Dial, G.D., 1998. Management factors associated with swine breeding-herd productivity in the United States. Prev. Vet. Med. 35, 255–264. Kunavongkrit, A., Heard, T.W., 2000. Pig reproduction in southeast Asia. Anim. Reprod. Sci. 60–61, 527–533. Lanada, E.B., Lee, J.A.L.M., More, S.J., Taveros, A.A., Cotiw-an, B.S., 1999. The reproductive performance of sows raised by smallholder farmers in the Philippines. Prev. Vet. Med. 41, 187–194.
210
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
Losinger, W.C., Bush, E.J., Smith, M.A., Corso, B.A., 1998. An analysis of mortality in the grower/finisher phase of swine production in the United States. Prev. Vet. Med. 33, 121–145. Manuel, M.F., Santos, A.V., Lucas, S., 1989. Prevalence of gastrointestinal helminths affecting backyard piggery farms. Philos. J. Vet. Anim. Sci. 15, 47–53. Martin, S.W., Meek, A.H., Willeberg, P., 1987. Veterinary Epidemiology: Principles and Methods. Iowa State University Press, Ames, IA. Masembe, F.N., 1985. A study of some factors affecting reproductive performance in female pigs in Kenya. MSc. Thesis, University of Nairobi, Kenya. Minitab, 1999. Minitab® User’s guide, Release 13 for windows. Minitab® Inc., USA. More, S.J., Lee, J.-A.L.M., Lanada, E.B., Taveros, A.A., Cotiw-an, B.S., 1999. Weight-for-age of grower pigs raised by smallholder farmers in the Philippines. Prev. Vet. Med. 41, 151–169. Munyua, S.J.M., Agumbah, G.J.O., Njenga, M.J., Kuria, K.J.N., Kamau, J.A., 1991. Causes of preweaning mortality in small-scale and medium-scale intensive piggeries in central Kenya (including Nairobi). Ind. J. Anim. Sci. 61 (2), 126–128. Noordhuizen, J.P.T.M., Frankena, K., van der Hoofd, C.M., Graat, E.A.M., 1997. Application of Quantitative Methods in Veterinary Epidemiology. Wageningen Press, Wageningen, The Netherlands. Polson, D.D., 1996. Key production factors influencing weaned pig output. In: Proceedings of the 14th International Pig Veterinary Society Congress, Bangkok, Thailand, 531 pp. Radostits, O.M., Leslie, K.E., Fetrow, J., 1994. Planned animal health and production in swine herds. In: Textbook of Herd Health, Food Animal Production and Medicine, 2nd ed. W.B. Saunders, Philadelphia, PA, pp. 435–526. Radostits, O.M., Gay, C.C., Blood, D.C., Hinchcliff, K.W., 1999. Veterinary medicine. In: Textbook of the Diseases of Cattle, Sheep, Pigs, Goats and Horses, 9th ed. W.B. Saunders, London. Straw, B.E., Dewey, C.E., Wilson, M.R., 1997. Differential diagnosis of swine diseases. In: Straw, B.E., D’Allaire, S., Mengeling, W.L., Taylor, D.J. (Eds.), Diseases of Swine, 8th ed. Blackwell Science, Osney Mead, Oxford, pp. 41–86. Tantasuparuk, W., 2000. Sow reproductive performance in Thailand (effects of climate, breed, parity, lactation length, weight loss during lactation and weaning-to-service interval). Ph.D. Thesis. Swedish University of Agricultural Sciences, Uppsala, Sweden. Taveros, A.A., More, S.J., 2001. A field trial of the effect of improved piglet management on smallholder sow productivity in the Philippines. Prev. Vet. Med. 49, 235–247. Vaillancourt, J.-P., Tubbs, R.C., 1992. Preweaning mortality. Vet. Clin. N. Am. Food Anim. Pract. 8 (3), 685–706. Vaillancourt, J.-P., Stein, T.E., Marsh, W.E., Leman, A.D., Dial, G.D., 1990. Validation of producer–recorded causes of preweaning mortality in swine. Prev. Vet. Med. 10, 119–130. Vrbanac, I., Balenovic, T., Yammine, R., Valpotic, I., Krsnik, B., 1995. Preweaning losses of piglets on a state farm in Bosnia and Herzegovina. Prev. Vet. Med. 24, 23–30. Wabacha, J.K., Maribei, J.M., Mulei, C.M., Kyule, M.N., Zessin, K.W., Oluoch-Kosura, W., 2004. Characterisation of smallholder pig production in Kikuyu Division, Central Kenya. Prev. Vet. Med. (in press). Wilkins, J.V., Martinez, L., 1983. Bolivia. An investigation of sow productivity in humid lowland villages. World Anim. Rev. 47, 15–18. Wongnarkpet, S., Morris, R.S., Pfeiffer, D.U., 1994. Pig productivity comparison between tropical and temperate climates. In: Proceedings of the 13th International Pig Veterinary Society, Bangkok, Thailand, 466 pp.
Health and production measures for smallholder pig production in Kikuyu Division, central Kenya J.K. Wabacha a,∗ , J.M. Maribei a , C.M. Mulei a , M.N. Kyule b,c , K.H. Zessin c , W. Oluoch-Kosura d a
d
Department of Clinical Studies, University of Nairobi, P.O. Box 29053, Nairobi, Kenya b Department of Public Health, Pharmacology and Toxicology, University of Nairobi, P.O. Box 29053, Nairobi, Kenya c Postgraduate Studies in International Animal Health, Freie Universität Berlin, Luisenstrasse 56, D-10117 Berlin, Germany Department of Agricultural Economics, University of Nairobi, P.O. Box 29053, Nairobi, Kenya
Received 21 August 2002; received in revised form 21 January 2004; accepted 6 February 2004
Abstract A longitudinal study was carried out in Kikuyu Division (a peri-urban area in central Kenyan highlands) between January 1999 and December 1999 to estimate the baseline parameters on reproductive performance of the sow, as well as health and productivity of grower and preweaning pigs of smallholder herds. Data were collected on 155 breeding pigs, 795 grower pigs and 801 preweaning piglets in 74, 50 and 40 smallholder herds, respectively, using record cards that were updated during monthly visits. The sow-level medians were: weaning-to-service interval 3 months; interfarrowing interval 6.4 months; number of live-born piglets 9.0; and number of piglets weaned per litter 7.5. The piglet crude morbidity incidence risk was 29%. The cause-specific incidence risks for the important health problems encountered in preweaned piglets were diarrhea (4.3%), pruritus (17.1%), and skin necrosis (4.2%). The estimated crude mortality incidence risk to 8 weeks of age was 18.7%. The cause-specific mortality incidence risks to 8 weeks of age for the important causes of mortality were overlying (9.9%), savaging (2.4%), unviable piglets (2.0%) and unknown (1.9%). Overall, 78.8% of the total live-born piglet mortality occurred during the first week postpartum with 69% of these deaths being caused by overlying. The grower-pig crude morbidity incidence risk was 20% and the cause-specific incidence risks of the important health problems encountered were gut edema (1.3%), pruritus (21.1%), and unknown (2.3%). The crude mortality incidence risk was 3.8% and the important causes were gut edema and unknown causes (cause-specific mortality incidence risks of 1.3 and 1.6%, respectively).
∗
Corresponding author. Tel.: +254-2-630-451; fax: +254-2-632-059. E-mail address: [email protected] (J.K. Wabacha). 0167-5877/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.prevetmed.2004.02.006
198
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
The median weight:age ratio and average daily weight gain for the grower pigs were 5.1 kg/month of age and 0.13 kg/day, respectively. For preweaning pigs, the median average daily weight gain was 0.13 g/day. © 2004 Elsevier B.V. All rights reserved. Keywords: Smallholder systems; Pigs; Health; Production; Kenya
1. Introduction In livestock production, suitable parameters for assessing health and productivity are those that measure production and also act as economic indicators (Martin et al., 1987). In breeding herds, the number of pigs weaned per sow per year is (from economic point of view) the most important measure of productivity (Radostits et al., 1994). This production parameter depends on litters/sow/year, farrowing rate, culling rate, weaning-to-service interval, number of live-born piglets and preweaning piglet mortality (Dial et al., 1992; Polson, 1996). For grower pigs, average daily weight gain is considered an important productivity measure (Radostits et al., 1994). Studies conducted in commercial piggeries in tropical areas indicate that sow reproductive performance is lower than in subtropical or temperate areas (Wongnarkpet et al., 1994; Kunavongkrit and Heard, 2000; Tantasuparuk, 2000). This poor performance has been attributed mainly to hot climate (leading to lower number of piglets at birth) (Kunavongkrit and Heard, 2000; Tantasuparuk, 2000). Studies on performance of the sow in smallholder herds in tropical regions are few (de Fredrick and Osborne, 1977; Wilkins and Martinez, 1983; Gatenby and Chemjong, 1992; Lanada et al., 1999). These studies demonstrated low performance of the sows as a result of pig diseases, substandard management, difficult in obtaining quality feeds and inappropriate service–delivery systems. Preweaning piglet mortality is an important cause of reduced sow performance in tropical smallholder herds (Wilkins and Martinez, 1983; Gatenby and Chemjong, 1992; Lanada et al., 1999; Taveros and More, 2001). However, information on the causes of preweaning piglet mortality is scanty. Piglet morbidity which contributes directly or indirectly to mortality (Cutler et al., 1997) also is not well studied in tropical smallholder pig herds and the available information is limited to cross-sectional studies that mainly focussed on parasitic diseases (de Fredrick, 1977b; Manuel et al., 1989; Kambarage et al., 1990; Esrony et al., 1997). The grower-pig performance in the smallholder herds in the tropics is also lower as compared to commercial production systems of the subtropics and tropical areas (de Fredrick and Osborne, 1977; More et al., 1999). The poor performance is attributed to inadequate and poor-quality feeds, poor genetics and diseases (de Fredrick and Osborne, 1977; More et al., 1999). Among the diseases, internal and external parasites are common as a result of climatic and management factors (de Fredrick, 1977b; Manuel et al., 1989; Kambarage et al., 1990; Esrony et al., 1997). In Kenya, few studies have been done on sow reproductive performance in smallholder herds (Masembe, 1985; Munyua et al., 1991). In those studies, passively derived data and convenience-sampling methods were used and high preweaning mortality was identified as
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
199
the main factor influencing sow performance. However, the methods used make us question data quality (Cameron, 2000). Studies on grower and preweaning pig health and productivity have not been undertaken in Kenya. To assess the economic importance of pig diseases and to design appropriate disease-control programs, information on productivity, disease incidence and mortality is necessary. We undertook a prospective longitudinal study on smallholder pig herds in a high-potential farming and peri-urban area in central Kenyan highlands. Our main objectives were to determine the reproductive performance of the sow, as well as the health and productivity of grower and preweaning pigs of smallholder herds in Kikuyu Division, central Kenya.
2. Materials and methods 2.1. Study area This study was undertaken in the high-potential farming and peri-urban area in Kikuyu Division of Kiambu District, central Kenya. The study area and the smallholder pig farms and their management practices were reported previously by Wabacha et al. (in press). 2.2. Selection of study farms The selection of the study farms was carried out as described previously (Wabacha et al., in press). 2.3. Data collection The selected herds were visited once-a-month by the first author and two recruited enumerators between January 1999 and December 1999. Data on sow reproductive performance were collected using sow-litter and dry-sow record cards. For the dry sow, the data collected included service date, farrowing date, date weaned and type and amount of feed provided (recorded in an open format). For the lactating sow, the data collected included farrowing date, number of live-born piglets, number of piglets born dead or mummified, date weaned, number weaned, piglet mortality, age at weaning and type and amount of feed provided. The data were recorded on the cards either from the recall of events that occurred between the visits or from records scribbled on the pen walls by some farmers. During the monthly visits, a detailed physical examination of the piglets and grower pigs in those herds with piglets and grower pigs was made and morbidity data recorded. The farmers provided additional information on morbidity and mortality occurring between the visits. Farmers had been motivated to keep close watch on their herds by being offered free treatment for any sick pig reported. Additionally, whenever possible, the first author also attended sick pigs at the farm at the request of the farmers. Clinical diagnosis of sick or dying pigs mainly was based on anti-mortem clinical signs observed during the monthly visits or from clinical history presented by the farmers. The small herd sizes allowed the farmers to follow the health events in their herds in detail. To avoid considering a case more than once, innovative ways of marking the upcoming cases (that included scraping off
200
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
of the hairs) were used—but in most cases, farmers were able to identify individual pigs. Data also were collected on type and amount of feed provided and preventive-medicine practices. During the monthly visits, unweaned piglets and grower pigs (up to a maximum of 40 kg) were weighed to the nearest 0.01 kg using a suspended weighing scale (Salter® , Salbrave, Kenya Limited). The data were recorded on prescribed record cards that were updated during the monthly visits. Rectal fecal samples and ear-wax scrapings from randomly selected preweaning and grower pigs from each of the selected herds were sampled once for fecal egg counts and mite identification, respectively. Fecal samples were taken from 83 preweaning and 204 grower pigs and analyzed using the modified McMaster technique (Anon., 1986). Ear-wax scrapings were taken from 117 preweaning and 154 grower pigs by scraping the inner aspects of the ear until traces of blood could be seen. The materials were examined under the light microscope for the presence of mites after digestion with 10% potassium hydroxide. 2.4. Measures of sow reproductive performance The measures of sow performance were preweaning piglet mortality (PWMT), weaningto-service interval (WTSI), interfarrowing interval (IFI), number of live-born piglets (NLBP), number of piglets weaned per farrowing (NWF), litters per sow per year (L/S/Y) and the number of piglets weaned per sow per year (PW/S/Y) (Dial et al., 1992) (Table 1). 2.5. Measures of health in preweaning and grower pigs Morbidity risk and mortality risk (Martin et al., 1987) were used as measures of health in preweaning and grower pigs. For preweaning pigs, the biologic period of risk was the period from farrowing to weaning while for grower pigs it was the period from weaning to the time the pigs left the herd through sale or mortality. The amount of disease (morbidity) was determined by recording any disease/condition with a recognizable clinical sign(s) while mortality was determined by recording all cases of death in pigs. 2.6. Measures of productivity in preweaning and grower pigs The measure of productivity used for the pigs was average daily weight gain (ADWG) in kg/day; additionally, the ratio of the measured weight and the reported age (weight:age ratio) was used for the grower pigs (Radostits et al., 1994; More et al., 1999). 2.7. Data management Data were entered in the data collection cards at the farm during the visits. Data files were created in Dbase (DBASE IV Plus, Ashton Tate, Torrance, California, USA) and Excel® 1997 (Microsoft Corporation, USA). The files were screened for any errors that might have occurred during the data entry and the errors were corrected by rechecking against the original data collection cards.
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
201
Table 1 Health and productivity measures for pigs in smallholder pig herds in Kikuyu Division, Kiambu District, central Kenya (January 1999–December 1999) Animal and measurement
Sows Weaning-to-service interval (month)
N
Minimum
Percentile
Maximum
25th
50th
75th
33a 22b
0.2 0.2
1.0 1.3
3.0 3.1
4.1 5.1
10.5 10.5
Interfarrowing interval (month)
29a 17b
5.0 5.7
5.9 6.0
6.4 6.9
7.5 7.8
13.0 13.0
Number of live-born piglets
95a 40b
0.0 4.5
7.0 7.6
9.0 9.0
12.0 10.4
20.0 17.0
Number of piglets weaned per farrowing
95a
0.0
5.0
7.5
9.0
14.0
40b
1.5
6.0
7.0
9.0
14.0
95a 40b
1.0 1.0
1.0 1.0
1.0 1.0
1.0 1.2
2.0 2.0
76a 40b
0.0 1.5
6.0 7.0
8.0 8.5
11.0 10.0
23.0 17.0
86c 40d 86c 40d 53c 28d
0 0 0 0 0.02 0.05
0 0 0 2.1 0.07 0.07
0 16 10 12 0.13 0.14
70 78 29 26 0.19 0.2
100 100 100 100 0.4 0.4
116c 50d 116c 50d 39c 17d 396e 26d
0 0 0 0 0.01 0.08 1.0 1.5
0 10 0 0 0.11 0.12 3.8 4.1
21 29 0 0 0.13 0.15 5.1 5.1
33 33 0 2.1 0.21 0.19 6.2 6.0
Litters/sow/year Piglets weaned/sow/year Preweaned piglets Crude morbidity incidence Crude morbidity incidence Crude cumulative mortality Crude cumulative mortality Average daily weight gain (kg/day) Grower pigs Crude morbidity incidence Crude morbidity incidence Crude cumulative mortality Crude cumulative mortality Average daily weight gain (kg/day) Weight:age (kg/month)
87 60 60 48 0.36 0.33 11.9 10.6
a
Sows. Herds. c Litters. d Herds. e Growers. b
2.8. Data analysis Data analyses were performed using Statistix® for Windows (Analytical Software, Tallahassee, FL, USA), Excel® 1997 (Microsoft Corporation, USA) and Minitab Statistical Software, version 13 for windows (Minitab Inc., USA).
202
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
2.8.1. Measures of sow reproductive performance A total of 155 breeding pigs were kept in 74 of the 76 herds that were enrolled for the longitudinal study. A number of parameters were only available for some sows and herds (Table 1) because sows in some herds were withdrawn from the study through death or sale while others did not farrow during the study period. The values of the parameters (section 2.4) for each sow in a herd were used as the sow-level measures of performance while the median values of the parameters for all the sows in the herd were used as herd-level measures of sow performance. 2.8.2. Measures of health in preweaning and grower pigs The cause-specific, age-specific and crude morbidity and mortality cumulative incidences for each litter in a herd were used as the litter-level measures of health. The median values of crude morbidity and mortality cumulative incidences for all the litters in the herd were used as herd-level measures of health. For crude morbidity, any pig that had at least one disease/condition was taken as a case. For the cause-specific morbidity, only new cases arising during the monitoring period were considered. The morbidity and mortality cumulative incidences for the piglets were calculated for the preweaning period of 8 weeks. Data on piglet health were collected in 40 out of the 74 herds with breeding pigs (Tables 1–3) because of the same reasons as given in Section 2.8.1. Grower pigs were kept in 50 out of the 76 herds that were enrolled for the study and the health parameters were determined in all the herds (Tables 1, 4 and 5). For the individual litters of grower pigs, the monitoring period lasted between 2 and 5 months. To take care of the different time-at-risk for the different litters of grower pigs, incidence rates were used to determine the grower-pig morbidities and mortality per grower-pig month-at-risk. Estimated cumulative incidence (CI(t) ) then was calculated using the formula: CI(t) = 1 − e−IR×t , where e is the base of the natural logarithm, and t indicates the time-unit of concern as described in Noordhuizen et al. (1997). 2.8.3. Measures of productivity in preweaning and grower pigs Piglet productivity was determined in only 28 herds (out of 76 with breeding pigs) because in addition to the reasons given earlier, piglets in some herds were weaned before two consecutive weights could be taken. For grower pigs, determination of average daily weight gain (kg/day) and weight:age ratio (kg/month of age) was only possible in 17 and 26 herds, respectively, due to censoring when the grower pigs were sold or when the study ended; others became too heavy to weigh with the suspended weighing scale during the subsequent visits. The average daily weight gain for the pigs in a litter was used as the measure of productivity. The ADWG was calculated as the difference in total litter weight between two consecutive monthly visits divided by the average number of pigs in the litters during the first and second visit and the number of days between the two visits. The ratio of the measured weight and the reported age of each grower-pig (weight:age) in kg/month of age was used as a measure of productivity to increase the number of herds where productivity of grower pigs could be determined, because the number of herds where ADWG could be determined was small. For each of the productivity measures, ADWG and weight:age ratio, the median of the parameters were used as herd-level measures of productivity.
Disease/condition
Callus on carpus Deformitya Diarrhea Herniaa Loss of claw Pneumonia Pruritis Skin necrosis Unthriftiness Unknown Age-specific Incidence risk (%) a
Age (weeks) 0 to <1
1 to <2
2 to <3
3 to <4
4 to <5
5 to <6
6 to <7
7 to <8
0 1 6 1 0 0 0 16 4 2
0 0 9 0 0 0 3 3 0 2
8 0 9 0 0 0 0 6 0 0
9 0 0 0 0 0 28 1 0 2
0 0 0 0 0 0 33 0 1 1
0 0 4 0 0 4 27 1 0 0
0 0 8 0 0 2 29 8 0 1
0 0 0 0 1 2 14 0 0 0
3
2
3
4
5
6
7
3
The cumulative incidence was not calculated because the conditions were present at birth.
Litter-level incidence risk (%) 2.4 – 4.3 – 0.14 1.3 17.1 4.2 1.3 0.77
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
Table 2 The distribution of morbidity cases by clinical manifestations, age in weeks, cause-specific and age-specific cumulative incidence for 801 preweaning piglets in 40 out of 76 smallholder pig herds in Kikuyu Division, Kiambu District, central Kenya (January 1999–December 1999)
203
204
Disease/condition
Overlying Pneumonia Predation Savaginga Scours Starvation Unviable piglets Unknown Age-specific Incidence risk (%) a
Age (weeks)
Litter-level crude cumulative mortality (%)
0 to <1
1 to <2
2 to <3
3 to <4
4 to <5
5 to <6
6 to <7
7 to <8
87 0 0 20 0 4 13 2
0 0 8 2 0 3 0 2
0 0 0 0 0 0 0 2
0 0 0 2 0 0 0 1
0 0 0 0 0 0 0 1
0 2 0 0 1 0 0 1
0 0 0 0 0 0 0 5
0 0 0 0 0 0 0 4
9.9 0.2 0.8 2.4 0.1 1.2 2.0 1.9
14.3
2.5
0.2
0.6
0.1
0.6
0.7
0.6
–
The 2 piglets in the fourth week were just found missing and the cause could not be ascertained.
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
Table 3 The distribution of mortality cases by clinical manifestations, age in weeks, cause-specific and age-specific cumulative mortality for 801 preweaning piglets in 40 out of 76 smallholder pig herds in Kikuyu Division, Kiambu District, central Kenya (January 1999–December 1999)
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
205
Table 4 The distribution of morbidity cases by clinical manifestations, age in months, cause-specific incidences and crude cumulative incidences for 795 grower pigs in 50 out of 76 smallholder pig herds in Kikuyu Division, Kiambu District, central Kenya (January 1999–December 1999) Disease/condition
Abscess Diarrhea Ear necrosis Gut edemac Pneumonia Pruritis Unthriftiness Unknown
Age (months) 2 to <3
3 to <4
4 to <5
5 to <6
6 to <7
7 to <8
0 4 0 14 12 133 1 19
0 1 0 0 0 95 2 2
0 0 1 0 0 31 0 3
1 10 0 0 0 17 0 0
0 0 0 0 0 24 0 0
0 0 0 0 3 18 0 1
IRa
CIb
0.1 0.22 0.001 1.3 0.4 23.6 0.2 2.1
0.1 0.22 0.001 1.3 0.4 21.1 0.2 2.3
a
Litter-level cause-specific morbidity incidence (%) per pig-month-at-risk. Estimated litter-level cause-specific morbidity incidence risk (%). c Neurological disorder, characterized by hoarse squeal, ataxia, swelling of the eyelids drooping ears, paresis and paralysis was tentatively diagnosed as gut edema. b
Table 5 The distribution of cases by clinical manifestations, age in months, cause-specific and cumulative mortalities for 795 grower pigs in 50 out of 76 smallholder pig herds in Kikuyu Division, Kiambu District, central Kenya (January 1999–December 1999) Disease/condition
Gut edemac Diarrhea Pneumonia PSSd Unknown
Age (months) 2 to <3
3 to <4
4 to <5
5 to <6
6 to <7
7 to <8
13 0 6 0 3
0 4 6 0 10
0 1 0 0 2
0 0 0 0 4
0 0 0 1 0
0 0 0 0 1
IRa
CIb
1.3 0.19 0.3 0.1 1.7
1.3 0.19 0.3 0.1 1.6
a
Litter-level cause-specific mortality (%) per pig-month-at-risk. Estimated litter-level cause-specific mortality (%). c Neurological disorder, characterized by hoarse squeal, ataxia, swelling of the eyelids, drooping ears, paresis and paralysis was tentatively diagnosed as gut edema. d Porcine stress syndrome, tentative diagnosis based on the history of death after exhaustion. b
3. Results 3.1. Measure of sow reproductive performance The measures of sow reproductive performance are shown in Table 1. At least 25% of the sows and herds achieved a WTSI of not more than 1 and 1.3 months, respectively. Twenty-five percent of the sows and herds achieved an IFI of not more than 5.9 and 6 months, respectively.
206
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
3.2. Diseases occurrence The diseases encountered in preweaning piglets were diarrhea, callus on carpus, loss of claws, pneumonia, pruritis, skin necrosis and unthriftiness. The prevalence for mange on mite detection was 10.3% while that for helminthosis was 27.7%. For grower pigs, the diseases encountered were abscesses, diarrhea, ear necrosis, gut edema, pneumonia, pruritis, unthriftiness, and other diseases whose tentative diagnosis could not be made. The prevalence for mange on mite detection was 19.5% while that for helminthosis was 50.5%. 3.3. Measures of health in preweaning and grower pigs The litter- and herd-level median crude morbidity cumulative incidences for the preweaning pigs were 0 and 16%, respectively (Table 1). The litter-level cause-specific cumulative incidence from birth up to 4 weeks was highest for skin necrosis (3%) and diarrhea (2.5%). From the fourth week postpartum to weaning, the cause-specific morbidity risk was highest for pruritis (17.1%) (Table 2). The litter- and herd-level median crude cumulative mortality of preweaning piglets was 10 and 12%, respectively (Table 1). At least 25% of the herds had a preweaning piglet mortality of not more than 2.1%, although mortality in some herds was as high as 100%. Preweaning piglet mortality was highest during the first week postpartum. Among the piglets born alive, overlying was the most frequent cause of death (9.9%) followed by savaging (3%) (Table 3). Overall, 78.8% of the total live-born piglet mortality occurred during the first week (with 69% of these deaths being caused by overlying). For the grower pigs, the litter- and herd-level median crude morbidity cumulative incidences during the period of the study were 21 and 28.5%, respectively (Table 1). The diseases with the highest incidence risk were pruritis (21.1%), gut edema (1.3%) and diseases of unknown causes (2.3%) (Table 4). At least 50% of litters and herds had crude cumulative mortalities of less than zero (Table 1). Among the causes of grower-pig mortality, gut edema (1.3%) and unknown causes (1.6) had the highest incidence (Table 5). 3.4. Measures of productivity in preweaning and grower pigs The measures of productivity in preweaning and grower pigs are shown in Table 1. The litter- and the herd-level median ADWG for the preweaning pigs were 0.13 and 0.14 kg/day, respectively. For grower pigs the median animal-level and herd-level weight:age was 5.1 kg/ month of age. The litter- and herd-level median ADWG was 0.13 and 0.15 kg/day, respectively.
4. Discussion The reproductive performance of the sow was low as compared to commercial piggeries in tropical countries including Kenya (Kabare, 1991; Wongnarkpet et al., 1994; Kunavongkrit and Heard, 2000; Tantasuparuk, 2000). However, the number of live-born piglets observed
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
207
was similar to the one observed in commercial piggeries in Kenya and other tropical regions (Kabare, 1991; Lanada et al., 1999; Kunavongkrit and Heard, 2000). The climatic stress due to high temperatures (especially from weaning until early gestation) can cause low litter size in tropical regions (Tantasuparuk, 2000). The NWF was lower than in commercial piggeries in tropical areas (Kabare, 1991) and this could be attributed to the high preweaning mortality observed. Similar observations have been made in other smallholder herds (de Fredrick, 1977a). The WTSI (median 3.0 months) observed in this study was much longer than observed in commercial tropical piggeries (Kunavongkrit and Heard, 2000; Tantasuparuk, 2000) and this could have contributed to the prolonged IFI observed in this study. The prolonged IFI (median 6.4 months), was in agreement with other observations in smallholder herds (Wilkins and Martinez, 1983; Gatenby and Chemjong, 1992; Lanada et al., 1999; de Fredrick and Osborne, 1977). The IFI is influenced by the lactation length, number of non-productive days and the gestation length (Dial et al., 1992; King et al., 1998). The non-productive days could be prolonged by poor nutrition offered to the sows and non-availability of the boar. The prolonged IFI could contribute to low number of litters weaned per sow per year. For the preweaning pigs, the cause-specific morbidity due to facial-skin necrosis, pruritis and scours were high and this was consistent with previous findings (Straw et al., 1997; Cameron, 1997; Kambarage et al., 1990). Facial necrosis commonly is observed during the first week postpartum and is as a result of infection by Staphylococcus hyicus and Fusobacterium necrophorum of wounds inflicted by piglets on each other during feeding (Cameron, 1997). Several factors cause pruritis in pigs (Davies, 1995; Hollanders et al., 1995; Cargill and Davis, 1997). Sampling of randomly selected pigs revealed high prevalence with sarcoptic mange with no other ectoparasites being detected; it thus appears that the pruritis in the studied herds was caused by sarcoptic mange. The preweaning piglet mortality (median 10%) observed in this study differs from that reported by Munyua et al. (1991) and Masembe (1985). They reported mean preweaning piglet mortalities of live-born piglets of 15.8 and 17.1%, respectively. The difference between the studies done in Kenya could be due to the sampling strategies used. We used a random sample of smallholder herds while the previous studies were based on convenient samples that also included both medium and smallholder herds. The high preweaning piglet mortality we observed was consistent with findings in other tropical smallholder herds (Lanada et al., 1999; Wilkins and Martinez, 1983; Gatenby and Chemjong, 1992; Taveros and More, 2001). Pig-management practices in smallholder herds are low (Lanada et al., 1999) and this could explain the high preweaning mortality observed in the tropical smallholder herds. Over three quarters of the overall mortality (78.8%) occurred during the first week of life and more than half of these deaths (69%) were caused by “overlying”. Although we classified this as a category on its own, “overlying” is interrelated with other factors—for example, a piglet which is getting little milk is likely to become weak and be “overlaid” by the sow (Vaillancourt and Tubbs, 1992; Cutler et al., 1997). The pattern of the causes of preweaning piglet mortality observed in this study differs from that observed in a previous study in Kenya (Munyua et al., 1991) in which gastrointestinal syndrome, starvation and pneumonias were the most important causes of preweaning piglet mortality in small and medium scale pig
208
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
herds. Those authors used convenient samples and relied more on secondary data than we did. The classification of the causes of preweaning mortality in this study was mainly based on the pig owners recorded causes of mortality between the monthly visits. These causes might need validation by performing post-mortem examinations (Vaillancourt et al., 1990; Vrbanac et al., 1995; Cutler et al., 1997). Due to the nature of the study design, this could not be accomplished. The observed crude morbidity risk (median 21%) of grower pigs seemed high to us. High morbidity risk would compromise the performance of grower pigs by contributing directly or indirectly to mortality or by reducing feed-conversion efficiency (Radostits et al., 1994). The high cause-specific morbidity due to pruritis (sarcoptic mange) was in agreement with findings in other tropical smallholder pig herds (Kambarage et al., 1990). Pruritis in grower pigs, due to sarcoptic mange, is a common problem especially in pigs kept in unhygienic conditions and where mange control measures are inadequate (Cargill and Davis, 1997). In this study, many cases of a neurological disorder (suspected to be gut edema from clinical presentation) were observed. Gut edema is a common problem in grower pigs, especially immediately after weaning (Radostits et al., 1999). The “unknown causes” contributed important morbidity of the grower pigs and they deserve further research to identify them. The other diseases recorded (for example, ear necrosis, unthriftiness and abscesses) are common disease/conditions in pigs (Straw et al., 1997). In the current study, the cause-specific morbidity for pneumonia’s and diarrhea were low and this was in agreement with findings in other tropical smallholder pig-production systems (de Fredrick, 1977b). The grower-pig crude mortality risk observed was high. The “unknown diseases” contributed the highest mortality risk and future recommendations to specifically delineate them are indicated. The high mortality risk of grower pigs would reduce grower-pig productivity in the smallholder herds. The median weight:age ratio observed (5.1 kg/month of age) was in agreement with observations in tropical smallholder herds in the Philippines (More et al., 1999). Poor growth of pigs in smallholder herds has been attributed to poor feeding (both in quality and quantity) (de Fredrick and Osborne, 1977; More et al., 1999). The data collected on preweaning and grower pigs morbidity and mortality mainly depended on farmer-reported morbidity and mortality. To ensure that the information was correct, the data were validated by gathering a detailed clinical history of pig morbidity and mortality at the farm during the monthly visits. However, it is still possible that there may have been some under reporting (especially with morbidity). Information based on farmer-reported morbidity and mortality is destined to have some errors (Losinger et al., 1998). Nevertheless, we believe that the data collected were representative of the diseases currently experienced on the smallholder pig herds in the study area. In conclusion, the reproductive performance of the sow (weaning-to-service interval, interfarrowing interval and number of piglets weaned per farrowing) was poor. The health and productivity of pigs in the studied smallholder herds were also suboptimal with the major health problems being mange and helminthosis. The variation in performance of the various herds indicates that an opportunity exists to raise productivity in the studied herds.
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
209
Acknowledgements We thank the smallholder pig farmers in Kikuyu Division, Kiambu District, for their enthusiasm throughout the study. The effort of the livestock extension officers (Ministry of Agriculture and Rural Development, Kikuyu Division) who facilitated the link with the farmers during farm recruitment and the subsequent visits is highly appreciated. This study was supported by Agricultural Research Fund (ARF) contract no. ARF/LSKP/1001004/1 and Akademischer Austausch Dienst (DAAD).
References Anon., 1986. Manual of Veterinary Parasitological Laboratory Techniques. Reference Book No. 418, 3rd ed. Ministry of Agriculture, Fisheries and Food. Helminthology, Her Majestys Stationery Office, London, pp. 1–68. Cameron, R.D.A., 1997. Diseases of the skin. In: Straw, B.E., D’Allaire, S., Mengeling, W.L., Taylor, D.J. (Eds.), Diseases of Swine, 8th ed. Blackwell Science, Osney Mead, Oxford, pp. 941–958. Cameron, A., 2000. Strategies for epidemiological surveys in developing countries. In: Proceedings of the International Workshop on Classical Swine Fever and Emerging Diseases, vol. 94, Vientiane, Lao, September 19–22, 1999, pp. 100–105. Cargill, C., Davis, P.R., 1997. External parasites. In: Straw, B.E., D’Allaire, S., Mengeling, W.L., Taylor, D.J. (Eds.), Diseases of Swine, 8th ed. Blackwell Science, Osney Mead, Oxford, pp. 669–683. Cutler, R.S., Fahy, V.A., Spicer, E.M., Cronin, G.M., 1997. Preweaning mortality. In: Straw, B.E., D’Allaire, S., Mengeling, W.L., Taylor, D.J. (Eds.), Diseases of Swine, 8th ed. Blackwell Science, Osney Mead, Oxford, pp. 985–1001. Davies, P.R., 1995. Sarcoptic mange and production performance of swine: a review of the literature and studies of association between mite infestation. Vet. Parasitol. 60, 249–264. de Fredrick, D.F., 1977a. Pig production in the Solomon Islands. I. Village pig production. Trop. Anim. Health Prod. 9, 113–123. de Fredrick, D.F., 1977b. Pig production in the Solomon Islands. II. Diseases and parasites. Trop. Anim. Health Prod. 9, 135–139. de Fredrick, D.F., Osborne, H.G., 1977. Pig production in the Solomon Islands. III. The influence of breed. Trop. Anim. Health Prod. 9, 203–210. Dial, G.D., Marsh, W.E., Polson, D.D., Vaillancourt, J.P., 1992. Reproductive failure: differential diagnosis. In: Leman, A.L., Straw, B.E., Mengeling, W.L., D’Allaire, S., Taylor, D.J. (Eds.), Diseases of Swine, 7th ed. Iowa State University Press, Ames, IA, pp. 88–137. Esrony, K., Kambarage, D.M., Mtambo, M.M.A., Muhairwa, A.P., Kusiluka, L.J.M., 1997. Helminthosis in local and cross-bred pigs in the Morogoro region of Tanzania. Prev. Vet. Med. 32, 41–46. Gatenby, R.M., Chemjong, P.B., 1992. Reproduction of pigs in the hills of eastern Nepal. Trop. Anim. Health Prod. 24, 135–142. Hollanders, W., Harbers, A.H.M., Huige, J.C.M., Monster, P., Rambags, P.G.M., Hendrikx, W.M.L., 1995. Control of Sarcoptes scabiei var suis with ivermectin: influence on scratching behaviour of fattening pigs and occurrence of dermatitis at slaughter. Vet. Parasitol. 58, 117–127. Kabare, N., 1991. Genetic and environmental aspect of reproductive performance and preweaning growth of three pig breeds in large-scale herds in Kenya. M.Sc. Thesis. University of Nairobi, Kenya. Kambarage, D.M., Msolla, P., Falmer-Hansen, J., 1990. Epidemiological studies of Sarcoptic mange in Tanzanian pig herds. Trop. Anim. Health Prod. 22, 226–230. King, V.L., Koketsu, Y., Reeves, D., Xue, J., Dial, G.D., 1998. Management factors associated with swine breeding-herd productivity in the United States. Prev. Vet. Med. 35, 255–264. Kunavongkrit, A., Heard, T.W., 2000. Pig reproduction in southeast Asia. Anim. Reprod. Sci. 60–61, 527–533. Lanada, E.B., Lee, J.A.L.M., More, S.J., Taveros, A.A., Cotiw-an, B.S., 1999. The reproductive performance of sows raised by smallholder farmers in the Philippines. Prev. Vet. Med. 41, 187–194.
210
J.K. Wabacha et al. / Preventive Veterinary Medicine 63 (2004) 197–210
Losinger, W.C., Bush, E.J., Smith, M.A., Corso, B.A., 1998. An analysis of mortality in the grower/finisher phase of swine production in the United States. Prev. Vet. Med. 33, 121–145. Manuel, M.F., Santos, A.V., Lucas, S., 1989. Prevalence of gastrointestinal helminths affecting backyard piggery farms. Philos. J. Vet. Anim. Sci. 15, 47–53. Martin, S.W., Meek, A.H., Willeberg, P., 1987. Veterinary Epidemiology: Principles and Methods. Iowa State University Press, Ames, IA. Masembe, F.N., 1985. A study of some factors affecting reproductive performance in female pigs in Kenya. MSc. Thesis, University of Nairobi, Kenya. Minitab, 1999. Minitab® User’s guide, Release 13 for windows. Minitab® Inc., USA. More, S.J., Lee, J.-A.L.M., Lanada, E.B., Taveros, A.A., Cotiw-an, B.S., 1999. Weight-for-age of grower pigs raised by smallholder farmers in the Philippines. Prev. Vet. Med. 41, 151–169. Munyua, S.J.M., Agumbah, G.J.O., Njenga, M.J., Kuria, K.J.N., Kamau, J.A., 1991. Causes of preweaning mortality in small-scale and medium-scale intensive piggeries in central Kenya (including Nairobi). Ind. J. Anim. Sci. 61 (2), 126–128. Noordhuizen, J.P.T.M., Frankena, K., van der Hoofd, C.M., Graat, E.A.M., 1997. Application of Quantitative Methods in Veterinary Epidemiology. Wageningen Press, Wageningen, The Netherlands. Polson, D.D., 1996. Key production factors influencing weaned pig output. In: Proceedings of the 14th International Pig Veterinary Society Congress, Bangkok, Thailand, 531 pp. Radostits, O.M., Leslie, K.E., Fetrow, J., 1994. Planned animal health and production in swine herds. In: Textbook of Herd Health, Food Animal Production and Medicine, 2nd ed. W.B. Saunders, Philadelphia, PA, pp. 435–526. Radostits, O.M., Gay, C.C., Blood, D.C., Hinchcliff, K.W., 1999. Veterinary medicine. In: Textbook of the Diseases of Cattle, Sheep, Pigs, Goats and Horses, 9th ed. W.B. Saunders, London. Straw, B.E., Dewey, C.E., Wilson, M.R., 1997. Differential diagnosis of swine diseases. In: Straw, B.E., D’Allaire, S., Mengeling, W.L., Taylor, D.J. (Eds.), Diseases of Swine, 8th ed. Blackwell Science, Osney Mead, Oxford, pp. 41–86. Tantasuparuk, W., 2000. Sow reproductive performance in Thailand (effects of climate, breed, parity, lactation length, weight loss during lactation and weaning-to-service interval). Ph.D. Thesis. Swedish University of Agricultural Sciences, Uppsala, Sweden. Taveros, A.A., More, S.J., 2001. A field trial of the effect of improved piglet management on smallholder sow productivity in the Philippines. Prev. Vet. Med. 49, 235–247. Vaillancourt, J.-P., Tubbs, R.C., 1992. Preweaning mortality. Vet. Clin. N. Am. Food Anim. Pract. 8 (3), 685–706. Vaillancourt, J.-P., Stein, T.E., Marsh, W.E., Leman, A.D., Dial, G.D., 1990. Validation of producer–recorded causes of preweaning mortality in swine. Prev. Vet. Med. 10, 119–130. Vrbanac, I., Balenovic, T., Yammine, R., Valpotic, I., Krsnik, B., 1995. Preweaning losses of piglets on a state farm in Bosnia and Herzegovina. Prev. Vet. Med. 24, 23–30. Wabacha, J.K., Maribei, J.M., Mulei, C.M., Kyule, M.N., Zessin, K.W., Oluoch-Kosura, W., 2004. Characterisation of smallholder pig production in Kikuyu Division, Central Kenya. Prev. Vet. Med. (in press). Wilkins, J.V., Martinez, L., 1983. Bolivia. An investigation of sow productivity in humid lowland villages. World Anim. Rev. 47, 15–18. Wongnarkpet, S., Morris, R.S., Pfeiffer, D.U., 1994. Pig productivity comparison between tropical and temperate climates. In: Proceedings of the 13th International Pig Veterinary Society, Bangkok, Thailand, 466 pp.