| Livestock Research for Rural Development 27 (4) 2015 | Guide for preparation of papers | LRRD Newsletter | Citation of this paper |
Data were collected on the health performance of indigenous chickens in two districts of the Lake Victoria Crescent Agro-ecological Zone (LVCAZ) of Uganda with the aim of determining the infectious causes of mortalities. A total of 240 chicken keeping households were engaged in the longitudinal study in which survival and mortalities were monitored in a longitudinal study for one year. During the study period 6,226 mature birds, which belonged to different sex categories,were found dead as a result of disease occurrence. Organ samples were taken from every dead bird for laboratory analysis. Additional data were collected on health history and post-mortem findings for birds which died.
The incidence rate of mortality was 0.230 per mature bird-months at risk. Of the total deaths 90% were diagnosed coming from single or mixed type of infections. Whereas in the baseline survey, responses based on farmer observed clinical signs ranked the infectious causes of chicken death as Salmonellosis (37%), Newcastle Disease (ND) (28%), mixed infections (16%) and Coccidiosis (14%) related infections in order of commonality and importance in the study areas, post-mortem and laboratory findings showed a slight deviation. Newcastle disease had the highest proportional mortality rate (46%) followed by mixed infections (30.3%). The proportional mortality caused by Salmonellosis, Coccidiosis fowl cholera, fowl pox and undiagnosed cases were 7.02 %, 4.58%, 0.321%, 1.27%, and 10.36%, respectively.Vaccinated birds were 3.7 times more likely to survive (P < 0.01, χ² = 306) during the scourge. Chicken mortality was also significantly linked to seasonal changes (P <0.05, χ² = 157) with critical times (March to May and October to late November) which are the two wet seasons, showing the highest mortality rates. Time between last vaccination against ND and disease outbreak incidence significantly (P < 0.01) affected mortality of chicken. A logistic regression model of last vaccination (against ND) time prior to incidence with the proportion of the death was significant (P < 0.01, χ² = 121.06). The model indicated that the shorter the last vaccination times prior to occurrence of incidences the higher the survival of flocks.
It was concluded that the main infectious causes of death in scavenging flocks were treatable (Bacterial and Coccidial) and immunizable Newcastle Disease, strategic control of which could help minimise mortalities considerably.
Keywords: backyard poultry, critical times, disease, seasonality, survival, Uganda
Indigenous chickens form the bulk of the rural poultry production in Uganda. Over 90% of Ugandan chickens are indigenous stock reared under the backyard system (Olaboro 1990), producing an average of 50 eggs per hen per year. The petite nature of chicken makes it a readily available and sure source of meat at meal times and functions. In many resource scarce households it could be described as a poor man’s insurance or even bank as it can easily be sold off or given out in compensation to settle debts. In many instances, chickens have been used as a vehicle to move poor households up and off the poverty ladder.
Indigenous chickens are mainly raised on a free range arrangement with bare a few supplements given. This leaves the birds to scavenge in order to survive andleads to mixing of flocks from different households. As a result, there is continuous contact between these birds of different age groups and types, leading to easy spread of diseases (Johnston, 1990). Indeed, diseases are the major cause of mortality in chicken in Uganda (Kugonza et al 2008), though other challenges also persevere such as predation and wasting.In the present study we aimed at determining the infectious causes of mortality in order to developappropriate interventions.
The study was conducted in two districts namely; Mubende and Mpigi both located in the Lake Victoria Crescent Agro-ecological zone. This area is sub humid in nature and receives bimodal rainfall, with temperatures ranging from 22° to 30°C annually. The inhabitants of the districts are mainly Baganda, of the Bantu ethnic group, and the production system of the area can generally be characterised as the Coffee-Banana system.
A stratified design was used to select the study households, with stratification done at district, sub-county, and parish and village level. Two districts chosen on the basis of indigenous chicken abundance, two Sub-counties, two parishes and two villages from each district were studied (Table 1). Only those households that were willing to participate, close to one another in one village and having an average of at least five chickens were engaged in the study. At least 25 and 10 households were enlisted in the study per parish and village respectively.
| Table 1. The number and proportions of households by administrative area and gender | ||
| Variable | Level | Number (%) |
| District | Mubende | 120 (50%) |
| Mpigi | 120 (50%) | |
| Sub-county | Kalwana | 60 (25%) |
| Kassanda | 60(25%) | |
| Muduuma | 60(25%) | |
| Mpigi Town Council | 60(25%) | |
| Gender | Male | 72 (30%) |
| Female | 168 (70%) | |
A study was conducted from February 2011 to January 2012 to determine the infectious causes of mortality in scavenging chickens. Active surveillance was conducted two districts (Mpigi and Mubende) onto a total of 240 households. The data collected included: demographic information (Household biodata, socio-economic), chicken production (flock sizes, housing, feeding), deaths (number of dead birds per month, signs before death, and Internal organ samples for post-mortem) and associated causes, frequency of vaccination and time between vaccinations against ND prior to the incidence of disease. The size of the study population was estimated based on the total number of birds recorded at the beginning and the end of 12 consecutive months. Two epidemiological measures were applied to estimate disease occurrence: incidence rate of mortality and proportional mortality rate. The incidence rate of mortality was estimated using the number of birds found dead during this period as the numerator and the total number of bird months at risk as the denominator. The proportional mortality rate was calculated as the number of deaths caused by a specific disease divided by the total number of deaths recorded (Thrusfield, 1986). Demographic and production data were recorded using interview guided questionnaires in a one farm visit while special clinical forms were used to record mortalities, clinical history and post mortem data as and when cases were reported by farmers. All participating farmers were tracked so as not to lose contact with them and rewarded with a token for their commitment each time the team of investigators visited them.
The team of investigators made monthly field visits to selected farmers to corroborate and verify the collected data by extension workers. Investigators comprising of avet epidemiologist, a pathologist, two animal scientists, a biostatisticianand two technicians supported by extension workers in each sub-county including a Veterinarian and two Para-veterinarians carried out the study. Prior to the study, two contact farmers were indentified in each village and given mobile telephones for communication with extension workers and researchers and ensuring compliance by the selected farmers. Furthermore, extension workers were trained in data collection in a two days workshop which was concluded by pre-testing of the data collection tools.
The extension workers were responsible for the collection of organ samples from dead birds including digestive tract, liver, spleen, heart, lung, trachea, and bursa of Fabricius. The motivation for reporting deaths and collecting samples by farmers and extension workers respectively was a financial incentive at the rate of Shillings 200 per dead bird for farmers and 50,000 shillings per day for extension workers to collect all of the dead birds from households.Each organ sample collected from a dead bird was placed separately in a polythene bag and all samples taken from a single dead bird were kept together in a polythene bag. Each bag was tagged to identify samples from a particular bird. Every sample had a submission form completed by the extension workers with information including sex, age of birds, medication history, vaccination history, post-mortem examination findings and tentative diagnoses based on clinical history and post-mortem examination findings. All samples collected in a particular area were kept frozen in a refrigerator at the ‘‘Extension worker’s office’’ until collected fortnightly and taken to the Department of Pathology at the Faculty of Veterinary Medicine, Makerere University for analysis.
Data were entered in MS Excel computer software worksheets and imported to SAS version 9.1 for analysis. Proc means and Proc Logistic were used to analyse the data. The information is mainly presented as means, percentages and charts. Significance was considered at P < 0.05. In our study, while calculating mortalities the number of deaths per household was taken as the numerator whereas the denominator was the total household flock size prior to the scourge. New acquisitionsinto the flock during and/or after the scourge were not considered for the mortality computation of that period but rather for the next month. Causal associations between factors such as sex of bird, whether vaccinated or not were assessed using odds rations or using a Chi-square test including determining independence and goodness of fit of the logistic model.
The logistic regression model used to link mortality to last vaccination against NCD prior to incidence (time) took the form:
Model:
Where:
Time = Time between vaccinations (of last
vaccination against ND prior to incidence in months)
p = the probability of mortality
a = constant
b = coefficient of time
Due to the monetary incentives given to both farmers and extension workers and good communication strategies, it was presupposed that all of the birds that died from disease were reported and collected. Overall, 27024 birds were recorded in the period of 12 month and a total of 6226 birds died table 2 gives the summary of monthly recordings.There were five major diseases recorded affecting the smallholder scavenging chickens reared in the study areas; Newcastle’s disease, ND; Coccidiosis, Cocc; Salmonellosis, Sal; Fowlpox (FP) and fowl cholera (FC). Notably, cases of mixed infections were not uncommon.
Study participants were largely female 168 (70 %), and almost all participants 216 (90%) attained some form of formal education. 178 (74%) of the respondents had specific chicken houses, in 20%of the households, chickens slept in kitchens while in 6%, chickens were sharing accommodation with humans especially in Mubende district.
Generally every household had a sort of ailment in their chicken flocks over the past 12 months. Of the total deaths 90% were diagnosed coming from single or mixed type of infections. There were no significant (P>0.05) variations attributed to differences in sex of chicken across all measurements. Whereas in the baseline survey, responses based on farmer observed clinical signs ranked the infectious causes of chicken death as Salmonellosis (37%), Newcastle Disease (ND) (28%), mixed infections (16%) and Coccidiosis (14%) related infections in order of frequency and importance in the study areas, post-mortem and laboratory findings showed a slight deviation. Newcastle disease (Table 2) had the highest proportional mortality rate (46%) followed by mixed infections (30.3%). Mortality rates due to ND associated causes ranging from 40% to 60% have been reported by Chowdhury et al (1982). The incidence rate of mortality for the study period was 0.230 per mature bird-months at risk (i.e. 230 deaths per 1000 bird-month at risk). The average proportional incidence ratesof mortality with corresponding ranges for the critical times or rainy seasons during above times were 0.52 (0.19 to 0.76) and 0.48 (0.23 to 0.85) respectively. Higher proportional rates of mortality were detected during March to May and resurgence again during October through November. Higher proportional rates of mortality in March and November attributed to most diseases except for Fowl pox and Fowl cholera indicated two epidemics of the diseases notably in ND in the study time. Generally very low incidence mortality rates were observed for bacterial infections. The proportional mortality caused by Salmonellosis, Coccidiosis fowl cholera, fowl pox and undiagnosed cases were 7.02 %, 4.58%, 0.32%, 1.27%, and 10.4 %, respectively. Both ND and Salmonella related infections were also reported earlier by Byarugaba et al (2002), as major causes of mortality in chicken in Uganda. Among the killer diseasesreported in this study, only ND could not be treated and just required vaccination.
Seven pathogen combinations were detected, of which only two resulted from infection by three pathogens. In the mixed infection category, ND presented the highest number of cases (1233) i.e. was involved with at least two different combinations of other diseases, of which the greater proportions (P < 0.05, χ²= 29.1) were found with salmonellosis (42%) and 32% with Coccidiosis.
Chicken mortality was also significantly linked to seasonal changes (P <0.05, χ² = 157.3) with critical times (March to May and October to late November) which are the two wet seasons, showing the highest mortality rates (Table 2). Generally all diseases showed seasonality except for fowl cholera and fowl pox which were fairly consistent.
| Table 2. Chicken population studied and disease observed over one year | |||||||||
| Period of observation | Number of birds recorded on visit day | Total number of deaths | Distribution of diseases | ||||||
| ND | Cocc | Salm | FP | FC | Mixed Infections | Un-diagnosed | |||
| Feb-11 | 1765 | 132 | 58 | 11 | 24 | 3 | 4 | 9 | 23 |
| Mar-11 | 1531 | 1146 | 679 | 26 | 37 | 0 | 4 | 297 | 103 |
| Apr-11 | 504 | 310 | 48 | 26 | 45 | 0 | 1 | 114 | 76 |
| May-11 | 721 | 140 | 24 | 18 | 24 | 0 | 2 | 39 | 33 |
| Jun-11 | 1213 | 65 | 11 | 12 | 11 | 1 | 2 | 16 | 12 |
| Jul-11 | 3442 | 55 | 4 | 5 | 2 | 0 | 5 | 25 | 12 |
| Aug-11 | 4113 | 37 | 2 | 15 | 0 | 1 | 0 | 13 | 6 |
| Sep-11 | 4672 | 771 | 341 | 11 | 71 | 1 | 0 | 228 | 119 |
| Oct-11 | 3806 | 944 | 356 | 39 | 148 | 8 | 12 | 291 | 90 |
| Nov-11 | 2832 | 2369 | 1312 | 61 | 33 | 0 | 46 | 778 | 139 |
| Dec-11 | 591 | 186 | 32 | 59 | 37 | 5 | 1 | 36 | 16 |
| Jan-12 | 1834 | 73 | 6 | 2 | 5 | 1 | 2 | 41 | 16 |
| Total (12 Months) | 27,024 | 6,228 | 2873 | 285 | 437 | 20 | 79 | 1887 | 645 |
| ND, Newcastle Disease; Cocc, Coccidiosis; Salm, Salmonellosis; FP, fowlpox; FC, fowl Cholera. | |||||||||
The average number of chickens per household fluctuated throughout the year, with highest peak being in August-September while the lowest was reported between April and May (Table 2). It is however important to note that there was another surge in the population of chickens in November. The deaths coincided with the flaring numbers and seasonal changes. The first rains fall in March which also happened to be the period with the highest mortality and in the second rains in October where mortalities are also high. It is therefore the high mortalities which are partly responsible for the small flock sizes of the indigenous chickens among chicken keeping households. There was also a clear linkage between mortality and seasonal changes as most deaths occurred during change from dry to rainy season. Similar findings are reported by Ban-Bo et al (2012), Bushra (2012) and Byarugaba et al (2012). During the rainy season many disease transmitting vectors, and hence the diseases become active, and this could explain the rampant deaths.
Vaccinated birds were 3.7 times more likely to survive (P < 0.01, χ² = 306) during the scourge (Table 3). Time between last vaccination against ND and disease outbreak incidence significantly (P < 0.01) affected mortality of chicken. A logistic regression model of last vaccination (against ND) time prior to incidence with the proportion of the death was significant (P < 0.01, χ² = 121.1). The model indicated that the shorter the last vaccination times prior to occurrence of incidences the higher the survival of flocks (Table 4 and 5).
The last vaccination time prior to the incidence of disease was the only quantitative factor which significantly (P < 0.01) affected mortality of chickens. Chicken mortality was also linked to seasonal changes; critical times (P <0.05, χ² =132) showing the highest rates. Seasonal effects of disease occurrence were noted and most farmers across all districts cited the rainy season to have registered the highest number of deaths.
| Table 3. Occurrence of Newcastle Disease in vaccinated and unvaccinated mature birds | ||||||
| Immunization status |
Newcastle Disease occurrence % | Total birds observed |
Odds ratio 95% CI |
χ² value | P value | |
| Positive | Negative | |||||
| Unvaccinated | 809 | 330 | 1139 | 3.73 | 306 | 0.00 |
| Vaccinated | 217 | 98 | 315 | |||
| Total | 217 | 98 | 1454 | |||
From the data collected about the proportions of chickens lost per household per month, a logistic model was fitted for prediction of mortality based on time from last vaccination against Newcastle’s Disease. The model parameters were all significant (Table 4).
| Table 4. Parameter estimates for the model for prediction of mortality | ||||
| Parameter | Estimate | Wald χ² | P value | |
| Constant | -2.29 | 111.5 | 0.01 | |
| Time | 0.89 | 121.1 | 0.01 | |
NCD which was reported to occur in the dry season, all the other diseases were reported to occur during the wet season. In the control of ND, it was noted that the shorter the last vaccination times prior to occurrence of incidences the higher the survival of flocks (Table 5).
| Table 5. Predicted mortality rates associated with last time of vaccination prior to occurrence of NCD incidence | |||
| Time between last vaccination to occurrence of NCD incidence (Months) | Predicted probability of mortality | Lower Critical Level | Upper Critical Level |
| 1 (n=107) | 0.11 | 0.10 | 0.12 |
| 2 (n=71) | 0.89 | 0.73 | 0.95 |
| 3 (n=32) | 0.94 | 0.83 | 0.96 |
| 4 (n=19) | 0.95 | 0.89 | 0.98 |
| 5 (n=11) | 0.99 | 0.98 | 0.99 |
| Numbers in brackets indicate number of households | |||
The predictions from the model showed that as the time between vaccinations against Newcastle’s Disease increased, the rate of mortality exponentially increased (Table 5). A thirty day period (1 month) between vaccinations had the least mortality while 150 days and beyond gave the highest, approximately 100%. A similar vaccination time has been recommended by Ssewannyana (2002) and Okitoi (2006). This is because the immunity developed after vaccination is short-lived and must be taken advantage of, if survival is to be ensured.
Ban-Bo B A, Kebkiba B and Mopate L Y 2012 Epidemiology of Newcastle disease and its economic impact in chad. European Journal of Experimental Biology, 2 (6): 2286-2292.
Bushra B 2012 The Status of Indigenous Village Chicken Production and Marketing System in Ethiopia. Poultry Industry Technical Articles. Available at http://en.engormix.com/MA-poultry-industry/meat-industry/articles/the-status-indigenous-village-t2392/471-p0.htm
Byarugaba D K, Olsen J E and Katunguka-Rwakishaya E 2002 Production, Management and Marketing Dynamics of the Rural Scavenging Poultry in Uganda. In: The Bangladesh Model and other experiences in Family Poultry Development, 2002. Available at http://infpd.net/econference/view/id/26p
Chowdhury F R, Sarker A J, Amin M M and Hossain M A 1982 Studies on Newcastle disease in Bangladesh. A research report, Bangladesh Agricultural University, Mymensingh.
Johnston J 1990 Health and Productivity of Village Poultry in Southeast Asia, ACIAR Working Paper No. 31, Canberra: ACIAR, 61 + appendices.
Kugonza D R, Kyarisiima C C and Iisa A 2008 Indigenous chicken flocks of Eastern Uganda: I. Productivity, management and strategies for better performance. Volume 20, Article #137. Retrieved January 14, 2015, from http://www.lrrd.org/lrrd20/9/kugo20137.htm
Okitoi LO 2000 Improvement of indigenous poultry production in western Kenya. Proceedings of the Endof-Project conference, KARI Headquarters, Nairobi, Kenya, 29 November – 1 December 2000, pp 227 –236.
Olaboro G 1990 Smallholder Rural Poultry Production in Uganda.Country Report. 35 pp.
SAS Institute Inc. 2003 SAS/STAT Software Release 9.1.SAS Institute Inc., Cary, NC.
Ssewannyana E 2003 Characteristics of rural chicken production in Apac and Kumi districts of Uganda. Uganda Journal of Agricultural Sciences 8 (10): 159-164.
Received 22 January 2015; Accepted 12 March 2015; Published 1 April 2015