Livestock Research for Rural Development 25 (11) 2013 Guide for preparation of papers LRRD Newsletter

Citation of this paper

Socio-economic assessment of indigenous goat production system in rural areas of Bugesera District in Rwanda

M Manzi, J Mutabazi, C D Hirwa and D R Kugonza*

Rwanda Agriculture Board, P.O. Box 5016 Kiga
* College of Agricultural and Environmental Sciences, Makerere University,
P.O. Box 7062, Kampala, Uganda


This study was commissioned to assess the production systems and opportunities for genetic improvement of goats in Rwanda. A total of 241 goat farms were visited and one respondent at each farm was interviewed using a structured questionnaire. The generated data were analysed using Generalised Linear Model procedures.

Most of the farms kept the indigenous Small East African goats, either in pure flocks or in combination with exotics. Tethering of goats while grazing/browsing and zero grazing in confinement sheds were the common management systems. On disease/parasite prevalence, 58% of the farms reported gastro-intestinal parasites as a top concern, while 17% had Contagious Echtyma. Most farms preferred controlled to natural mating. The risk of inbreeding was high because most male kids were castrated for sale in nearby markets. Farmers who kept Records were rare (4%) and 50% of these kept breeding records. Nearly all respondent had no training on goat production. Therefore, the study identified training in improved goat husbandly and investment in research in improved goat genotypes as pertinent interventions for improved goat productivity in Rwanda

Key words: breeding, genetic improvement, selection


Indigenous goats have important socio-economic roles in the livelihood strategies of the poor farmers, especially those in rural and hard-to-reach areas. Those roles include their use as savings, insurance, security, accumulation and diversification of assets, social and cultural functions. They are also valued for their productive performance, adaptation and disease resistance. Small ruminants in general represent about 30% of the red meat and 21% of the total milk produced in sub Saharan Africa worth about 1.3 and 1.2 billion dollars respectively (Winrock International 1992). Other attributes that make them favourable to pastoral production include the small body size, therefore requiring low nutrient levels for survival. They are also suitable for subsistence systems because they produce meat in small readily usable quantities (Indetie et al 1998).

Despite the merits mentioned above, the indigenous goats in Rwanda as is typical for the Small East African goat types are known for their low productivity and efforts to upgrade them are in synch with the national development goals. However, replacement and indiscriminate crossbreeding of indigenous goats may cause genetic erosion (loss of genetic diversity and reduction of adaptive value) and opportunities for efficient utilisation of the existing goat’s genetic resources for improvement of livelihood may be missed (Das et al 1992). In various instances, there has been an increasing tendency to introduce exotic germplasm and concentration on a narrow range of supposedly more profitable breeds (Hammond 1994). About 30% of indigenous animal genetic resources are therefore at risk of extinction due to indiscriminate crossbreeding (Scherf 2000) which is associated with erosion of breeds that are naturally adapted to the local environments.

Goats are a very valuable genetic resource that is suited for low-input agricultural production systems. They require low inputs and are easy to manage, making them suitable for the resource poor rural households. These characteristics enable them to continue providing milk and meat even when cattle have succumbed to drought (Rege 1994). The East African indigenous goat has been characterised as a small, very hardy animal which are bi-coloured or multi-coloured. Almost all the indigenous goats in East Africa belong to the Small East African goat group (Devendra and Burns 1996), though a few other major breeds such as the Borana (Galla), Mubende and Kenya Dual Purpose goats exist.

The objective of the present study was to determine the production system of indigenous goats in selected sites of Rwanda with the aim of identifying gaps to inform uptake pathways, facilitate policy formulation and strengthening capacity for improvement and investment in small ruminant development programs.

Materials and methods

Study location

Bugesera district is located in the low altitude zone of Rwanda, at 30o 30–30o 25 E and 2o05 –2o 30 S and at an altitude of 1,400 m a.s.l (Munyemana 2001). The climate of the area is essentially equatorial but is categorized as semi-arid according to total annual precipitation. The local communities in the area are basically agro-pastoralists but mainly keep more goats than cattle. Average daily temperature is 25oC in wet season and 30oC in dry season and a relative humidity of 74%. The rainfall received is bimodal, with short rains (season A) falling between September and December and long rains (season B) extending from March through May. The dry season extends between June and August (Fig. 1).

Figure 1. Monthly rainfall (mm) distribution in Bugesera district
Study Design

A survey was conducted using a structured questionnaire to determine the production system of indigenous goats in Bugesera District of Eastern Rwanda. A total of 241 goat keeping farmers in Rwanda were interviewed one-on-one. The survey included baseline data on the economics of goat production where goats are kept and purposive sampling was done of the major goat producing regions to obtain baseline data on production traits and economics of production of indigenous goats. Data was obtained for the household management system of goat in relation to socio-economic parameters for husbandry practices, breeding and selection, disease control, production parameters, flock dynamics, marketing and extension materials.

Sampling procedure

The stratified sampling technique was used to implement the survey. At the highest level the Eastern Province was chosen because of the existence of a wide phenotypic diversity of goat genetics (Fig. 2) and the adequacy of indigenous goat populations for meaningful sampling. This was based on historical background and the availability of information on the sites identified. The sample sizes were dictated by the resources available for the sampling process.

Figure 2. A flock of indigenous goats showing a wide array of coat colours

Within the province, the survey was conducted in one (Bugesera) out of seven districts because the district has the highest number of goats and proximity to the Goat Breeding Centre at Karama Station. For the same reason 4 out of 14 Sectors within the District were selected for the study. All villages within the target sectors were sampled. Within villages 30% of households keeping at least 5 goats (237 out of 709) were randomly selected for questionnaire administration (Table 1).

Table 1. Categories of goat keeping households is villages selected for goat characterization survey

No. of goats/household





































































Grand total






The sampling layers were random at farmer level, herd level and purposive at animal level (selecting specific animals).The data on production system collected were entered in MS Excel computer software and the frequencies obtained were analysed using NPARIWAY and GLM procedures of SAS (2004).

Results and Discussion

Characteristics of the surveyed households

In total, 241 respondents were interviewed and of all, 33.2% were household heads, 44% were spouses and 15% were children (daughters and sons). Others were hired managers (2.5%) and elderly dependants /parents of the household heads. The female headed households were 9.1%.These findings show that the majority of farmers in the study area are actually resident and not absentee or cell-phone farmers. Majority of the respondents were female (55.2%), plausibly because traditionally, small ruminant production is predominantly a women’s activity. Jaitner et al (2001) reported that in the Gambia, 71% of goat owners were found to be women. Women are more likely to own small ruminants probably because of their determination to increase their economic autonomy and their bargaining power within the household. Thus owning small ruminants contributes to their empowerment. The operations undertaken by female members of the household include feeding, watering, maintenance of cleanliness and manure collection. Construction of small ruminant structures, breeding management, animal health issues and marketing arrangements were men’s responsibilities.

The majority of the heads of household were either primary school leavers (52.8%) and or illiterate (38.2%), while a modest 7.5% attended secondary school. In a way, these results indicate that most goat farming activities were being done by illiterate farmers, which has serious bearings on production methods and management ability of the farmers. High illiteracy rates tend to hinder adoption of suitable technology and do not make it easy to communicate to such producers messages of technical nature. Adult literacy training may need to be arranged for the studied community to enhance their literacy skills.

Land ownership, goat grazing patterns and general management system

Most of the land used for goat farming was individually owned (87.6%), having been either traditionally inherited or purchased. Other lands were communally owned (5.4%), or borrowed (5.8%).Tethering was the dominant grazing system in most households (45%). Goats were being tethered close to the homesteads in the morning from 0900–1200 hrs, then brought back home and either kept in sheds/pens or tethered on pegs and thereafter supplied with crop residues. Goats would normally be taken back to the grazing area where they would be tethered till evening.  Free range grazing either on individually owned, communally owned or public land was rather common.

The high prevalence of free range grazing was due to availability of public grazing land. However, free range management practices cannot survive the on-going reduction of farm sizes as a result of a high population density and subsequent increased pressure on the land. Such practices are also not in line with government’s initiative to intensify and modernize production systems, to maximize production per unit area as enshrined in the country’s vision for 2020. Free range systems lead to overgrazing, and the risk of disease outbreaks is very high. Few farmers practiced zero grazing system, and probably these were the ones who had attained secondary school education, and hence had some knowledge of modernized goat production. Substantial work is therefore required for the transformation of the production systems from traditional extensive grazing to the desired zero grazing system.

GGoat production in Rwanda is focused on attainment of self-sufficiency in meat and manure. In most areas of the country, especially in places with high human population pressure, there has been overuse of the available arable land that has led to depletion of soil nutrients, hence the need to replace/build up soil fertility through the use of animal manure, compost and/or chemical fertilizers. This could probably be the reason why semi-intensive was the most popular production system (66%), so as to maximise efficiency in the use of manure.

Breeds of livestock reared and their roles /strong>

The indigenous Small East African breed was the dominant type kept by 97% of all the households surveyed (n = 234). This gives an impression that limited efforts have been put in place to improve indigenous goats either by crossbreeding or substitution by exotic breeds; however, efforts are on-going, only that they need to be harmonised and strengthened. On the other hand, 0.4% kept exotic goats, 0.8% kept both local and exotics, while 1.7% were rearing a combination of local and crossbreds. Only 1% of the surveyed households kept sheep probably because mutton traditionally is not eaten by many people in Rwanda.

Farmers kept goats mainly because of the goats’ ability to thrive in conditions of inadequate feeds, high prolificacy and high market value (Figure 3). Other reasons were income (41.5%)and cultural benefits (9.9%). Recent studies in Kenya (Kosgey et al 2006) ranked regular cash income as the most important purpose of small ruminants for both smallholders and pastoral/extensive farmers. Knowing the reasons for keeping small ruminants is a prerequisite for deriving operational breeding goals (Jaitner et al 2001). Indeed, ignorance of this aspect has been a major constraint in the lack of success in genetic improvement programs attempted in the tropics (Sölkner et al 1998; Rewe et al 2002). Considering the importance that farmers attach to goat production, our findings suggest that genetic improvement programmes if well planned have a good chance of success.

Figure 3. Purpose for keeping goats

Most of households surveyed in addition to keeping goats, they also kept cattle (44%), chickens (43%), pigs (4%), rabbits (3%) and equines (0.5%). However, with this information it should be recalled that in selection of households one of the criteria was to select farmers keeping goats, thus this probably has led to some of farmers keeping other livestock species to be left out. Jaitner et al (2001) and Seleka (2001) found out that small ruminant production ranked close to cattle. The majority of farmers in the study had no special housing for their animals (54.7%), adult stock were normally tied to pegs fixed onto the floor, while, kids was left free. Thirty eight percent (38.6%) of the farms, housed animals in the confined sheds constructed for this purpose. The houses are mostly the ordinary floor type.

Disease prevalence, treatment and prevention

Gastro intestinal parasites (58%) were the major animal health problem facing goats in the study areas. Other prevalent diseases were contagious Echtyma (33%) and pneumonia (8%). In Kenya’s smallholder and extensive pastoral systems, the most common diseases were tick-borne diseases (Kosgey et al 2001). Interestingly, in this study it was also found that 35% of farmers use Nilzan and 31% use Albendazole while 19% use both traditional medicine and Nilzan and only 7% use traditional medicine in treatment of gastrointestinal parasites. Generally, majority of farmers practiced irregular deworming (65.2%) against endo-parasites while 21.4% practiced no control measures. Elsewhere, the control of ecto-parasites is mainly by cleaning (64.2%) while 17.9% practiced no control measure. Vaccination of goats was not reported;  this service has been not successful due to meagre budget allocation from the government. The veterinary drugs and vaccines are expensive and in most cases unavailable. With most farmers owning both cattle and goats, cattle receive the highest priority in resource allocation. Maximum productivity in a given system of production emerges when disease control is optimal (Gatenby 1986). Hence healthcare is an important problem to consider before implementation of any genetic improvement program.

Mortalities were mainly occurring in April and May, according to 47.6% of 147 respondents who conceded to losing animals. Some 12.9% reported deaths mainly occurring in April and other months after May, while 39.5% reported that deaths mainly occurred in the other ten months with none dominant. Most deaths occurred in these months mainly due to rain only (22.6%), gastrointestinal parasites only (28.5%), or these two factors in combination with others (48.9%).

Breeding methods and sources of breeding buck and breeding management

The majority of farmer (78.4%) practiced controlled breeding as compared to natural method (Table 2). Few farmers keep bucks for breeding in Bugesera. Normally nearly all born male kids are castrated when less than three month of age for improved meat quality. As result, the majority of farmers borrow (62.2%) or hire (10%) bucks for breeding (Table 2). The inbreeding problem is likely to be encountered in future because the majority of farmers in the study area did not seem to realize this problem of using one buck common for all without employing different strategies to go around it. The present study is in agreement with the observation by Kosgey (2006) in Kenya where keeping of goats breeding purposes was lowly ranked. However, contrary to the present findings uncontrolled mating was the most common practice.

Table 2: Reproduction and breeding


Proportion of households (%)

Breeding method







No response



Do you practice








Source of breeding buck




Own herd









No response



Decision maker on

Household head












Veterinary officer



The major reasons revealed by farmers for using natural breeding were that its asimple method (easy to manage), while others said that it’s the only breeding method available (no alternative), also some said that animal mates while in grazing area and did not know when they do. Controlled mating was performed mainly with the objective of avoiding unwanted breeding, avoiding inbreeding (16%), improve success of conception (12%), improvement of flock performance (11%) and increasing fertility (6%). It appears that some farmers are aware of the dangers of inbreeding, however, from Table 5, majority of farmers depend on borrowed bucks for breeding without putting in place strategies for controlling it.  

Majority of the farmers (43%) reported that they replace their buck depending on availability of alternative bucks, while 25% and 16% respectively reported to replace their buck after 1 year and 2 years (Table 3). The implications of this are vast, but critically, it means that majority of farmers did not have strategies in place for buck replacement. This has got negative implication because the most critical part of any meat goat operation is the selection of a herd sire (breeding buck). A high quality buck can produce high quality offspring even mated with an average doe (Kugonza et al 2001a; Kugonza et al 2001b).

Table 3: Years that a breeding buck is kept before buck replacement



Percentage of households

1 year



1-2 years



1-3 years









2-3 years



2-4 years



3 years



Depending on availability of the buck



The reasons advanced by most farmers (61%) for replacement of their bucks were fear of inbreeding followed by improvement of the performance of flock (16%). Others (11%) reported to replace the buck because of old age. Forty three percent of the farmers reported to have encountered breeding problems with their goats, compared to 57% who experienced none. The most critical goat breeding problem revealed in the study area was lack of a breeding buck (59%) followed by failure of goats to conceive (repeat breeding) (29%) and abortions (12%).The lower percentage of farmers with goat breeding problems is probably in line with those without their own breeding bucks who have to depend on borrowing or hiring and as a result, it is difficult for these farmers to plan their breeding strategies. Due to the low number of bucks, some farmers probably take their does for buck service when the oestrus is ending.

Observation of heat is the most important aspect attributable to the farmers in influencing fertility. There are various methods of heat detection practiced by rural farmers in the study area as shown (Table 8). Almost all the farmers know and combine several heat behavioural attributes to detect oestrus in goats. However, the majority of the farmers perceive bleating and mucus discharge to be the most important signs of heat (Table 4), and few (6%) used loss of appetite and mounting of others goats to be sign of heat detection. No single farmer mentioned use of standing to be mounted as the behavioural sign indicative of oestrus. Because standing when mounted is the most reliable sign of oestrus, the majority of farmers in the study area are therefore not able to correctly pin point and predict occurrence of oestrus.

Table 4: Signs of detecting heat in indigenous goats



Percent (%)

Bleating and vagina discharge



Bleating and mounting others



Loss of appetite and mounting others



Bleating alone



Bleating and mounting others



Vagina discharge, bleating and mounting others



Mounting others and vagina discharge



Behaviour change (active) and becomes difficult to manage



Bleating and swollen vulva



Mounting other and frequent urination



Most births were reported to be occurring between June and August, by 62.2% of 193 respondents, while according to 10.9%, there was no dominant one month, while 26.9% of the respondents were distributed over the other nine months from September to May. Goats are seasonal breeders, however there are breeds of goats in which, like in sheep,  breeding activity never stops completely when kept in a suitable environment (Gordon 1997) and photoperiods have been implicated in seasonality in goat breeding ( Chemineau and Delgadillo 1994).

Record keeping

Very few farmers (4%) in the study area kept records out of which only 2% of the farmers responded to be keeping breeding records on kidding date, date of service after kidding, and number of kids born. Records and record keeping is one of the most critical areas in improved goat productivity as evaluation is based on performance records. Without record keeping, one cannot monitor events in the flock related to the reproductive parameters. However it has been found out that it is unlikely that performance recording is logistically feasible in large numbers of smallholder flocks (Baker and Gray 2003).  

Feeding systems

The survey results indicated that 50% of the farmers provided preferential feeding to their goats, while 47% never provided any kind of special feeding. The main preferential feeds included bean soup, Napier grass, sorghum porridge and other crop residues provided in various combinations depending on the availability. Farmers prioritised these preferential feeds to the lactating goats. Roughage was being used by many farmers who rarely purchased concentrates, thus confirming that small ruminants tend to be kept in low-input systems. Although the feed quality and quantity of many tropical grasses is often inadequate (Carles 1983; Gatenby 1986; Charray et al 1992), it would appear from this survey that farmers are doing their best to attend to the nutrition of their stock from their limited means.

Selling goats, goats products and setting the price

Majority of farmers (91%) reported to be selling their goats throughout the year. This contradicts the commonly known phenomenon that farmers especially under semi intensive system sale their goat at the beginning of rainy season mainly because of demand of labour. This supports the results other reports indicating that ad hoc sales of animals to meet emergencies prevail (Seleka 2001). Several reasons for selling goats were revealed and all were related to using sales from goat to solve household problems that ranged from paying for school fees, seeds, fertilizer, community health insurance, house construction, buying food, paying dowry and paying for on farm labour and only 1.2% to sell for the purposes of destocking (culling).  

No single farmer reported even to be selling goat milk or milk products, and this as reported earlier has got a lot to do with the culture of Rwandans for whom goat milk is not traditionally a common food. However, other goat products were reported sold, other than selling mainly live animals as breeding stock, meat, skin and manure were respectively the most sold goat products. The majority of farmers (58%) based on the size of the goats to set the price followed by quality of meat. Castrated goats normally have quality meat and thus are more expensive as compared to un-castrated ones.

Plan to change goat flock sizes

From the survey results it was clear that majority of farmers (84%) in the study area were willing to increase the current stock kept. The reasons revealed were mainly to increase income for the households and that the increased revenues would enable them solve their day to day requirements including paying for School fees, to construct bigger goat houses, paying for community health insurance, paying for seeds and labour on farm. Further, some farmers reported plans to increase number of goats so as sell goat and buy cows thus shifting from goat keeping to cattle keeping. Others said that by increasing the numbers they will be able to get enough fertilisers and also because those with enough land they want to increase the number so as to maximise the profit goat sales. However, some few (7%) farmers said that they do not intend to increase the number of goats because mainly because they do not have enough land to sustain their animals and also that they had limited personnel to look after high number and others believed higher numbers are associated with higher goat mortalities.  

Farmer training and skills

Fifty four percent (54%) of the farmers reported to be getting information about breeding, feeding, production, marketing and any other techniques from local radios. Others reported that they get information from local authorities (sector veterinarians) and neighbours while some revealed to look for any information anywhere (use skills acquired traditionally). Ninety five percent (95%) of the farmers reported to have not received any training and only 3.3% attended any such training. In such situation, local knowledge pertaining to management and breeding practices are applied across board and as such, survival and reproductive traits are upheld in this traditional system (Bernstein and Jacobs 1983; Kosgey 2004).  

Indigenous goat breeds are the dominant type kept by ninety seven percent (97%) of the total households surveyed. The goats in tropics are often kept in large numbers and are valued for their ability to survive drought periods. But production levels are low due to:-poor nutrition, diseases, parasites and harsh environment (Rege 1994; Rey and Das 1997).

The main constraints to goat farming are low prices and lack of market for the animals, limited feed resources during drought years and high mortality of the kids and lambs. From the researcher perspective, the main constraints are lack of initiatives by the central and district government to assist the small ruminant farmers in development of small ruminant systems through research, extension and commercialisation. Funding of small ruminant development programs have been minimum resulting in small efforts by the NGOs or from the research institutions.



The Authors gratefully acknowledge the support of the Association for Strengthening Agricultural Research in East and Central Africa (ASARECA) in funding this study as part of the regional LFP06 project.


Bernstein R H and Jacob A H 1983 Socio-economic aspects of range research in Kenya. A Plan for Reaching the Ultimate Target Group. Winrock, International.

Chemineau P and Delgadillo J A 1994 East African Herders Development and Change 32: 401-433.

Das S M and Sendalo D S 1992 Comparative performance of improved meat goats in Malya, Tanzania. Small Ruminant Research Network in Africa. In: In: Rey B, Lebbie S H B, Reynolds L (Editors). Proceedings of the First Biennial Conference of the African Small Ruminant research Network, 10-14 December 1990. ILCA, Nairobi, Kenya.

Devendra C and Burns K 1996 Goat production in the tropics. Longman Ltd, London.

Gatenby R M 1986 Sheep Production in the Tropics and Sub-Tropics. LongmanInc., New York, 351 pp.

Hammond K 1994 Conservation of domestic animal diversity: Global overview. Proc. 5th World Congr. Genet. Appl. Livest. Prod., Guelph, Vol. 21, 423- 430.

Indetie D, Karimi S, Wandera F, Lebbie S and Mwai O 1998 Phenotypic characteristics of east African goats in Kajiado districts of Kenya. Sixth Biennial KARI Scientific Conference.

Jaitner J, Sowe J, Secka-Njie E and Dempfle L 2001 Ownership pattern and management practices of small ruminants in the Gambia—implications for a breeding programme. Small Ruminant Research 40, 101–108.

Kosgey I S 2004 Breeding objectives and breeding strategies for small ruminants in the tropics. Ph.D. Thesis, Wageningen University, The Netherlands, 272 pp.

Kosgey I S, Baker RL, Udo H M J and Van Arendonk J A M 2006 Successes and failures of small ruminant breeding programmes in the tropics: a review. Small Ruminant Research 61: 13–28.

Kugonza D R , Bareeba F B and Petersen P H 2001a Response to phenotypic screening of Mubende meat goats. MUARIK Bulletin 4: 18–24.

Kugonza D R , Bareeba F B and Kirembe G 2001b Characterisation of meat goat management systems in three districts of central Uganda. MUARIK Bulletin4: 26–29.

Maijala K 1983 Genetic diversity of European Livestock Breeds. In: Simon D L, Boucehuquer D (eds). EAAP Publications, No 66.

Munyemana J M V 2001 Evaluation of some agroforestry species for resistance to the termites on the grounds of Bugesera, case of the Kayovu hill, Memoir, UNR, Faculty of Agronomy, Butare, New, pp 26-27.

Rege J E O 1994 Indigenous African Small Ruminants: A case for characterization and improvement. In: Lebbie S H, Rey B and Irungu S(eds)Proceeding of the second Biennial Conference of the African Small Ruminants Research Network, AICC, Arusha, Tanzania pp 205-211.

Rege J E O, Yapi-Gnanore C V and Tawah C L 1996 The indigenous domestic ruminant genetic resources of Africa. Paper presented at the 2nd All Africa Conf. on Anim. Agric., April 1-4, 1996, Pretoria South Africa.

Rewe T O, Ogore P B and Kahi A K 2002 Integrated goat projects in Kenya: impact on genetic improvement. In: Proceedings of the Seventh World Congress on Genetics Applied to Livestock Production, vol. 33, Montpellier, France, 19-23 August, 2002, pp. 385-387.

Rey B and Das S M 1997 A system analysis of inter-annual changes in the pattern of sheep flock productivity in Tanzania livestock research centres. Agricultural Systems 53(2-3): 175-190.

SAS (Statistical Analysis Systems) 2004 The GLM Procedure. SAS Users Guide: 9.1.3 edition, SAS Institute Inc. Cary, NC, USA.

Scherf B D 2000 World Watch List for Domestic Animal Diversity. 3rd Edn., Food and Agriculture Organization of the United Nations, Rome, Italy, pp: 726.

Seleka T B 2001 Determinants of short-run supply of small ruminants in Botswana. Small Rumin. Res. 40, 203-214.

Sölkner J, Nakimbugwe H and Valle Zarate A 1998 Analysis of determinants for success and failure of village breeding programmes. Proceedings of the 6th World Congress on Genetics Applied to Livestock Production Armidale. 273-280.

Winrock International 1992 Strategies for increasing meat production: Small ruminants research in Kajiado and Baringo districts. KARI/ILRI collaborative research Project Report. 223 pp.

Received 6 August 2013; Accepted 22 October 2013; Published 1 November 2013

Go to top