Economic analysis of cattle fattening systems based on forage tree legume diets in eastern Indonesia

Research and government agencies in eastern Indonesia have identified 2 systems with potential to increase productivity and incomes of small-holder cattle producers: improved cattle feeding practices through forage tree legumes (FTL); and the development of more efficient and specialized cattle-fattening systems. Extensive research has been conducted on production and technical aspects of FTL-fattening systems, but there is a gap in research on economic incentives for households to adopt the systems. This paper provides an economic analysis of a leucaena-fattening system in a village in West Timor. It draws on trial data from associated technical research projects and detailed semi-structured interviews with farmers and other stakeholders to populate a bio-economic model built for the research. Under all measures of profitability, leucaena-fattening systems in representative households are profitable in the wet season. Importantly, ’returns to person days’ are higher than off-farm incomes. The activity generates cash income, increasingly required to meet cash expenses in modern rural Indonesian society. However, returns vary considerably between households, are considerably lower in the dry season and, as would be expected, are sensitive to relative prices of feeder and finished cattle.


Introduction
The province of Nusa Tenggara Timur (NTT) in eastern Indonesia faces substantial development challenges.The province is one of the least developed in Indonesia, with a per capita GDP one-quarter of the national average.In 2017 incomes of 25% of the rural population of NTT were below the poverty line (Rp 329,136 or AU$ 32 per month), compared with the national average of 13% (BPS 2018).Agriculture is a central economic activity in NTT, and livestock production makes up 16% of agricultural GDP.There are 60,000 livestock producers in NTT, the majority of whom own cattle (DGLAHS 2013), threequarters of which are small-holders with 1-10 head (Mullik 2012).In particular parts of NTT, cattle sales can represent more than 80% of the family's cash income (Nimmo-Bell and ICASEPS 2007).Cattle also play a social role for ceremonies and as a source of 'savings' that can be cashed-in to meet large cash outlays including housing, school fees, health and transport.
In response to high beef prices over a sustained period, cattle numbers have almost doubled in the past decade from 555,000 in 2007 to more than one million in 2017 (BPS 2018).However, productivity is low as indicated by annual turnoff rates of just 13%, due to low weaning rates and long periods of slow growth to reach sale weight, for either slaughter or live cattle export (Waldron et al. 2012).Cowcalf production is conducted mainly in extensive grazing systems in NTT and various measures have been taken to improve reproduction and reduce calf mortality (Copland et al. 2011).The emergence of a cattle-fattening sector has potential to increase growth rates to reach sale weights at an earlier age and to create demand for feeder cattle (from the cow-calf sector).Of particular interest in relation to this paper, cattle-fattening has the potential to generate positive cash flows that are increasingly required in a society transitioning from a subsistence to a cash economy.
Forage tree legumes, particularly leucaena, have been identified as a central feed source for the development of the small-holder cattle-fattening sector.This has given rise to a body of literature on the impacts on cattle growth of feeding tree legume forage (e.g.Dahlanuddin et al. 2014Dahlanuddin et al. , 2019; Shelton and the Project Team 2017).
However, there has been a dearth of studies to verify the economic incentives for households to take up and sustain the systems, which is the focus of this paper.

Sites
The economic analysis on which this paper draws was conducted across 3 sites in the provinces of NTT and NTB, which have differing characteristics.This paper focuses on the West Timor village site of Oebola, where Bali bulls are fattened in pens on a leucaena-based diet.Leucaena is strip-planted with corn.The system is widely applicable to other areas of West Timor including Kupang and Amarasi, which are the largest cattle-producing areas in NTT.Results for the Sumbawa site of Jati Sari are reported in this issue (Dahlanuddin et al. 2019), including cattle systems, history, adoption, productivity, economics and meat output from the village.

The model
To assess household structures and incentives for FTLbased fattening, a bio-economic model was developed for a representative cattle-fattening household in Oebola.It is a partial budget, insofar as it focuses on the activity of leucaena production and cattle fattening.It is also a steady-state budget, with production and returns assessed over a specified fattening period, which is almost always less than 365 days.The focus on leucaena-fattening means that the model accounts for virtually all direct cost and revenue items.However, the budget does not account for environmental externalities including soil enrichment and reduction in over-grazing.
It is problematic to conduct a 'with and without' economic analysis of leucaena-fattening systems.Smallholders did not fatten cattle as a specialized activity prior to the extension of the systems.As elaborated in the 'scenarios' section of the paper, it is economically unviable to fatten cattle on a diet of crop residues and grasses, and grain is prohibitively expensive.Thus this analysis begins with a detailed discussion of a single scenario of fattening with leucaena, namely a representative (typical) household in Oebola fattening 4 cattle in the wet season, with average prices over the period.Alternative scenariosbased on season, weight gains and pricesare then examined to test production and income effects.The model reports on various measures of profitability, the most relevant being 'returns to labor'.All budget items and formulae are explicitly stated in Excel spreadsheets that are publically available on request.All values are expressed in Rupiah and the exchange rate adopted is Rp 10,000 to AU$ 1 in 2014 when the research and fieldwork was conducted.

Data
Production data for the research were drawn from ACIAR project LPS/2008/054, which monitored 8 households in Oebola during 2013 and 2014 with a total of 30 head of cattle between them (Pakereng 2015).Price data collected from the monitored households were cross-referenced with meetings with traders and butchers and with weekly beef price data (MoA, various years).Costs and sales data were collected through focus groups and semi-structured interviews with 5 farmers.Village-level data were used to place the farmers in context and to establish a 'representative' or 'typical' fattening household, which is reported in this paper.Fieldwork was undertaken in August 2012, May 2014 and July 2015.

Background on Oebola village
The budgeting for NTT focuses on Oebola village, Fatuleu Sub-District, Kupang District, West Timor.The system is based on corn cropping with strip-planting of leucaena and the fattening of Bali bulls in pens owned by individual households, all of which immigrated from other parts of Indonesia.In 2015 village statistics reported 276 households with an average of 4.2 household members.Household cropland sizes were 0.5-1 ha per household but some households had up to 2 ha split into parcels.Cattle were grazed collectively on village land and sometimes government forest land.Ninety-five percent of households earned a living from agriculture for both own consumption and sales.The main crops were corn (a single crop in the wet season), pumpkins and beans.Livestock included cattle, pigs and chickens.There were 1,453 cattle in the village, i.e. an average of 5.3 head per household.Cow-calf production in grazing systems predominated and calves were usually taken through to slaughter age.Many households bought-in feeder cattle to use in specialized feeding operations and most households fattened only 1 or 2 head at a time, with a maximum of 8.

Cattle fattening in Oebola
The budget results are summarized in Table 1 and detailed throughout the rest of the paper.Based on average results of monitored households (Pakereng 2015) the representative household sourced 4 bulls from its own herd or bought them in, with a starting live weight of 189 kg.Feeding periods were variable but averaged 170 days.There was a large difference in feed regimes and weight gains between seasons.In the wet season, bulls were fed a diet amounting to 2.5% of their body weight, comprised of 80% FTL (60% leucaena, 20% gliricidia), 17.5% native grasses and leaves and 2.5% corn silage.Average daily weight gain (ADWG) was 0.4 kg/d and bulls were sold at 257 kg live weight.In the dry season bulls were fed only 40% FTL (30% leucaena, 10% gliricidia) plus 60% native grasses and leaves and ADWG was only 0.2 kg/d, with bulls sold at 223 kg live weight.

Revenues
Cattle sales accounted for 98% of all cattle revenues for the household.Prices for both feeder and finished cattle were Rp 29,000/kg live weight in July 2015 (with price variations discussed in scenarios below).The model also accounted for revenue from manure and timber.While these were only 2% of total revenues, they were significant as a percentage of value added from fattening (finished cattle cost minus feeder cattle cost) at 7 and 12%, respectively.Of the manure produced (35% of DM intake), 10% was sold (Rp 250/kg dry), 20% was used for fertilizer (valued based on the substitution of urea and NPK fertilizers) and 50% was unused (which can have negative environmental impacts).The remainder (20%) was used for biogas which was valued based on reduction in household labor spent collecting firewood (1 hour per day) and kerosene usage for cooking and light in the household.The branches of leucaena trees were used for firewood.If 2 branches were used per cut (every 120 days) then 3,600 branches were collected over the fattening period with a value of Rp 100,000.

Capital investments
The representative household made capital investments in items used for multiple household activities over extended periods.The cost (both cash and labor) was amortized over the economic life of the asset and attributed to cattle fattening over the fattening period.
Table 1.Revenues, costs and returns (in Indonesian rupiah; 1 AU$ = 10,000 Rp) on leucaena-based cattle fattening for a representative household in Oebola village, West Timor, and weight gain and price scenarios.Highlighted cells refer to the scenarios analyzed in the paper and the key indicator of 'returns to labor'.Source: Author calculations; all values are expressed for a fattening period for the number of cattle specified, except 'returns to labor', which are expressed on a per day basis.
To meet dietary requirements of the 4 bulls, the household required a total of 1,200 trees with a 120-day interval between harvests, strip-planted on the land of the household.Planting costs included fencing, purchase of seeds, nursery (poly bags, bedding, shade cloth) and transplanting (labor and transport).The modest up-front costs (Rp 308,000 in cash and 13 person days labor) were negligible when depreciated over 40 years.
The costs of constructing a pen (nails, wire, timber, cement, sand, gravel, reinforcing, troughs, roof and other items) were higher than for trees (Rp 1.35 million) but also low when depreciated, given a lifespan of over 20 years and fattening 160 cattle.The 'design capacity' of the pen was 5 head.Given the actual number (4 head) and time not on feed (26 days) capacity utilization was 72%, which increased overhead costs per head.
The cost of a motorbike (used to transport feed and marketing of cattle) was high but the machine was used for fattening for only 20% of the time and was depreciated over a life-span of 15 years.Most households had a well that lasts 15 years at a cost of Rp 1,500,000 plus the costs of meals for workers that dig it.Biogas facilities (pits and converters) are commonly installed in West Timor to use effluent from the pens.Equipment is free (as part of a government program) but costs are incurred for meals for installers and household labor, especially to dig the pit and for maintenance (cleaning out pits and lines).
When costs of all capital items were amortized, depreciation costs totaled Rp 186,000 per fattening period.Together, these made up <1% of total costs and were eclipsed by other costs, so appear to be small.However, it is important to note that these are up-front costs (in land, labor and capital) that can be significant for households when first investing and can be a barrier to adoption.While loans are available through formal and informal channels including banks with subsidized loans, traders and profit-sharing arrangements with other households (Waldron et al. 2012), households usually use their own limited savings for these infrastructure items.

Production costs
Production costs are incurred specifically for cattle fattening on a daily basis or within the fattening cycle and are directly linked to production volumes.Feeder cattle purchase costs accounted for 94% of total costs.Selfproduced bulls were valued as an opportunity cost (that could otherwise be sold) or as a cash cost when bought in.Costs of cattle purchases, mainly from a nearby live cattle market, included search costs (telephone, fuel and household labor), trucking and broker fees.Cattle were sold to traders at the farm gate and so incurred negligible sale costs.
After leucaena has been established there are no additional cash costs but there are significant labor costs for collection and transport.In the wet season all members of the household traveled an average of 1 km (range of 0.5-5 km) to collect forage twice per day, taking 1.5 hours.Motorbike fuel was Rp 700 per day.The collection of native grasses and leaves in the wet season was less timeconsuming because it comprised the smaller part of the diet but was more labor-intensive to collect from scattered trees and bending over to cut grass.Labor costs to collect native grasses and leaves in the dry season were high (2 hours per day).Corn stalks were fed after harvest (at the end of the wet season), which required labor for cutting, transport and storage over a few days but little after that.
Several households in the group paid Rp 240,000 per year for access to group water supplies (access, pipe maintenance, fuel for pump) of which about 30% was used for cattle fattening.The household spent 30 minutes per day collecting and distributing water to the troughs in the pen.The household spent another hour per day in cleaning pens and cattle management.
Veterinary costs included vaccination (for anthrax and haemorrhagic septicaemia to allow for live export), one medical check (from local vet), a vitamin supplement and a small amount of salt.The total veterinary treatment costs (Rp 364,000 over a fattening period) were the second highest cash outlay.
Crop shading and moisture competition were included as additional costs of production.When 1 ha of corn is planted in the wet season and strip-cropped with leucaena, it is assumed that the grain yield (2,400 kg/ha) is reduced by 10%.Valued at Rp 3,000/kg, the forgone revenue is Rp 720,000 or Rp 335,000, when allocated over a fattening period.

Returns to cattle fattening
The returns (or profits) were estimated in different ways to provide different measures of profitability.Subtraction of costs from revenues provided 'Gross profit', which was positive in the wet season (at more than Rp 7 million over the fattening period) but declined to less than half of this in the dry season.Capital costs were deducted from gross profits to give 'Net profit'.Few households took out loans, but an opportunity cost was applied to the money invested in cattle that could otherwise be invested.The interest rate on a deposit in a savings account was used to value the opportunity cost of capital (8%).For large and expensive inputs like feeder cattle, the opportunity costs of capital were significant.
The next section on family labor reports the labor inputs into cattle fattening.The majority of labor was used on feed collection and watering, followed by work in pens, then by cattle marketing.Labor input into infrastructure was allocated over the fattening period and was small.The total labor inputs were converted into hours per day in cattle fattening (i.e. 4 hours for 4 cattle in the wet season), then converted into an 8-hour working day (i.e.half a day).
'Net returns' were divided by total labor input to derive 'Returns to labor', which provided the most useful indicator of profitability.This provided an indication of the profits from cattle production that a household was making from their own labor and management, and allowed comparison with other farm and off-farm work.Results for the representative household suggested that returns to cattle fattening in the wet season were positive (Rp 61,463), which compared favorably with average offfarm work (Rp 45,000).Comparisons were not so favorable in the dry season.At Rp 12,097 per day, income was at or below the poverty line.However, it must be considered that cattle can be produced all-year-round compared with off-farm work, which can be seasonal or inconsistent.Farmers may also be attracted to the customs and pride of running their own enterprise.
Budget results from 1 ha of corn in Oebola using (low) yields from 2015 and 2016 suggested that returns to person days were comparable with cattle fattening in the wet season.Because of its central role in household consumption and cash sales and its agro-climatic suitability, farmers in Oebola continue to grow corn.While strip planting of leucaena reduces corn yields (by 10%), it is integrated into the corn cropping, and not a substitute activity.

Scenarios
While the discussion above examined an average household in 2 seasons, there is large variation between households and natural and market conditions.A range of scenarios are reported in Waldron et al. (2015) including changes to rations, weight gains, fattening period, price, capital investment, labor cost and sales channels.This paper examines just the major variablesweight gain and price.Weight gain.Profitability of feeding in the wet season was far greater than in the dry season for the representative household.By far the most important determinant was the difference in ADWG (0.4 vs. 0.2 kg/d) due to diet and compensatory weight gain early in the wet season.The labor cost in collecting native grasses and leaves in the dry season was also slightly higher than collecting FTL leaves in the wet season.The differences in ADWGs lead to returns to labor of Rp 61,463 in the wet season more than 5 times the returns in the dry season (Rp 12,097).
Households in the group with highest gains in the wet season (0.8 kg/d) recorded very high returns (Rp 153,788), while those with lowest weight changes (-0.2 kg/d) operated at a heavy loss 024).Prices.Profitability is also sensitive to market trends and especially the relative prices of feeder and finished cattle, which are a function of market conditions, weather, household conditions, the skills of buyers and sellers in appraising cattle (visually) and timing (ceremonies, the issue of export permits or when school fees are due).If finished cattle prices are 15% higher than feeder cattle prices, returns increase strongly by 86%, but if they are 15% lower than feeder prices, cattle fattening is barely a break-even activity (Rp 9,170).Returns without FTL.It is also useful to examine returns to fattening without leucaena.This is done in the modelling by assuming a diet of improved grasses (80%) and corn stover (20%).However, this would not be possible throughout the wet season because corn is harvested at the end of the wet season, unless corn stover was carried over from the previous season or was purchased in.While the stover could be stored and used in the dry season, quantities of grass available would be insufficient or very time-consuming to collect.Weight gains are reduced to 0.15 kg/d, which is a generous assumption given comparisons in various feed systems and locations (Quigley et al. 2009;Panjaitan 2012;Dahlanuddin et al. 2014).Time to chop stover increases from 0.1 to 0.5 hours, and to collect grass and leaves from 1 to 2.5 hours.In this case, 'Returns to person days' are very low at Rp 4,821, indicating that cattle fattening is not biologically or commercially viable without leucaena.

Conclusions
Analyses reported in this paper confirm the intuitive understanding that cattle fattening on a leucaena-based diet is biologically and economically viable for most small-holders in West Timor under most conditions.Leucaena is a low-cost input, provides feed through the dry season (albeit at a lower yield) and generates reasonable weight gains.Cattle fattening is capitalintensive, but allows for rapid turnover of both cattle and capital.Cattle fattening is not land-intensive, can be done under various ownership (owner-keeper) arrangements, and is relevant for a wide range of households.These factors explain the growth of leucaena-fattening systems in NTT.
Growth in the sector could be disrupted if circumstances differed.Beef prices were buoyant for most of the 2010s and market fundamentals remained strong, but prices have fluctuated in recent years (for policy reasons, including trade policy and domestic support), which may affect incentives.Cattle fattening with leucaena is relatively labor-intensive.In countries and regions where there are good alternative opportunities for work and wages, small-holders can be drawn out of cattle production (Waldron et al. 2018).While cattle fattening may provide reasonable returns for small-scale farmers (such as the representative household in this study), profits of AU$ 6 per day will not make them rich.Future growth in the sector may see the emergence of more large, skilled and entrepreneurial fatteningtrading households, which could utilize leucaena sometimes through land and cattle-/land and laborsharing arrangements.

Final
Report.Australian Centre for International Agricultural Research (ACIAR), Canberra, ACT, Australia.bit.ly/33JxicJWaldron S; Pham L; Smith D; Hieu PS; Dong XX; Brown C. 2018.Macro developments in the China and Southeast Asia beef sector.Animal Production Science 59:1001-1015.doi: 10.1071/AN17434 (Accepted 7 March 2019 by the ILC2018 Editorial Panel and the Journal editors; published 3 September 2019) © 2019 Tropical Grasslands-Forrajes Tropicales is an open-access journal published by International Center for Tropical Agriculture (CIAT), in association with Chinese Academy of Tropical Agricultural Sciences (CATAS).This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license.