The rapid uptake of treated crop residue technology by smallholder farmers in central China can be attributed to several favourable factors of which its profitability is the most significant. The paper discusses the method of analyzing input:output data generated by FAO/UNDP Project CPR/88/057 to determine the supplement levels which yield maximum profit to the farmer. In the present cost:price environment they occur between 1 - 2 kg cottonseed cake per head per day, depending on whether maximum profit per day or per head is desired. The authors also indicate the important parameters determining profitability, particularly the cost of protein supplement, stressing the need to continually monitor them in the dynamic agricultural economic environment.

The recent explosion in the uptake of supplemented treated straw technology for cattle production in China is due to a number of favourable coinciding factors, but fundamentally because it is sufficiently attractive to farmers to promote its rapid adoption. This is firstly because it is generally profitable for fattening cattle in the current economic environment, is easy to adapt from the traditional methods of feeding livestock, it fits snugly into the farming system, raw materials are abundant and still cheap and, for all these reasons, it is a low risk innovation. In addition it is strongly supported by government because: i) it reduces the pollution resulting from the traditional straw burning after harvest; ii) it increases livestock production using a hitherto waste material; iv) it raises farmers' incomes; and v) it substitutes a non-human food source for the human food grains normally used in intensive animal production systems in China. This support takes the form of trained technical assistance at the village level, ready availability of credit at concessional interest rates (0.7% per month) through the Agricultural Bank, and assured supplies of urea for treating cereal straw: 200 000 tinnes in 1992 at subsidised rates.

An FAO pilot project in the Provinces of Henan and Hebei during 1990 - 1992 developed the production parameters and husbandry guidelines (Finlayson 1992; Dolberg 1992) for its wider adoption.

The project areas are located in the eastern central plains of China, which are characterised by flat to slightly undulating land with soils of alluvial origin which are fairly rich in organic matter. Despite below-zero temperatures in winter the basic resources support two crops annually; wheat sown in the autumn on practically the whole area, followed by cotton and some maize as a summer crop. Cropping intensity is close to 2.0 on the arable area. Underground water is exploited for supplementary irrigation of the wheat during the spring in parts of the area, and also vegetables and medicinal crops which are grown on small areas, especially near urban centres.

The crops are heavily fertilised with inorganic (up to 1.5 tonnes/ha of urea or ammonium phosphate) and organic manures to maintain high production levels (wheat up to 4.5 t/ha). The cereal crop residues are usually removed or burnt, rarely incorporated due to lack of time between crops and the denitrification effect on the following crop. Land preparation and wheat seeding and harvest are generally mechanised, although animal and human draught are also common, and manual labour is used for the cotton and maize harvest.

Village-based smallholders cultivate the major part of the area with farm sizes proportional to family labour availability, but typically in the range of 0.3 to 1.5 ha, while larger state farms occupy the remainder. Most families own some livestock which comprise mainly small animals for domestic use and sale, of which pigs and poultry are common, with rabbits and birds less so. Large animals include horses, donkeys, and cattle for breeding and fattening, although some are used for draught, but not every farmer raises larger animals and those who do have only small numbers, typically a donkey or horse and 1-3 cattle, depending on labour resources and motivation.

Maize grain is produced mainly to feed the pigs and chickens, while the stover is kept for the larger animals as semi-dry `silage'. In recent years the wheat straw and cottonseed cake (CSC) have been increasingly used to fatten cattle, and the manure returned to the crop area. The farming system is thus fairly stable and seems sustainable, although highly dependent on two off-farm resources: organic fertiliser for crops, and urea (or ammonia) for the treated straw. To that extent crop and livestock production based on this technology are in competition for a limited resource.

Most farmers are intuitively quick to grasp potentially profitable activities - and Chinese farmers are no exceptions - and so many try out new techniques which appear attractive even without the benefit of scientific refinement. In fact, the technology has been used in central China since 1985. However, because the application was not guided by local applied research, implementation by farmers has tended to be somewhat haphazard, even though profitable. In addition, recommendations by Chinese technicians were based on out- moded livestock feeding technology and with little regard for economic considerations, especially the concept of input optimisation for maximum profit.

It is not possible to determine the most profitable input level, or combination of inputs in the case of several animal feeds, and hence make recommendations to farmers, using physical input:output responses alone. However, the physical data are very necessary as the basis for the economic calculations upon which sound, profitable recommendations can be made. It follows that the more reliable and consistent the physical response function, the more precise will be the economic analysis results and hence husbandry recommendations arising therefrom.

The on-farm and on-station trials undertaken by the Chinese technicians and farmers provided very useful and quite reliable input:output response information. Data from two trials (Zhang Weixian et al 1994) only are analyzed in this paper. The actual trial data were submitted to regression analysis enabling calculated response curves to be compared with the actual results. Although many production functions can be explained by a quadratic second-degree polynomial (Gomez and Gomez 1988), due probably to substitution of the supplement for straw at the higher levels of CSC intake, these data fitted closely an exponential expression of the form Y = a - be^-x with high reliability. This mathematical expression of the results enabled missing data to be determined, and irregular data to be corrected. Table 1 sets out the relevant data for each location.

Because the actual growth rate at 2 kg/day of CSC in Henan (655 g/d) was apparently inconsistent, it was omitted from the data series used to determine the regression. To clarify the point and also confirm the different responses in each trial, they should be repeated. The calculated figures were used to determine the input optimisation/profitability response data.

The main variable financial parameters used in the calculations for each area are set out in Table 2. In Hebei Province farmers use both forms of nitrogen, and so the parameters for each are presented for completeness. In some parts of Hebei, straw has a commercial value.

The main financial parameters are set out in Table 2. However, the cost of the pens, animal health and the tank for urea treatment are included in the cost calculations.

Table 1: Growth rate data used in the economic analysis
HENAN (Urea)			HEBEI (NH3)
CSC (X)	Daily gain, g/d		CSC (Y)	Daily gain, (g/d)
kg/d	Actual	Calculated	kg/d	Actual	Calculated
0	250	236	0	63	135
1	600	639	0.25	370	319
2	655	786	0.5	529	463
3	845	841	1.0	-	663
4	833	861	1.5	781	783
			2.0	829	857
			2.5	892	901

Regression equation: Y=873 - 637e^-x R2 = 0.99
Regression equation: Y=970 - 835e^-X R2 = 0.98

Table 2: Main financial parameters in the project areas(Ø/kg,1992)
Parameter	Henan	Hebei	Hebei
	Urea	NH3	Urea
Cattle purchase	3.6	3.6	3.6
Cattle sale	3.8	3.6	3.6
Straw value	0.00	0.00	0.06
Urea	1.1		1.1
NH3		1.2
CSC, normal	0.4	0.4	0.4
, scarce	0.6	0.6	0.6
Plastic	8	8	8

The treated straw intake figures at air-dry equivalent weight for each level of supplement are expressed as a % of cattle live weight, based also on the trial data (Table 3). These are used to determine total straw intake and hence N-treatment quantities and costs.

The profitability analysis is based on an initial live weight of 180 kg and a constant sale weight of 450 kg for all levels of supplement. It is also assumed that the cattle maintain uniform growth rates throughout the fattening period, although this needs to be verified.

It should be noted that in some areas, sale and purchase price (per kg liveweight) of cattle are not the same. Urea is at market price, not at the concessional government price, while straw values may differ depending on the proximity to a paper factory or to an inter-farm straw market. Farm labour is not included because it is generally agreed that this input is offset by the manure value - approximately Ø60/head. Tables 4 and 5 below summarise the financial performance of the cattle fattening activity in the central plains environment, using the Henan trial data to demonstrate the methodology, and Hebei (NH4) shows the profit results only.

Table 3: Intake of treated straw for each level of supplement
Supplement, kg/d	NH4	Urea
	air-dry as % liveweight
0	3.1	3.1
0.5	2.6	*
1.0	2.45	2.7
1.5	2.3	*
2.0	2.1	2.3
3.0	1.9	*
4.0	*	2.1

Table 4: Effect of supplement level on financial performance (Urea-treated Straw - Henan)
Item	Unit	Data
Supplement	kg /d	0	1	2	3	4
Daily gain	g /d	236	639	786	841	861
Gross income	Yuan/hd	1710	1710	1710	1710	1710
Total costs	Yuan/hd	1620	1154	1176	1259	1338
Feed costs*	Yuan/hd	663	386	428	517	598
PROFIT
Per head*	Yuan	90	556	534	451	282
Per day*	Yuan	0.08	1.31	1.56	1.40	0.90
Break-even gain	g/d	144	218	300	394	530
Break-even
CSC	Ø/kg	-	1.71	1.18	0.87	0.62
Fattening period	Months	38	14	11.4	10.7	10.4
PROFIT
Per head**	Yuan	90	471	397	258	32
Per day**	Yuan	0.08	1.11	1.16	0.80	0.10
--------------------	--------	------	------	------	------	------

Note: * CSC = Ø0.40/kg; ** CSC = Ø0.60/kg.

Because of the periodic seasonal scarcity of cottonseed cake, profit was calculated at two price levels representing normal and scarce supply situations. Some parts of Hebei Province recently suffered high prices due to low cotton production. In addition, profit was determined per head for the duration of the fattening period, and per day. The `break-even' daily gain is the growth rate required to cover variable, mainly feed, costs. Break-even CSC is the price of the cottonseed cake which would result in zero profit. Total costs include the purchase price of the cattle.

Table 5: Effect of supplement level on financial performance: NH3-treated straw, Hebei
Supplement	kg/d	0	0.5	1	1.5	2.5
Profit
Per head*	Yuan	-719	404	501	511	471
Per day*	Yuan	0.36	0.69	1.23	1.48	1.57
Per head**	Yuan	-719	346	420	408	321
Per day**	Yuan	-0.36	0.59	1.03	1.18	1.07

The input level of maximum profit can also be determined directly from the physical response curve (Pervaiz et al 1989). It occurs where the tangent to the production function curve is equivalent to PX/PY from the equation, Y1*PY = X1*PX, at the margin where Y1 and X1 are both equal to unity. (P= price; X = input unit; Y = output unit).

The economic analysis confirms the profitability of the technique in the present economic environment. It also demonstrates clearly that the level of input at which maximum production occurs is considerably higher than that at which MAXIMUM profit occurs. This means that feeding beef cattle at more than this optimum rate results in a lower PROFIT to the farmer and also involves a waste of resources, particularly cottonseed cake. In fact, the level of inputs at which maximum profit and maximum production coincide occurs only when the cost of the input to the farmer is zero.

It is important to recognise that maximum profit per head (around 1 kg/day) does not coincide with maximum profit per day (around 2 kg/day) and these results are consistent at each trial, even though the shape of the curves differ. This means that farmers who are content to fatten one animal per year, for example, should use the lower level of supplement, whilst those who wish to fatten animals continuously, that is, obtain maximum turn-over, should use the per day rate of supplement. Furthermore, to take advantage of seasonal price variations, an additional strategy is to purchase weaner cattle when prices are low and sell when they are high.

The results also demonstrate that the optimum rate of supplement is sensitive to its cost. Within the present range of CSC prices, the profit margin is clearly reduced at the higher price, but the maximum profit input level does not change. However the data can be extrapolated to show that as the cost of CSC goes beyond Ø0.60/kg, maximum profit will occur at lower rates of supplement. Indeed it is quite easy to establish a suite of CSC prices and inputs within which maximum, and any profit occurs. Similarly the profit relationship is affected by other parameters including the price of straw, difference between purchase and selling price of the animals (in Hebei this is a significant factor), cost of urea and ammonia, etc. This means that because these factors change, even though not dramatically, they have to be continually monitored and the profit response curves adjusted accordingly to determine recommendations about appropriate levels of supplement, in each province at least annually. This will become even more important in future as the price of cottonseed cake rises in response to increasing demand. Because of the dynamic and variable economic environment it is not possible to establish a `fixed' supplement level `recipe', even for one province.

The break-even growth rates are important indicators of supplement level responses, especially for the farmer, as they enable non- profitable animals which, because of ill health or low genetic growth potential, to be identified and disposed of early if weighings are made regularly. Similarly, it is worth noting that animals that are identified with high genetic growth potential should be retained for breeding instead of being slaughtered for meat.

Finally the results demonstrate the critical importance of undertaking this type of economic analysis before making recommendations to farmers about production input levels, not only for feeding cattle but for crop production and any other system where output response to increasing doses of inputs is non-linear. For this reason also such research needs to involve the economic considerations even in the planning and design stages.

(Received 1 September 1994)