巴西最高产量和最大经济产量研究


Tsuioshi Yamada
POTAFOS P. O. Box 400 POTAFOS P. O. Box 400 POTAFOS P. O. Box 400 POTAFOS P.O.Box 400 13400 Piracicaba-SP Brazil 13400 Piracicaba-SP Brazil 13400 Piracicaba-SP Brazi


　　原文摘要 ABSTRACT

　　最高产量研究在巴西农业研究中可以说是特殊的例子。这使得最大经济产量研究几乎不可能。因此本文将一些农民为达到100吨/公顷甘蔗，3000公斤/公顷早稻和3100公斤/公顷棉花（纤维+棉籽）的高产量所采取的管理措施作一介绍。

　　Maximum Yield Research or MYR is almost an exceptional case in Brazilian agriculture research. This fact makes Maximum Economic Yield or MEY results almost impossible. Thus, in this paper is reported the management practices done by some farmers to reach high yields aiming at the production of 100 t/ha of cane, 3,000 kg/ha of upland rice and 3,100 kg/ha of cotton (fiber + seed).

　　1.引言 INTRODUCTION

　　随着农业出现不仅是个体农民之间，而且是国家之间的竞争，Holt（1987）指出美洲农业的生存战略包括：综合性、多样性、产品分化和成为低成本生产者。这也适用于巴西农业。在现代市场竞争的重压下，只有效率高的农民才能生存。既使他们情况也不妙，他们中许多人被迫部分转向牛肉生产或其它生产。当然，同在美国一样，如果他们想继续从事农业，还要使用其它一些谋略。

　　Within the emerging reality in agriculture of competition not only among individual farmers but also among states and nations, Holt (1987) points out the following strategies for the survival of the American agriculture: integration, diversification, product differentiation and being the low-cost producer. This is also true for the Brazilian agriculture. Under the pressure of the present highly competitive market only the efficient farmers are surviving. Even for them the situation hasn't been easy making many of them to shift partially to beef production or other activities. Certainly, the other strategies will have to be used here, as in the States, if they want to continue in the agriculture.

　　2. 巴西最高产量研究（MYR） MYR IN BRAZIL

　　从事最高产量研究的人员及其他了解其内容的人士普遍接受的定义是：以多学科体系来寻找所研究的地区达到最大产量潜势涉及的变量及其交互作用。因最高产量并非最经济，所以仅对研究者有用，而最大经济产量（MEY）对农民有用（Wagner和Dibb，1990）。

　　The accepted definition by MYR researchers and others who understand its intended purpose is "the study of variables and their interactions in a multidisciplinary system that seeks to achieve full production potential of the research site". Since maximum yields are not the most efficient, they are for the research and maximum economic yields (MEY) are for farmers (Wagner and Dibb, 1990).

　　因难于组织一批多学科人员一起工作，所以巴西只有一个研究符合最高产量研究的概念。即对大豆的一项研究，其结果曾由Yamada（1990）提交。

　　Due to difficulties in having multidisciplinary teams working together, only one research is being done in Brazil under MYR concept. It is the one on soybean which results were already presented by Yamada (1990).

　　3.巴西最大经济产量结果 MEY RESULTS IN BRAZIL

显然，没有最高产量研究便没有最大经济产量结果。因此这里提供的只是一些高产农民获得的结果。Yamada（1990）在京都最高产量研究会议上提交了有关玉米，大豆和咖啡的结果，本文再提供甘蔗、旱稻和棉花的结果。 　　Without MYR, evidently there is no way to have MEY results. Thus, what will be presented here, are yield data obtained by some top farmers. Results on corn, soybean and coffee were already presented by Yamada (1990) in Kyoto MYR Symposium and in this paper will be presented results on sugarcane, upland rice and cotton.

　　4.甘蔗 SUGARCANE

　　4.1.影响生产力的因素 Factors affecting productivity

　　由于制糖和酿酒双重用途，甘蔗种植面积已从1972年的1800万公顷增加到1992年的4200万公顷，使巴西成为世界最大的甘蔗生产国。

　　Thanks to its dual purpose of sugar and alcohol production, sugarcane area increased from 1.8 million ha in 1972 to 4.2 million ha in 1992 making Brazil the world largest sugarcane producer.

　　同时，改良品种和改进管理措施使圣保罗州的糖厂生产提高了53%，由1972年的51吨甘蔗/公顷上升到1991年的78吨/公顷（Nunes Jr.，1992）。根据Demattê(1992)的报告，一些糖厂5~6茬甘蔗平均可产83吨/公顷。

　　At the same time, the improved cultivars and the better management practices allowed an increase of 53 % in the productivity among Sao Paulo State Sugar Mills, from 51 t/ha in 1972 to 78 t/ha of cane in 1991 (Nunes Jr., 1992). According with Demattê (1992) some sugar mills are obtaining an average of 83 t/ha of cane in 5-6 cuts.

　　以前甘蔗占用高肥力土壤，现已转向低肥力土壤。这些土壤上推荐使用的技术措施有：石灰与石膏同时施用以加深根系，热法磷酸盐作磷肥，并使用酒糟、滤饼和蔗渣等糖厂副产品作为有机质和必需养分来源。

　　After having occupied previously the high fertility soils, sugarcane is now moving toward low fertility soils. Among the technologies recommended for these soils are: lime associated with gypsum for a deeper root system, P-fertilization with thermophosphate and the use of sugar mill by-products like vinasse, filter cake and bagasse, as source of organic matter and some essential plant nutrients.

　　4.2.改良品种 Improved cultivars

　　70年代CB 41-76占总播种面积的35%，80年代初，NA 56-79占43%。80年代中期以后，种植者有了更高产的品种，这是增产的主要原因（表 1）。可发现含糖量提高了23%，从每吨甘蔗含糖93.6公斤增到115.0公斤。

　　In the seventies the cultivar CB 41-76 was responsible up to 35% of the total planted area and in the early eighties NA 56-79 occupied 43 %. After mid-eighties more productive cultivars were offered for the growers, which were in great account responsible for the additional yield increase shown in Table 1, according with Nunes Jr. (1992). It can be seen a 23% increase in sugar content, from 93.6 to 115.0 kg of sugar per ton of cane.

(表：表1 蔗糖生产力和栽培品种和演进(Nunes Jr.,1992) )

　 Year 年份
1972 1977 1983 1988 1991*
Cane甘蔗(t/ha)吨/公顷 51.00 58.00 61.00 76.00 78.00
Sugar糖(t/ha)吨/公顷 4.77 5.41 6.27 8.49 8.27
Sugar糖(kg/t)公斤/吨 93.60 93.30 102.80 111.70 115.00
Cultivars 栽培种 (%)
CB 41-76 5.20 35.00 7.00 0.10 0.10
NA 56-79 4.00 15.00 43.00 25.70 11.70
SP 70-1144 0.00 0.00 9.00 28.60 24.20
SP 71-1406 0.00 0.00 0.30 16.00 24.00
SP 71-6163 0.00 0.00 0.10 13.00 17.40

注释：* Estimate.估计值
　　4.3.石灰和石膏配施 Lime associated with gypsum

　　大约70%的甘蔗种在酸性土壤上，盐基饱和度低，其对甘蔗生产的主要限制为根系所及的土壤体积太小，一般深不过60厘米，虽然根系原本能深及120~200厘米。Morelli等（1992）在瘠薄暗砖红壤上，进行了石灰和石膏同施试验，以克服这种问题，0~25厘米和25~50厘米土层化验结果和粘粒含量见表2。

　　Seventy percent of the area cultivated with sugarcane are of acid soils with low base saturation. Under such situation the main constraint for sugarcane production is the small volume of soil occupied by the root system, no deeper than 60 cm, although it can reach 120-200 cm. Lime associated with gypsum were tested by Morelli et al. (1992) to overcome this problem, in a dystrophic dark red latosol with analyses and clay contents in the layers 0-25 cm and 25-50 cm presented in Table 2.

(表：表2 化学分析和粘粒含量(Morelli et al.,1992) )

Depth 深度 pH O. M. 有机质 P 磷 K 钾 Ca 钙 Mg 镁 Al 铝 CEC阳离子交换量 Clay 粘粒
(cm)厘米 (H2O) (%) (ppm) (meq/100 cm3) 毫克当量/100厘米3 (%)
0-25 4.9 1.4 3 0.06 0.34 0.14 0.59 3.5 12
25-50 4.9 0.9 2 0.04 0.12 0.05 0.65 2.8 16

　　试验采用4个水平石灰用量（0、2、4、6吨/公顷）和 4个石膏用量（0、2、4、6吨/公顷）。所用为白云质石灰，含28%氧化钙（CaO），15%氧化镁（MgO），有效碳酸钙当量（ECCE）为63%。石膏是磷石膏，磷酸生产的副产品，含28%CaO，10%硫（S）和l%五氧化二磷（P2O5）。甘蔗品种SP70-1143，播种每公顷施39公斤N，120公斤P2O5和120公斤K2O。种后第8个月每公顷追施32公斤N和40公斤K2O。以尿素，磷酸一铵和氯化钾形式施用。

　　Four doses of lime (0, 2, 4 and 6 t/ha) were tested with four doses of gypsum (0, 2, 4 and 6 t/ha). The lime was dolomitic with 28% of CaO, 15% of MgO and 63% of E.C.C.E. (effective calcium carbonate equivalent). The gypsum was phosphogypsum, by-product of phosphoric acid production with 28 % of CaO, 14 % of S and 1% of P2O5. The sugarcane cultivar SP 70-1143 received at planting time 39 kg/ha of N, 120 kg/ha of P2O5 and 120 kg/ha of K2O. Eight months after the planting received topdressed application of 32 kg/ha of N and 40 kg/ha of K2O. These nutrients were supplied as urea, MAP and potassium chloride.

　　表3为试验中前4茬所获平均产量。从中可看出石灰和石膏对甘蔗产量的单独作用等同。石灰和石膏配施效果更好，最好的是施用4吨石灰和2吨石膏/公顷。其成本相当于13.2吨/公顷甘蔗，可由第一茬收获的甘蔗全部支付。石灰和石膏的交互作用由其在土壤剖面中更好地分布得以解释，这由施用后29个月所测土壤化验结果给出。

　　The average productions of the first 4 cuts obtained in the experiment are presented in Table 3. It can be seen that the isolated effect of lime and gypsum on cane production was the same. Better results were obtained with the association of lime and gypsum. The best combination was 4 t/ha of lime plus 2 t/ha of gypsum. Its cost was equivalent to 13.2 t/ha of cane which was completely paid by the yield increase already in the first cut. The interaction lime ′ gypsum was explained by the better distribution of Ca and Mg along the soil profile which were detected by the soil analyses made 29 months after the application.

(表：表3 瘠薄暗砖红壤上石灰和石膏对甘蔗产量(4次收割平均值)的影响(Morelli等,1992) )

Lime石灰 Gypsum 石膏 Average 平均
0 2 4 6
Cane 甘蔗 (t/ha)吨/公顷
0 99.05 106.3 111.42 111.97 107.18
2 110.25 114 117.25 114.42 113.98
4 112.65 121.37 118.17 118.25 117.61
6 110.25 117.37 113.65 118.35 114.90
Ave.平均 108.05 114.76 115.12 115.75 　

　　4.4.热法磷酸盐磷肥 P-fertilization with thermophosphate

　　60年代初Alvarez等（1965）在圣保罗州5个不同地点所做的研究表明，甘蔗对磷肥有巨大反应。其所获年均产量见于表4。热法磷酸盐效果最好，比无磷处理增产49%。

　　Research done during the early sixties by Alvarez et al. (1965) in five different locations in Sao Paulo State showed the great response of sugarcane for P-fertilization. The average yield obtained in these locations is presented in Table 4, where the best result was given by thermophosphate, with a yield increase of 49% over the treatment without P.

(表：表4 五个地点3 种施磷量(50,100和150公斤/公顷P2O5)甘蔗平均(第一次收割)(Alvarez等,1965) )

Treatment 处理 Cane甘蔗(t/ha)吨/公顷 Index指数
NK 氮钾 63.8 100
NK + Single superphosphate氮钾+普钙 86.2 135
Bicalcium phosphate 磷酸二铵 88.7 139
Thermophosphate 热制磷肥 94.9 149
Bone meal 骨粉 87.9 138
Olinda rock phosphate 磷矿粉 80.4 126
Alvorada rock phosphate 磷矿粉 81.6 128
Araxa rock phosphate 磷矿粉 76.5 120
Bauxite phosphate 铝矾土磷肥 69.9 110

　　最近Morelli等（1992）在圣保罗州伦索伊斯?保利斯塔含有3毫克/公斤 P的石英砂质新生土上做了研究，评价热法磷酸盐撒施和沟施对土壤化学性质和甘蔗生产的影响。他们的结论是用量相同时，撒施产量更高（表5），每公顷撒施P2O5200公斤加沟施100公斤最经济。土壤分析表明，热法磷酸盐提高0~20和20~40厘米土层取样中的磷、钙和镁的水平。

　　Recent work was done by Morelli et al. (1992) in a quartz psamments with 3 ppm of P in Lencóis Paulista, Sao Paulo State, to evaluate the effects of thermophosphate, broadcasted and furrow applied, on the soil chemical properties and on the sugarcane production. They concluded that at the same amounts, higher productivity were obtained with broadcasted treatments, and that the treatment with 200 kg of P2O5/ha plus 100 kg of P2O5/ha furrow applied was the most economical for cane production. The soil analysis showed that thermophosphate increased P, Ca and Mg levels in the soil samples taken at 0-20 and 20-40 cm layers.

(表：表5 热制磷肥撒施和沟施获得的甘蔗产量(Morelli et al.,1992) )

P2O5 broadcasted撒施磷肥 P2O5 furrow applied 沟施磷肥(kg/ha)公斤/公顷
0 100 200 300
(kg/ha)公斤/公顷 (t/ha) 吨/公顷
　 Plant-cane 新植甘蔗
0 68.7 100.9 104.2 127.5
200 147.7 169.1 171.8 171.3
400 158.1 168.7 172.5 173.1
　 Ratoon-cane 截根甘蔗
0 44.7 63.6 72.9 77.1
200 91.7 97. 3 100.2 100.8
400 104.7 106.3 109.3 112.3

　　4.5.甘蔗工业有机废料的利用 Use of organic residues of the sugarcane industry

　　制糖和酿酒业的特点是产生大量副产品。Orlando Filho等（1983）指出，每吨甘蔗在不同类型的酿造厂可生产如下产品：

　　Sugarcane industry for sugar and alcohol production is characterized by the great volume of by-products it generates. Depending on the distillery type the following amounts of products are produced by one ton of cane, according with Orlando Filho et al. (1983):

　　4.5.1.糖、酒制造厂 Distillery for sugar and alcohol production

　　糖94公斤 94 kg of sugar

　　酒精12升 12 l of alcohol

　　酒糟156升 156 1 of vinasse

　　滤饼30公斤 30 kg of filter cake

　　蔗渣250公厅或在蒸汽机中燃烧生成的灰分6公斤。

　　250 kg of bagasse or 6 kg of ashes, if burned in the steamers.

　　4.5.2.酿酒厂 Distillery for alcohol production

　　酒精70升 70 1 of alcohol

　　酒糟910升 910 1 of vinasse

　　蔗渣250公斤或在蒸汽机中燃烧生成灰分6公斤

　　250 kg of bagasse or 6 kg of ashes, if burned in the steamers.

　　酒糟是最重要的一种副产品。甘蔗工业产生的大量酒糟，在70年代初是环境污染的主要原因，当时一些工厂将其倾入河中。在认识到酒糟在改善土壤肥力上的重要性后，几乎全部被用在蔗田中，表6示出每立方米酒糟中养分平均含量。最常使用在甘蔗截根苗上，配施氨水或尿素等氮肥，酒糟用量视其含K2O量而定。建议截根苗甘蔗使用一般化肥量的两倍，在圣保罗州为120~180公斤K2O/公顷（Demattê等，1992）。

　　Among the by-products, vinasse is by far the most important. The high volume of vinasse generated by the sugarcane industry was a cause of great environmental concern during the early seventies in Brazil, when some plants dropped it in the rivers. After realizing vinasse importance in improving soil fertility it has been completely used in sugarcane fields. The average content of nutrients per cubic meter of vinasse is presented in Table 6, made by Rodella et al. (1980). It is used mostly in ratoon-cane with N supplementation, as aqua ammonia or urea. Vinasse doses vary according with its K2O concentration. For ratoon-cane it is recommended the double amount of the usual chemical fertilization which in Sao Paulo State ranges from 120 to 180 kg of K2O/ha (Demattê et al., 1992).

(表：表6 不同类型蔗汁酒糟的化学成分(Rodella等,1980) )

Nutrient 养分 Molasses must糖渣汁 Mixed must 混合汁 Sugar cane juice must甘蔗汁
N 氮(kg/m3)公斤/米3 0.57 0.48 0.28
P2O5 五氧化二磷 0.10 0.09 0.09
K2O 氧化钾 3.95 3.34 1.29
CaO 氧化钙 1.85 1.33 0.13
MgO 氧化镁 0.98 0.58 0.21
O.M. 有机质 37.30 28.97 22.31
pH 4.20 4.40 3.70

　　滤饼是蔗糖生产的副产品，有机质、钙、磷、氮、铁含量高，而钾、镁含量低，见于表7。

　　Filter cake is a residue of sugar production, with high content in organic matter, calcium, phosphorus, nitrogen and iron, although very poor in potassium and in magnesium, as can be seen in Table 7, from Coleti et al. (1986).

(表：表7 滤饼化学成分(Coleti等,1986) )

Determination 测定值 Filter cake滤饼 Determination测定值 Filter cake滤饼
Dry matter 干物质 (%) 48 Mg镁(%) 0.03
Organic matter有机质(%) 10 Fe铁(%) 3.54
N 氮(%) 0.8 Cu铜 (ppm) 15
P (total)全磷 0.8 Zn 锌(ppm) 13
K 钾(%) 0.08 Mn锰 (ppm) 117
Ca钙 (%) 3.12 C/N 碳氮比 7.5

　　建议施于蔗田，因其可部分取代氮、磷肥（表8），并且用在砂质土要好于粘质土。

　　It is recommended mostly for application in cane-plant, because it can partially substitute N and P fertilization, as observed by Coleti et al. (1986). It is observed also that better results were obtained in sandy than in clay soil.

(表：表8 滤饼在砂质和粘质土壤上的效果(Coleti et al.,1986) )

Nutrients 养分 (kg/ha)公斤/公顷 Wet filter cake 湿滤饼 (t/ha)吨/公顷 Cane 甘蔗 (t/ha)吨/公顷
N P2O5 K2O 1st cut 第1茬 2nd cut 第2茬 Ave. 平均
Yellow Red Latosol - sandy phase黄红壤-砂质
0 0 0 0 69 72 70
0 0 0 15 92 95 93
0 60 60 15 99 96 97
0 60 120 15 115 98 106
20 120 120 0 97 94 95
Roxo Latosol- dystrophic clay 红壤-贫瘠粘土
0 0 0 0 121 92 107
0 0 0 15 150 102 126
0 0 120 10 140 96 118
0 60 60 10 141 94 117
20 120 120 10 137 96 116
20 120 120 0 124 98 111

　　推荐用量为：The recommended doses are:

　　25~35吨/公顷：施于种植行内，或

　　25-35 t/ha: in the planting row or

　　60~100吨/公顷：播前撒施翻混入土，和

　　60-100 t/ha: broadcast application and incorporated into the soil before planting and

　　20~30吨/公顷：施于截根苗行间

　　20-30 t/ha: application between the rows in ratoon cane.

　　滤饼平均含1.3~1.5公斤K2O/立方米，卡车最大经济运输距离为8~10公里。这样，在有喷灌的条件下，平均用量为每公顷约300立方米。

　　For an average content of 1.3-1.5 kg of K2O/m3 , the maximum economic distance for truck distribution is in the range of 8-10 km. With sprinkler application the average dose used is around 300 m3/ha.

　　4.6.高产管理措施 Management practices for high yields

　　4.6.1为获得平均每周期5茬产量100吨/公顷，根据土壤质地推荐以下措施（Demettê,1992）：

　　To achieve an average yield of 100 t/ha in the five cuts/cycle the following practices are reconunended (Demattê, 1992) depending on the soil texture:

　　粘土 clay soil:

　　园盘耙耕（32英寸园盘耙） disking (32" disk)

　　 施用石灰和石膏 liming and gypsum application

　　 深松 subsoiling

　　园盘耙耕（26英寸园盘耙） disking (26" disk)

　　播种时施肥 planting + fertilizer

　　砂壤土（量达30%粘粒） sandy-loam soil (up to 30% clay):

　　园盘耙耕（32英寸园盘耙） disking (32" disk) liming and gypsum application

　　施用石灰和石膏 liming and gypsum application

　　耕翻 plowing

　　园盘耙耕（26英寸园盘耙） disking (26" disk)

　　播种时施肥 planting + fertilizer.

　　4.6.2.假如在pH7时测定阳离子交换量，按盐基饱和度的60%计算石灰用量。

　　Lime doses are calculated for 60% of base saturation, considering CEC measured at pH7.0.

　　4.6.3.按照阳离子交换量和盐基饱和度推荐石膏用量（表9）

　　 　　Gypsum doses are recommended according to CEC and base saturation.

(表：表9 甘蔗的石膏用量(Demattê, 1992)* )

CEC阳离子交换量(meq/100 ml)毫克当量/100毫升 v (%) Gypsum石膏 (t/ha)吨/公顷
＜ 2 ＜ 10 2.0
　 10-20 1.5
　 20-40 1.0
2-5 ＜ 10 3.0
　 10-20 2.0
　 20-40 1.5

注释：*Demattê,J.L.I.(ESALQ, Piracicaba-SP) Personal communication,1992. 私人通信
4.6.4.播种时热法磷酸盐磷肥一般推荐用量为：

　　The usual fertilizer recommendations with P as thermophosphate, at planting are:

　　低肥力土壤 Low fertility soils

　　40-150-120或 40-150-120 or

　　0-90-120 +30吨/公顷滤饼或 0-90-120 + 30 t/ha filter cake or

　　40-120-0 +酒糟 40-120-0 + vinasse

　　高肥力土壤 High fertility soils

　　 0-0-120 +30吨/公顷滤饼或 0-0-120 + 30 t/ha filter cake or

　　0-0-0 +酒糟 0-0-0 + vinasse.

　　Malavolta（1992）播种时施用以下微量元素：5公斤/公顷硼和锌，和2.5公斤/公顷铜。

　　Malavolta (1992) recommends to apply the followings micronutrients at planting time: 5 kg/ha of B and Zn and 2.5 kg/ha of Cu.

　　4.6.5.一般推荐为截根苗施用：

For ratoon-cane the most common fertilizer recommendations are: 　　100-0-100或 100-0-100 or

　　100-0-0 + 酒糟 100-0-0 + vinasse.

　　4.6.6.Malavolta提出的100吨/公顷甘蔗叶片分析参数见表10。

　　Leaf analysis parameters associated by Malavolta (1992) with 100 t/ha of cane is presented in Table 10.

(表：表10 产量100吨公顷的种植园观察到的养分浓度(Malavolta,1992) )

Nutrients 养分 Plant-cane** 新植苗 1st ratoon** 第1茬截根 2nd ratoon** 第2茬截根
N 氮（%） 1.47 1.64 1.86
P 磷 0.16 0.15 0.21
K 钾 1.49 1.08 0.87
Ca 钙 0.39 0.28 0.68
Mg 镁 0.35 0.2 0.35
S 硫 0.17 0.10 0.12
B 硼 (ppm) 20 1.10 9
Cu 铜 19 6 6
Fe 铁 271 66 62
Mn 锰 256 52 99
Zn 锌 12 5 16

注释：* Malavolta, E. (CENA/USP, Piracicaba-SP), personal communication, 1992. 私人通信 ** Samples taken 4 months after beginning of vegetation.营养生长开始4个月后取样
　　4.6.7.一周期第5次收割后建议播种绿肥菽麻(Crotalaria Juncea)。

　　After the end of the 5 cuts/cycle green manuring with Crotalaria juncea is recommended.

　　5.旱稻：Progresso农场 UPLAND RICE: THE PROGRESSO FARM CASE

　　巴西中部旱稻平均产量约1.5吨/公顷。该地区的Progresso农场的产量却为3吨/公顷。

　　The average yield of upland rice in Central Brazil is around 1.5 t/ha. In this very same region Progresso Farm is obtained more than 3,000 kg/ha.

　　Progresso农场座落于帕雷西斯高原，马托格罗索州首府库亚巴以北380公里，海拔400米，南纬13度。天然植被为低肥力黄砖红壤上生长的Derrado草（大草原）。9月~5月降雨量为2000~2600毫米。

　　Progresso Farm is located in the Parecis Plateau, 380 km north of Cuiabá, the capital of, Mato Grosso State, 400 m altitude and 13°S latitude. The natural vegetation is of cerrado (savannah) over yellow and red latosols of low fertility. Rainfall is in the range of 2,000-2,600 mm from September to May.

　　1976年开始伐林垦地。14年来在1800公顷土地上连续耕种。头4年种稻，但产量从第二年的2520公斤/公顷降为第四年的1500公斤/公顷后，就使用了石灰（3吨/公顷）和绿肥孟加拉黧豆(Stizolobium aterrimum)。其后种大豆，但产量徘徊在2200公斤/公顷，当时农场主Munefune Matsubara先生通过EMBRAPA寻求帮助研究。

　　The land-clearing was started in 1976 and now, it has 1,800 ha under continuous cultivation since the last 14 years. Rice was grown during the first 4 years, but after yield reduction of 2,520 kg/ha in the 2nd year to 1,500 kg/ha in the 4th year, the land was limed (3 t/ha) and green manured with Stizolobium aterrimum. After that soybean was planted but yield stagnated in the level of 2,200 kg/ha when the owner, Mr. Munefume Matsubara looked for the research help, through EMBRAPA.

　　1986年，为了新技术开发，该农场建立了一块50公顷的研究用地，第二年扩为100公顷。

　　In 1986' a 50 ha pilot unit for research was implemented in the farm aiming at the development of new technologies and in the next year it was increased to 100 ha.

　　研究要解决的最重要问题是：（1）积累和维持土壤肥力；（2）避免土壤硬磐；（3）控制土壤侵蚀。

　　The most important problems research had to solve were: (1). build-up and maintenance of soil fertility; (2). avoid soil hardpan and (3). control soil erosion.

　　在试验和商品生产地区获得以下结果（Bouzinac S.，1992）。

　　The following results have been obtained in the experimental and in the commercial areas (Bouzinac, S., 1992)6:

　　5.1.新垦土地 Newly cleared land

　　因有残留树木和树根，头两年中新垦土地上唯一可行的土壤耕作措施是重型园盘耙和平地园盘耙。

　　In newly cleared land due to the rest of trees and roots, the only possible soil tillage management is through heavy disks and leveling disks in the first two years.

　　0~20厘米土层平均粘粒含量为35~50%，pH 5.0，钙+镁 0.4毫克当量/l00克土，铝1.5毫克当量/100克土，0.5ppm磷和20ppm 钾。

　　The average clay content of the soils varies from 35 to 50%, with pH 5.0, 0.4 meq/100 g of Ca + Mg, 1.5 meq/100 g of Al, 0.5 ppm of P and 20 ppm of K in the 0-20 cm layer.

　　培肥土壤使用2~3吨/公顷白云质石灰石和1500公斤/公顷 YOORINBZ（含硼和锌的热法磷酸盐），园盘耙混，这种培肥措施足以满足3季作物。

　　Fertility correction is made with 2-3 t/ha of dolomitic limestone and 1,500 kg/ha of YOORINBZ (thermophosphate with boron and zinc), which are disk incorporated. This correction is enough for up to 3 crops.

　　所用水稻品种为高产品种IRAT216或Rio Verde 183，行距为18~25厘米，播种时每公顷只施用氯化钾 100公斤。

　　The rice cultivars used are the high yield varieties IRAT 216 or Rio Verde 183, the spacing of 18-25 cm between the rows and with 100 kg/ha of KCl as the only fertilizer applied at planting.

　　根据生长周期和所见发育状况推荐以下追肥：

　　According to the crop growth cycle and visual development the following top dressed fertilizers are recommended:

　　 播后25~30天追施100公斤/公顷硫酸铵和

　　 100 kg/ha of ammonium sulphate 25-30 days after planting and

　　 播后50~60天追施50公斤/公顷尿素和50公斤/公顷氯化钾。没必要用除草剂。

　　50 kg/ha of urea + 50 kg/ha of KCl 50-60 days after planting. There is no need of herbicides.

　　1990年用这种管理方法在大面积上得到以下产量：

　　With this management the following yields were obtained at commercial areas in 1990:

　 9月播种的：4550公斤/公顷

　　seeding in September: 4,550 kg/ha

　　 11月播种的；3350公斤/公顷

　　seeding in November: 3,350 kg/ha.

　　观察到一个有趣的副效应，当使用热法磷酸盐时会减轻稻梨孢(Pyricularia orizae)引起的稻瘟病。

　　One of interesting side effect observed with thermophosphate use is the reduction of rice blast infection caused by Pyricularia orizae.

　　种稻两年后，种大豆前用2吨/公顷石灰对土壤作了纠正。

　　After two years with rice, a new soil correction is made with 2 t/ha of limestone, before using the soil for soybean crop.

　　5.2.使用10年以上的土壤 Land with more than 10 years of use

　　这种土壤最重要的问题是重型园盘耙造成的30厘米深处的硬磐。打破硬磐土壤措施使得：

　　In these soils, the most important problem is a hardpan, about 30 cm in depth, caused by heavy diskings. Soil management for breaking-down the hardpan allows:

　　有机质、磷、钾、钙、镁和其它养分在更深土层重新分布；

　　redistribution of organic matter, P, K, Ca, Mg and other nutrients to deeper layers;

　　水分更好地入渗，减少侵蚀危害，改善储水能力；

　　better water infiltration redqcing erosion damages and improving water storage capacity;

　　扎根更深有助于克服干旱造成的问题。

　　deeper root system to help overcome drought problem caused by dry spells (veranicos).

　　在约1公顷的试验小区上，大豆连作水稻试验三种不同管理措施：普遍园盘耙耕、免耕、雨季开始时深翻。所得结果见表11，深耕是最好的措施。

　　In research plots of about 1 ha each three different management practices were tested for rice after soybean: usual disking, no-till and deep plowing at beginning of the rainy season. The results obtained are in Table 11, where deep plowing is by far the best management.

(表：表11十几年传统耕作后采用3种不同的管理措施地块的水稻产量(Bouzinac,1992) )

Year 年份 Management practices管理措施(kg/ha)公斤/公顷
Disking圆盘耙 No-till免耕 Deep plowing深耕
86/87 2,110 2,375 2,375
87/88 2,795 2,070 3,846
88/89 1,810 1,854 3,114
89/90 890 1,100 3,575
90/91 1,571 876 2,556
×(86/91) 1,835 1,655 3,093
Production cost生产成本(US$/ha)美元/公顷 330 345 381
Net profit纯利 (US$/ha)美元/公顷 -48 -99 73

注释：*Bouzinac, S. Persona1 communication, 1992. 私人通信
　　根据这些结果，可对商品生产区建议如下：

　　Based on these results, following recommendations were developed for the commercial area:

　　i）播种期 planting dates:

　　9月~10月：头年大豆收后，早熟品种；

　　 September/October: for early cultivars, after soybean in the previous year;

　　12月：9月播种的绿肥收后，晚熟品种。

　　December: late cultivars, after green manure sown in September;

　　ii）间距:17~25厘米 spacing: 17-25 cm;

　　iii）播种时施肥：与新垦地的相同

　　planting fertilization: the same as that in the newly cleared land;

　　iv）除草剂 herbicides:

　　0.6~0.8升/公顷FURORE控制禾本科杂草，施于水稻4~5叶期和

　　0.6-0.8 1 /ha of FURORE (Fenoxaprop 70-100 g a. i. /ha) for Graminae species control, applied at rice 4-5 leaf stage and

　　1.0升/公顷2-4D胺，控制宽叶杂草，施于一周后。

　　1.0 l /ha of 2-4 D amine for broad leaf species control, one week later.

　　v）追肥：施除草剂后立即追施第1次100公斤/公顷硫酸铵氮肥，播种后50~60天施用50公斤/公顷尿素和50公斤/公顷氯化钾。

　　top dressed fertilization: application of the 1st dose of topdressed N - 100 kg/ha of ammonium sulphate soon after the herbicides and 50-60 days after planting the additional 50 kg/ha of urea + 50 kg/ha of KCl.

　　Seguy和Bouzinac（1991）的文章中有关于Progresso农场更详细的资料。

　　Detailed information about the soil management at Progresso farm can be seen in Seguy and Bouzinac, 1991.

　　6.棉花-以Mine家庭农场为例 COTTON: THE MINE FAMILY CASE

　　Mine农场耕种1500公顷，其中360公顷租用邻居土地。1100公顷棉花是主要作物，其余种玉米或大豆。

　　The Mine family explores a total area of 1,500 ha, from which 360 ha are rented from neighbors. Cotton is the main crop with 1,100 ha, and the rest is cultivated with corn or soybean.

　　1988和1989年巴西棉花平均产量是1270公斤/公顷，其时，Mine家在850公顷土地上获得3967公斤/公顷（纤维+棉籽）。

　　The average cotton yield (fiber + seed) in Brazil for the years 1988 and 1989 was 1,270 kg/ha. Mine Farm got 3,967 kg/ha in an area of 850 ha, in that same period.

　　Mine农场的保本收成是3100公斤/公顷或1140美元/公顷。他们的经营边际利润至少是20%。

　　The break-even point at Mine Farm for cotton production is 3, 100 kg/ha or US$ 1,140. 00/ha and their management is done aiming at a profit margin of at least 20%.

　　玉米和大豆是棉花种植3年后的轮作作物。

　　Corn or soybean are planted as a rotation crop after three years of cotton.

　　最近15年对土壤重施磷肥，所用施肥措施为：

　　The soil was heavily fertilized with P during the last 15 years. The fertilization scheme used was:

　　pH7时测定阳离子交换量，按土壤盐基饱和度的65%施石灰。

　　 liming for 65% of base saturation in a CEC measured at pH 7.0;

　　播种时施用400公斤/公顷3-15-15。

　　at seeding: 400 kg/ha 03-15-15;

　　追肥：200公斤/公顷

　　top dressing: 200 kg/ha

　　20-0-20（第1次） 20-00-20 (1st)

　　80公斤/公顷尿素（第2次） 80 kg/ha urea (2nd)

　　80公斤/公顷尿素（第3次） 80 kg/ha urea (3rd).

　　所用品种为IAC-12，定苗为6~7株/米，行距1米，便于机械收获。

　　The variety used was IAC-12, with final stand of 6-7 plants/m and 1.0 m between rows, for mechanical harvest.

　　通过综合治虫措施防虫。

　　Pests are controlled through IPM - Integrated Pest Management.

　　他们对未来的打算：考虑到种棉花需高度集约，场主打算将来缩减棉花面积，转而多种大豆和玉米和风险更少的牛肉生产。

　　Their planning for the future: considering the high investments needed to grow cotton the family trends for the future is to reduce cotton area, to diversify into larger proportion with soybean and corn and to move to beef production which are less risky.

　　7.结论 CONCLUSIONS

　　7.1.世界农业面对激烈竞争环境的严酷现实，农民生存的最好战略之一是成为最高经济产量（MEY）生产者。

　　In the crude reality of the highly competitive environmental scenario facing world agriculture, one of the best strategy for farmer survival is to be a MEY producer.

　　7.2.巴西的先进农民领导了这一潮流，自己研究发展最大经济产量的生产技术，如上述甘蔗，早稻和棉花的范例。

　　In Brazil, top farmers are leading this trend, doing their own researches in order to develop technologies for the MEY production, as presented for sugarcane, rice and cotton.

　　7.3.但在一些场合，这些农民也难以生存。一些人转向多种经营，或种其它作物，或养牲畜，或两者兼营。

　　However, in some cases, even for these farmers the survival has been difficult. Some are moving to diversification, either to other crops or to animal production, or both.

　　7.4.合作化可能在促进立体农业方面是重要一步，这是农民和消费者的利益所在。

　　Cooperativism could be of key importance for the promotion of verticalization in agriculture, which would be for the best interest of farmers and consumers.

　　参考文献 REFERENCES

　　1. Alvarez, R., J.C. Ometto, A.C.P. Wutke, H.V. Arruda, e E.S. Freire. 1965. Adubacao da cana-de-acúcar. XI - Experiências com diversos fosfatos. Bragantia, Campinas, 24(9): 97-107.

　　2. Coleti, T., J.M. Lorenzetti, P. G. Freitas, J.C. Carbini e L.A. Warder. 1986. Uso de produtos de compostagem em comparacao com torta de filtro na adubacao da cana-planta e seus efeitos na soca em solo de alta e baixa fertilidade natural. STAB, Piracicaba, 4: 33-39.

　　3. Demattê, J.L.I. 1992. 0 uso agronomico de resífduos x fertilizantes na cultura de cana-de-acúcar, In: Simpósio sobre '0 solo como meio de descarte e degradacao de resìduos", Y.X Reuniao Brasileim de Fertilidade do Solo e Nutricao de Plantas. Anais. In print.

　　4. Holt, D. 1987. A competitive strategy for Americanagriculture. AAAS Symposium on Competition in Global Agriculture, Chicago.

　　5. Morelli, J.L., A.E. Dalben, J.O.C. Almeida e J.L.I. Demattê. 1992. Efeitos de c-alciao e gesso na produtividade da cana-de-acúcar e nas caracterìsticas quìmicas de um latossolo de textum média álico. Revista Brasileira de Ciência do Solo, Campinas, In print.

　　6. Morelli, J.L., E.J. Nelli, J.R. Baptistella e J.L.I. Demattê. 1991. Termofosfato na produtividade da cana-de-acúcar e nas propriedades quìmicas de um solo arenoso de baixa fertilidade. Revista Brasileira de Ciência do Solo, Campinas, 15:57-61.

　　7. Nunes Jr., D. 1992. Variedades de Cana-de-Acúcar: Perspectivas de Substituicao e Expectativas de Ganhos de Produtividade. STAB, Piracicaba, 10(3):9-13.

　　8. Orlando Filho,J., G.M.A. Silva e E.J.A. Leme. 1983. Utilizacao agrìcola dos resìduos da agroindúdstria canavieira. p. 229-264. In: J. Orlando Filho (coord.), Nutricao e adubacao da cana-de-acúcar no Brasil, Planalsucar, Piraciraba.

　　9. Rodella, A.A., C. Parazzi e A.C. Cardoso. 1980. Composiqio da vinhaga. In: Simpósio de Tecoologia do Acúcar e do álcool. STAB Sul 3, águas de Sao Pedro. Anais. p.243-56.

　　10. Séguy, L. e S. Bouzinac. 1991. Gestao e manejos dos solos e das culturas, na fronteira agrìcola dos cerrados úmidos do centro-oeste brasileiro. Destaques 1991 e sìntese atualizada 1986-1991. Convênio RPA/IRAT-CIRAD.

　　11. Wagner, R. E. and D. W. Dibb. 1990. Global MYR perspective. Proceedings of Symposium on Maximum Yield Research. 14th International Congress of Soil Science, Kyoto, Japan: 1-8.

　　12. Yamada, T. 1990. Maximum Yield Research in Brazil. Proceedings of Symposium on Maximum Yield Research. 14th International Congress of Soil Science, Kyoto, Japan: 24-31.