Agronomic-economic research on soils of the tropics

AGRONOMIC-ECONOMIC RESEARCH

ON SOILS OF THE TROPICS

1978-1979 Report

Soil Science Department
North Carolina State University
Raleigh, N.C.

under
Contract AID/ta-C-1236
with the
U.S. Agency for International Development

AGRONOMIC-ECONOMIC RESEARCH
ON SOILS OF THE TROPICS

1978-1979 REPORT

Soil Science Department
North Carolina State University
Raleigh, North Carolina 27650, U.S.A.

supported by
Contract AID/ta-C-1236
with the
U. S. Agency for International Development

November, 1980

DEDICATION

Dr. John L. Malcolm is AID Project Monitor for the
Tropical Soils Research Program and has contributed
immeasurably to the Program's success through his
positive encouragement, administrative guidance and
scientific direction since its inception. To Dr. John L.
Malcolm, for his unique and outstanding contribu-
tions to the Program, we give our heartfelt thanks and
dedicate this report.

ACKNOWLEDGEMENTS

We wish to take this opportunity to give special thanks to Mrs. Dawn Silsbee for her excellent
typing of the text of this report. Thanks are also due to Ms. Bertha Monar and Mrs. Jan Holman
for their excellent preparation of the tables and to Mrs. Ann Matrone for the excellent figures
contained herein. Appreciation is also given to Mrs. Thomasene Bennett for her fine typing of
some of the drafts, to the authors for their presentations and especially to Mr. Jot Smyth for
helping edit the drafts and final manuscripts.

John J. Nicholaides, III and Pedro A. Sanchez
Coordinators, Tropical Soils Research Program

PERSONNEL

Administration
Charles B. McCants, Department Head
John J. Nicholaides III, Program Coordinator1
Pedro A. Sanchez, Program Coordinator2
Bertha I. Monar, Adminstrative Secretary
Dawn M. Silsbee, Bilingual Secretary
Lynn Dickey, Research Technician
Mary Jo Stephenson, Research Technician
Ralph Edelberg, Research Technician

Amazon Jungle of Peru
Dale E. Bandy, Project Leader
J. Hugo Villachica, former Research Assistant
J. R. Benites, Research Assistant
Julio Alegre, Research Assistant
Miguel Ara, Research Assistant
T. Yengle, Research Technician
Stanley W. Buol, Professor
D. Keith Cassel, Professor
John J. Nicholaides III, Assistant Professor
Pedro A. Sanchez, Professor

Cerrado of Brazil
K. Dale Ritchey, Project Leader3
T. Jot Smyth, Research Assistant
Fred R. Cox, Professor
George C. Naderman, Jr., Associate Professor
Pedro A. Sanchez, Professor

Extrapolation
Dale E. Bandy, Project Co-Leader
Ruben Mesia, Project Co-Leader
Stanley W. Buol, Professor
Fred R. Cox, Professor
Robert E. McCollum, Associate Professor
Gordon S. Miner, Associate Professor
John J. Nicholaides III, Assistant Professor

Intercropping
Robert E. McCollum, Project Leader
Clifton K. Hiebsch, Research Assistant

Economic Interpretation
Arthur J. Coutu, Project Leader
Diane Hernandez, Economist

1From 1977-present.
2From 1971-1976, 1979-present.
3Cornell University staff member employed on
cooperative project.

TABLE OF CONTENTS

Page

1 INTRODUCTION
1.1 H highlights ....... .. ... ...... ... .... .. .. .. .. .. ... ... .. .. .. .. .... .. 3
1.2 Collaborating Institutions and Individuals ................................ 9
2 CERRADO OF BRAZIL (JOINT EMBRAPA-CPAC/CORNELL/NCSU RESEARCH)
2.1 C rop W weather ..................................................... 13
2.2 Residual Effects of Lime Rates and Depth of Incorporation. ................. 17
2.3 M movement of Ca and M g ............................................. 29
2.4 K and M g Feftilization .............................................. 33
2.5 K Movement in a Clayey Red-Yellow Latosol ............................. 37
2.6 Sources, Rates and Placement of P Fertilizers on a Clayey-Red
Yellow Latosol ................................................ 39
2.7 Residual Effects of P Rate, Placement and Time of Application ............ .. 49
2.8 Effects of P Sources on Pastures ....................................... 51
2.9 N Fertilization..................................................... 61
3 AMAZON JUNGLE OF PERU (JOINT MAA-INIA/NCSU RESEARCH)
3.1 C rop W weather ..................................................... 69
3.2 Annual Crop Varietal Adaptation Experiments. ........................... 78
3.3 Forage Species Adaptation Experiment .................................. 103
3.4 Management of Annual Crops Experiment ............................... 105
3.5 Deep Lime Experiment ............................................ 127
3.6 K and Mg Fertilization ....................... .......... ............... 137
3.7 Multiple Cropping-N Experiment.................... ................... 141
3.8 Effect of Clearing and Continuous Cultivation on Soil
Physical Properties ................................................ 159
3.9 Minimum Input Experiment .......................................... 163
3.10 Lim e Experim ent .................................................. 181
4 EXTRAPOLATION ACTIVITIES
4.1 Yurimaguas Small Farmer Extrapolation Program. ......................... 197
4.2 Pasture Fertilization in Pucallpa ....................................... 218
4.3 Extrapolation to Bolivian Savannas..................................... 225
5 INTERCROPPING
5.1 Effects of Removing Corn from a Corn-Soybean Intercrop Upon Soybean
Yields and Yield Equivalency Ratios ................ ................ 244
6 ECONOMIC INTERPRETATION
6.1 Economic Interpretation of Agronomic Data ............................. 267
7 COMMUNICATION OF RESULTS
7.1 Publications....................................................... 277
7.2 M ailing List ....................................................... 28 1

INTRODUCTION

Sr. German Gonzalez, a Peruvian small farmer, explains to Dr. Carlos Valverde,
Deputy Director of INIA, and Ing. Ruben Mesia, Peruvian head of the Yuri-
maguas Station and Extrapolation Program, that the small farmers of the area
are accepting the continuous cropping technology developed by the coopera-
tive MAA-INIA/NCSU research. San Juan, Loreto, Peru. December, 1979.

3

This is the seventh formal report of North
Carolina State University's Tropical Soils Re-
search Program, supported by the U. S. Agency
for International Development under Contract
AID/ta-C-1236. This report covers the period
from late 1977 through 1979.
The overall objectives of the Tropical
Soils Research Program are to: 1) Develop
economically-sound soil-crop management sys-
tems for tropical rainforests and acid savannas,
2) validate these systems on small farms pre-
sently under shifting cultivation, and 3) refine
means to extrapolate results to other areas of
the world with similar agronomic and socio-
economic conditions. Presented herein is evi-
dence of an exciting and successful Program
which is meeting these objectives.

HIGHLIGHTS
The most exciting highlight centers on
the development of agronomically-sound and
economically-viable continuous crop production
systems at the Yurimaguas Experiment Station
and their extension to and acceptance by small
farmers of the surrounding area. The improved
technologies developed by this Program are
energy- and scale-neutral. They are benefiting
both labor-intensive and capital-intensive far-
mers. The Program's research by both farmers
and national leaders of Peru.
The benefits of this Program, however,
clearly transcend political boundaries. Nowhere
is this fact more evident that in the successful
application of the results by national institutions
in other parts of Peru, in Brazil and in the
savannas of Bolivia. The Program, at the requests

of the local USAID Missions, has provided sup-
port to related agricultural development projects
in Guatemala, Sao Tome and Guinea-Bissau.
Worldwide, these results are being used to deter-
mine overall research priorities for agricultural
research and development strategies for the re-
mainder of this century by the international
donor community.
We invite the reader to come with us
through the following pages detailing the
Tropical Soils Research Program's results which
are beginning to have a positive impact on the
world in which we live.
Acid Savannas
Agronomic-economic research continued at
EMBRAPA's Cerrado Research Center (CPAC)
near Brasilia. Residual effects of lime, P sources,
K and Mg were investigated, as was movement
of Ca, Mg and K into Oxisols representative of
vast areas of tropical savannas.
Residual effect of lime was evident six years
after application to the Typic Haplustox at
CPAC near Brasilia. Yields of corn and soybeans,
the sixth and seventh crops following applica-
tion of 2 t lime/ha and incorporation to 30 cm
in 1972, were double those without lime appli-
cation. Corn yields increased approximately
200 kg/ha with each 10% decrease in Al satura-
tion. Soybean yields increased markedly when
Al saturation was decreased below 40%.
Movement of Ca into the subsoil of Typic
Haplustox is an important finding as liming,
with proper source, can be accomplished over
time without the special equipment and addi-
tional energy required for deep incorporation of
lime. Calcium movement to at least a 75 cm

4

depth was greater when source was CaCI2, which
was greater than CaSO4, which was greater than
CaCO3. This movement was, as expected, in pro-
portion to the solubilities of these compounds.
Movement of CaSO4 from 228 kg/ha of simple
superphosphate surface applied and incor-
porated raised pH in the 30-45 cm depth by 0.5
units.
Residual K was beneficial to the second and
third crops. Value of extra corn grain to 62 kg
K/ha applied a year earlier was nine times the
cost of the K fertilizer. Corn responded to re-
sidual K up to 500 kg/ha. The third crop, soy-
beans, responded to 140 kg residual K/ha. Soy-
bean yields were 37% greater with 97 kg residual
Mg/ha than with 7 kg residual Mg/ha. Soil test
critical level by double acid extractant was 0.10
meq/100 cc for both corn and soybeans. Move-
ment of K was found to the 60-75 cm depth in
the Typic Haplustox.
Potassium movement was also noted in the
Typic Acrustox which had 25% more clay but
half the effective CEC of the Typic Haplustox.
Potassium was detected in the 75-90 cm depth
in a P deficient treatment. In contrast, higher K
uptake in the high P treatment probably reduced
the amount of K leached below the 60-75 cm
depth.
Direct application of rock phosphate under
certain conditions can be a cost-effective alterna-
tive to large initial P fertilizer amendments on
high P-fixing soils of the Brazilian Cerrado. A
low reactivity Brazilian rock phosphate, Patos de
Minas, at 352 kg P/ha plus 44 kg P/ha as simple
superphosphate (SSP) produced an identical
second soybean crop yields to the 352 kg P/ha

broadcast SSP treatment at 50% less cost on the
Typic Acrustox. Residual evaluation of this
study is continuing.
Evaluation of residual effects of P rate,
placement and time of application to the Typic
Haplustox revealed that the 560 kg P/ha as SSP
broadcast applied in 1972 produced corn yields
for 8 years superior to other rates, placements
and combinations. Soil test P values also re-
flected this difference. The economic optimum
was greater than 600 kg P/ha over the eight-year
period. The optimum P rate based on the first
two crops was 220 kg/ha; if a farmer had applied
that amount, total production over the eight
crop period would have been 18 t/ha less than at
the 600 kg P/ha rate. This fact stresses the neces-
sity of long-term agronomic studies.
The tolerance of Brachiaria decumbens to
soil acidity makes feasible its production with
rock phosphates solubilized by the soil acidity.
Araxa rock is 40% the cost of simple superphos-
phate in Brazil, yet relative yields from Araxa vs.
SSP for the cumulative production of 10 cuts
over four years was 85% at 600 kg P/ha on the
Typic Haplustox.
The fifth and sixth consecutive corn crops
on the Typic Haplustox responded up to 140 kg
N/ha, with respective yields of 5.76 and 5.21
t/ha, 1.85 and 3.45 t/ha for the respective fifth
and sixth crops. Corn was planted prior to the
rainy season in the sixth year to maximize
utilization of the "N flush" released by wetting
the soil after the dry season. Corn planted on a
newly cleared Typic Acrustox responded up to
150 kg N/ha with a 2.5 t/ha yield. At 0 kg N/ha,
only 0.33 t/ha corn was realized. Acidity, P

5

deficiency, a 40-day drought and some microbial
immobilization of the fertilizer N limited corn
yields on the Typic Acrustox.
Tropical Rainforests
In addition to the continuous cropping ex-
periments which have shown it possible to pro-
duce 17 consecutive crops on the same plot of
land since 1972, new agronomic research was
begun at the Yurimaguas Experiment Station.
This included experiments of varietal adaptation
with important annual crops and forage species,
management of annual crops, deep incorpora-
tion of lime, K and Mg fertilization, N manage-
ment in multiple crops, clearing and continuous
cultivation effect on soil physical properties,
minimum inputs and residual lime management.
All studies were conducted on the Ultisol which
is representative of a large portion of the
Amazon basin.
The minimum input study revealed that a
14-month kudzu fallow can regenerate a soil
equal to or better than a 15-20 year forest
fallow as far as the first crop is concerned.
However, 40 kg N/ha were required even for the
first crop after fallow as neither the forest or
kudzu ash contained enough N for good corn
growth and production. Also, fertilizer amend-
ments are required for all subsequent crops, even
after kudzu fallow. Tillage is important as incor-
poration of ash and fertilizers significantly in-
creased crop yields.
Residual lime management studies showed
that application of 4 or 8 t lime/ha caused Mn
deficiencies and that no more than 2 t lime/ha
should be used for initial lime applications.
At 4 t lime/ha, however, maximum respective

yields of corn, peanuts and rice were realized
at 12, 17 and 21 months after application. Thus,
an obvious rotation beginning 12 months after
liming with 4 t/ha would be corn-peanuts-rice.
Soil physical properties were affected by
clearing and continuous cultivation. Lowest bulk
densities were found in the virgin forest soil
compared with those of soils continuously
cropped for 1/2, 1, 4, 5 and 6 years and 15-year
forest fallow. Greatest mechanical impedance
was noted in soil bulldozed six years earlier.
Mean infiltration rate was greatest in the virgin
forest soil. Soil in the pathway of the forest was
lightly compacted.
The multiple cropping-N study was revealed
that corn grown with peanuts yielded 850 kg/ha
more than corn grown with rice, suggesting that
some N fixed by peanuts became available for
corn. Rice intercropped with corn yielded 880
kg/ha more than rice intercropped with cassava,
probably due to the lack of cassava canopy
shading effect. Peanuts intercropped with corn
outyielded those intercropped with cassava by
130 kg/ha, probably also due to lack of cassava
shading. All monoculture crops outyielded those
same crops when intercropped. Corn response
was to 160 kg N/ha. Higher N levels in rice were
associated with increased blast incidence. Advan-
tage of multi-crops over monoculture by LER
was up to 116%, while by ATER highest effi-
ciency was only 6%.
The K-Mg fertilization study indicated that
without continual monitoring of soil chemical
properties, Mg readily can become limiting,
especially when K is applied prior to each
planting. Tentative Mg critical levels were 0.20

6

and 0.38 meq/100 cc for soybeans and corn,
respectively. Tentative critical Mg/K ratio for
both crops was 1.2. In a newly cleared area, soy-
beans responded only to 25 kg K/ha and 12 kg
Mg/ha. Lack of additional response was due to
high initial K and Mg contents in the recently
cleared soil.
Deep incorporation of lime compared with
shallow had no beneficial effect on corn yields
during the first rainy trimester increased corn
yields by 36% during the following drier season
and produced greater corn yields in the follow-
ing rainier season. Subsoil acidity and percent
Al saturation were reduced and root penetra-
tion into the subsoil was increased by deep
placement of 4 t lime/ha. Thus, the deeper the
lime placement, the less chance for reduced corn
yields during short droughts during any season
in Yurimaguas.
Evaluation of adaptation of corn, soybean,
rice, sugarcane, cowpea and mungbean varieties
to the udic soil moisture regime of the Amazon
jungle area revealed some potentially better
varieties than those currently grown. PMC 747,
the National Corn Program's designated hybrid
for the Peruvian jungle yielded well, had good
photosynthate distribution and excellent ear
filling qualities, but had many barren stalks and
were susceptible to lodging. Entries from the
line "Amarillo Planta Baja" seem to have the
greatest potential for the area.
Five soybean cultivars outyielded Jupiter,
the local check, by more than 40%. Improved
Pelican was the only variety to meet all four
selection criteria of yield, nodule abundance and

activity, and seed viability. At high inputs
Hardee and Davis were superior, while at low
inputs Tunia and Improved Pelican were the
best.
Nine rice introductions from IITA were
resistant to blast, the primary limiting factor to
rice production in Yurimaguas. These varieties
have other advantages over the local tall
Carolino variety (low yield potential and lodging
susceptibility) and the introduced, semi-dwarf
IR4-2 (low consumer acceptance due to small
grain size and high amylose content). The eco-
type Tox 340-1-1-1-1 was perhaps the most
promising.
Best sugarcane and juice qualities were
found in cultivar NCO-310 as is had low fiber
content, excellent Pol, low percentage reduc-
tors, very good Brix, optimum purity and low
moisture at harvest. These findings should prove
beneficial as most sugarcane is better adapted to
the ustic soil moisture regimes. The predominate
local cultivar, P05-2878, was only 16th best of
19 cultivars, in terms of cane and juice quality.
Some determinate and indeterminate cow-
pea lines from IITA outyielded the local varie-
ties, with four indeterminate entries producing
50% more yield than the local. Preliminary
results with mungbeans from AVRDC indicated
that at least one cultivar, Pangasa, has good
potential for the area.
First year evaluation of 11 legume and 3
grass accessions for local pastures revealed
Pueraria phaseoloides 9900, Desmodium ovali-
folium 350 and Stylosanthes guianensis 136 to
be the most promising legumes and Andropogon

7

gayanus 621 among the most promising grasses
for the area. The A. gayanus 621 produced
56 t/ha of dry matter during the first year. This
initial test will be supplemented by grazing pres-
sure trials in 1980.
The management of annual crops study
showed corn production to be best during the
periods with lower rainfall and higher solar
radiation (March 15-September 15). Peanut
yields were less influenced by planting date than
either corn or soybeans. Soybeans planted in
July, September or October yielded nearly
3.5 t/ha. Based on this and previous studies,
rotations of rice-corn-soybeans, rice-peanuts-
soybeans, and rice-cowpeas-corn, with rice plant-
ing from September 15-January 15 are now
recommended for the area. This information
also shows that relatively small but consistent
changes in rainfall pattern make major differen-
ces in determining the most appropriate crop-
ping sequence in humid tropical conditions.
Extrapolation of Results to Other Regions
The extrapolation component has developed
from a dream of the early years of the Program
to one of its most important activities. Contin-
uous farming systems developed at the Yuri-
maguas Experiment Station have been trans-
ferred to and accepted by farmers of the
surrounding area. Also, national research pro-
grams-one on pasture in Pucallpa, Peru, the
other in peanut and soybean production in the
Bolivia savannas-have benefited by the Extra-
polation program.
The Yurimaguas small farmer extrapolation
program placed research demonstration trials on
the land of small farmer leaders in communities

surrounding Yurimaguas. The farmers compared
their own traditional system (I) with imrpoved
agronomic practices without fertilizers and lime
(II) and with improved agronomic practices with
lime and fertilizer (III). Results showed a) that
small farmers of the region can continuously
crop land where only to shifting cultivation was
possible before, b) that greater than 10 t
grain/ha/yr can be realized by these farmers
using the improved technologies, c) that this im-
proved production is economically feasible, and
d) that farmers are accepting the improved
agronomic practices. Thus, the first year's results
have shown that seven years of research on the
Yurimaguas Station has produced a continuous
cropping package that is not only agronomically
and economically superior to the traditional
system, but also is transferable and acceptable to
the small farms of the surrounding area.
Extrapolation work with IVITA in Pucallpa
found that Brachiaria decumbens produced 35
t/ha with 400 kg N/ha, with preliminary
optimum economic N fertilization rates for in-
tensive milk production being 250 kg N/ha in
the rainy season and 72 kg N/ha in the dry
season. B. decumbens gave no response to liming
at either low or high fertilization levels. It was
concluded that B. decumbens has a tolerance to
at least 50% Al saturation.
Extrapolation work in the Bolivian savannas
near San Ignacio de Velasco used experience
with somewhat similar soil-crop-climatic condi-
tions in Peru and Brazil and soil testing proced-
ures to determine needed fertilizer rates. Phos-
phorus was correctly predicted to be the repre-
sentative of much of the surrounding area. Also,

8

correctly predicted was 13 kg P/ha to reach
maximum peanut yields (3.25 t/ha). For
maximum soybean yields (1.73 t/ha), 26 kg
P/ha were required. Usual second crop peanut
yields in the area are less than 1 t/ha. With only
improved management, yields were increased to
2.5 t/ha and with only 13 kg P/ha yields were
increased further to 3.25 t/ha. Soil test critical
levels by the modified Olsen technique were
9 and 10 ug P/cc for peanuts and soybeans,
respectively.
Thus, it was possible to build past expe-
rience, soil characterization and soil testing to
create adaptive research trials to find rapid
answers to crop production problems in new
lands of Bolivia. It is felt that similar approaches
can be used successfully in other areas of the
world and the Program welcomes the oppor-
tunity to cooperate in other areas, as we also
learn in the process.
Intercropping
The postulation that intercrop advantage of
corn and soybeans could be augmented by har-
vesting the corn plants from the mixture before
critical reproductive stages of the soybeans was
tested on three Ultisols in eastern North
Carolina. Under drought conditions at one loca-
tion, LER's indicated that the intercrop utilized
land area 60% more efficiently than monocul-
tures and ATER's revealed a 30% area-time
advantage of the intercrop. In two locations, de-
layed removal of corn from the intercrop
resulted in a) soybean yields decreasing from
73% of their monoculture check to 45%, b) LER
decreased from 1.46 to 1.26 and c) ATER
averaged 1.09 for the three removal dates (soy-

bean flowering, pod formation, bean formation)
as compared with 1.01 when corn was not
removed. Increasing intercrop corn populations
increased corn yields and LER, decreased soy-
bean yields, and had no effect on ATER. Gain in
weight by interplanted soybeans was linearly re-
lated to percent of incoming photosynthetically
active radiation intercepted.
Economic Interpretation
The economic feasibility of the improved
agronomic systems when used on small farms, as
discussed in the extrapolation section, was very
positive. Economic interpretation of agronomic
data from the Yurimaguas Experiment Station
revealed that high marginal rates of return
(between 60 and 120%) are consistently yielded
by soybean, peanut and rice cultivation, while
lower marginal rates of return on corn and cow-
peas suggested a subsistence rather than market
orientation. High marginal rates of return were
found for up to 2 t lime/ha for most crops.
Economic evaluation of intercrops vs. monocul-
ture, composts vs. inorganic fertilizers, and fer-
tilization rates is continuing. The improved con-
tinuous cropping technologies developed on the
Station and transferred to the small farmers of
the area have been demonstrated to be very
economically feasible and acceptable to the
farmers.

9

COLLABORATING INSTITUTIONS AND
INDIVIDUALS
The reported research has been conducted in
close cooperation with several national and
international institutions and involves a high
degree of collaboration.
In the Cerrado of Brazil, research is con-
ducted jointly with Cornell University and the
Empresa Brasileira de Pesquisa Agropecuaria
(EMBRAPA) at the Centro de Pesquisa Agro-
pecuaria dos Cerrados, located about 40 km
north of Brasilia. The USAID Affairs Office in
Brasilia and the Interamerican Institute of Agri-
cultural Sciences provided valuable logistical
support. EMBRAPA has assigned Mr. Edson
Lobato as project leader to represent Brazil.
Cornell and N. C. State staff stationed at the
Cerrado Center form an integral part of the
Center's research staff.
In the Amazon Jungle of Peru, field research
is conducted at the Yurimaguas Experiment
Station which is part of the Instituto Nacional
de Investigaciones Agrarias (INIA). Supporting
laboratory work is conducted at the La Molina
Experiment Station and the National Agrarian
University. INIA has assigned Dr. Carlos
Valverde as project leader, to represent Peru.
Dr. Valverde has been very effective in expedit-
ing administrative matters with the Peruvian
Government as well as providing scientific
leadership. The International Potato Center
(CIP) plays a major role in providing administra-
tive and logistical support. In turn, the program
grows its potato trials at Yurimaguas as the low-
land tropical station for adapting potatoes to the
region. The Instituto Veterinario de Investi-

gaci6n del Tr6pico y de Altura (IVITA) in
Pucallpa has been a major collaborator in soils
pasture research in the Peruvian jungle areas
under the direction of Dr. Jos6 Toledo.
In the Bolivian savannas, the cooperative
research is conducted in eastern Bolivia with the
Centro de Investigaci6n Agricola Tropical
(CIAT) and the Corporaci6n Regional de
Desarrollo Santa Cruz (now CORDECRUZ).
Supporting laboratory work is with CIAT's lab-
oratory in Santa Cruz. Ing. Eduardo Hinojosa,
M.S., was named technical collaborator by our
Bolivian counterparts.
The following individuals from the different
cooperating institutions provided substantial
administrative support or are coauthors of some
of the research projects. We gratefully ack-
nowledge their assistance.
BRAZIL
Jose Ireneu Cabral, President of EMBRAPA
Almiro Blumenschein, Director of EMBRAPA
Elmar Wagner, Chief, CPAC, EMBRAPA
Wenceslau G. Goedert, Technical Chief, CPAC
Delmar Marchetti, Administrative Chief, CPAC
Edson Lobato, EMBRAPA Project Coordinator
Gilberto Paez, Head of the Data Processing
Department, EMBRAPA.
Walter Couto, CIAT-CPAC, EMBRAPA
Waldo Espinosa, Soil-Plant-Water Relations Spe-
cialist, CPAC, EMBRAPA
Leo Nobre de Miranda, Soil Fertility Specialist,
CPAC, EMBRAPA
Djalma de Souza, Soil Chemist, CPAC,
EMBRAPA
Claudio Samzonowicz, Soil Fertility Specialist,
CPAC, EMBRAPA

10

Frank Campbell, Chief, USAID Affairs Office-
Brazil.
Alfredo Lopes, Professor of Soil Science, Univer-
sity of Lavras.
David R. Bouldin, Professor of Soil Science,
Cornell University.
Douglas J. Lathwell, Professor of Agronomy,
Cornell University.
PERU
Javier Gazzo F., Director Ejecutivo, INIA
Carlos Valverde S., Project Coordinator for INIA
and Director Ejecutivo Adjunto, INIA
Manuel Llaveria, Director del Centro Regional
de Investigaciones Agrarias IlI-Tarapoto.
Ruben Mesia P., Head, Yurimaguas Experiment
Station.
Hugo Villachica, Professor of Soil Science,
National Agrarian University
Ricardo Sevilla, Corn Program, National Agrar-
ian University.
Richard L. Sawyer, Director General, Interna-
tional Potato Center.
Roger Rowe, Deputy Director General, Inter-
national Potato Center.
Carlos Bohl P., Executive Director, Interna-
tional Potato Center.
William Hamann, Assistant Executive Director,
International Potato Center.
Oscar Gil, Controller, International Potato
Center.
Veronica de Franciosi, Assistant to Executive
Officer, International Potato Center.
Leonard Yaeger, Director, USAID/Peru.
John O'Donnell, Multisector Loans Officer,
USAID/Peru.
Loren Schulze, Acting Food and Agricultural
Officer, USAID/Peru.

BOLIVIA
Carlos Vaca Diez, Director Ejecutivo, Centro
de Investigaci6n Agrfcola Tropical
Emma Viruez, Directora del Laboratorio, Centro
de Investigaci6n Agricola Tropical.
Fernando Prado, Gerente General,
CORDECRUZ
Jorge Gomez, Gerente, Unidad Proyectos Rurales
y Agropecuarios, CORDECRUZ
Jaime Aguilera, Jefe, Departamento de Des-
arrollo Agropecuario, CORDECRUZ
Oscar Moreno, Jefe, Proyecto de San Ignacio,
CORDECRUZ
Eduardo Hinojosa, Agronomo, CORDECRUZ
Luis Aguirre, Jefe, Estaci6n Experimental de
San Ignacio.
Abe Pena, Director, USAID/Bolivia
Daniel Chaij, Rural Development Officer,
USAID/Bolivia.
Richard Peters, Agricultural Adviser, USAID/
Bolivia.
David James, Chief of Party, Consortium for
International Development.

11

CERRADO OF BRAZIL

Dr. Ady da Silva (EMBRAPA/CPAC), Dr. Dave Bouldin (Cornell), Dr. and Mrs. Dean Peterson
USAID/Washington) and Dr. Edson Lobato (EMBRAPA/CPAC) discuss the cooperative EMBRAPA-
CPAC/Cornell/NCSU research at CPAC, Brasilia, Brazil, 1978.

13

This section of the biennial report covers
agronomic research conducted cooperatively
by scientists from Brazil's Centro de Pesquisa
Agropecuaria dos Cerrados (CPAC), Cornell
and North Carolina State Universities from
1976 through 1978. Formal research involve-
ment of NCSU's Tropical Soil Research Pro-
gram in Brazil terminated with that portion of
the USAID contract December 31, 1978.
Informal research involvement continues via
communication with the Brazilian scientists
who trained and/or worked with the Program
during 1972-1978.
Studies on residual effects of lime and
depth of incorporation, Ca and Mg movement,
K and Mg fertilization, K movement, P sources
and rates, residual effects of P, effect of P
sources on pastures and N fertilization are
reported herein.
All field experiments, unless otherwise
specified, were conducted on a clayey Dark
Red Latosol (Typic Haplustox, fine, kaolinitic,
isothermic) located on the second erosion
surface at CPAC near Brasilia. Soil properties
are as described in previous annual reports and
are typical of acid Oxisols of tropical savannas.
In general, these soils are well-drained, rela-
tively deep, acid, low in organic matter, P, K,
Ca, Mg and have a relatively high P retention
capacity. These Oxisols are not as susceptible
to soil compaction as are the sandier Ultisols of
the Amazon jungle.

2.1 CROP WEATHER
W. Espinosa, Denilo C. Gomes and L. A.
Rocha Batista
Pertinent climatological data for the two
agricultural years 1976-1978 are presented in
Table 2.1:1. The rainfall data for these years as
compared with the 35-year average and the
solar radiation data are presented in Fig. 2.1:1.
Rainfall
Total rainfall for the agricultural year
1976-1977 was 1656 mm, somewhat higher
than the average for the previous three years
(1405 mm). A 40-day drought began February
5, and ended March 17, 1976. Because many of
the crops at CPAC reached flowering stage
during the drought, serious reductions in yield
occurred.
Total rainfall for 1977-1978 was 1474 mm,
following the usual pattern for the area. A
17-day drought began January 19, and ended
February 4.
Intensive rains often occur at the beginning
and end of the rainy season, complicating
efforts of soil conservation due to the lack of
soil cover on cultivated cropland at that time.
The largest rainfall per 24-hour period was
96.4 mm occurring January 17, 1977 with an
intensity of 78.4 mm/hour.
Water Balance
The water balance data for 1976-1977
shown in Fig. 2.1:2 indicate the extent of the
40-day drought occurring in February and

14

Table 2.1:1.

Meteorological data for the agricultural years 1976-77 and 1977-78.
EMBRAPA/CPAC. (150 36'S, 470 42'W, altitude 1010 meters).

Temperature (oC) prav e
T rate () Precip- Evapor- e at ve Solar
Month Max. Min. Avg. itation ation Wind Humidity Radiation

1976
July
August
September
October
November
December
1977
January
February
March
April
May
June
July
August
September
October
November
December
1978
January
February
March
April
May
June

26.4
28.8
28.0
26.3
25.6
26.5

26.1
23.3
29.9
27.2
25.6
26.4
27.1
27.2
29.6
29.2
28.5
27.8

27.4
25.8
27.0
26.4
26.7
24.2

Source: Relatorio Tecnico Anual, CPAC, Brasilia,D. F.

12.7
15.3
17.1
17.2
18.1
18.2

18.0
18.2
18.3
18.4
15.6
15.8
14.4
16.4
17.8
18.0
18.5
18.5

18.3
19.5
17.8
17.7
15.1
14.0

19.5
21.0
22.9
22.4
22.3
23.4

21.5
21.8
22.5
21.9
20.6
20.4
20.7
21.9
23.7
23.6
23.5
23.2

23.6
22.4
21.9
22.0
20.9
19.1

(mm)
12.1
3.6
140.7
160.5
321.7
243.4

388.8
50.2
108.3
154.6
53.6
18.8
0.0
6.6
9.6
98.9
170.4
226.6

287.1
231.3
257.7
133.1
53.2
0.0

(mm)
182.5
224.1
147.1
147.3
106.1
120.6

131.4
145.2
178.6
136.7
127.4
112.7
143.6
200.1
173.9
169.9
148.6
123.5

138.3
142.9
143.1
118.3
107.8
157.0

(m/s)
0.99
1.10
1.18
1.01
1.00
0.69

1.31
1.02
0.72
0.87
0.73
0.66
0.94
0.64
1.00
1.11
0.75
0.58

0.65
0.71
0.86
0.77
0.72
0.67

(%)
51
47
61
65
74
67

70
62
52
61
55
54
42
38
46
32
67
64

68
68
70
68
58
55

(cal/cm /day)
432.5
462.6
348.9
379.6
358.5
376.7

371.4
406.4
337.8
327.9
378.0
308.0
406.8
374.2
361.8
385.2
381.5
351.0

379.7
342.8
386.2
313.7
340.1
354.4

15

J A S O N D J F M A M J
MONTH

Figure 2.1:1.

Rainfall and solar radiation patterns at CPAC for
the two cropping seasons, 1976-1977 and 1977-1978.

E
E

I
I-
z
0

w
a-
Qr.

I
z

0

cuJ
E
UO
0
U

I
I-
z
0
YE
W
a-

z
0
HJ

-J
o0
C,)
0

16

J A S O N D J F M A M J
1976 1977
MONTH AND YEAR

Figure 2.1:2.

Meteorological water balance
on Class A pan evaporation.
Tecnico Anual, CPAC, 1978.

for 1976-77 based
Source: Relatorio

0
0

W

Q-
W

I-
QW
W

E
E

17

March 1977. Fig. 2.1:3 shows the severity of
the 17-day drought which began January 19,
1978 and substantially reduced grain produc-
tion for maize plants flowering during the
period. Low relative humidity and high solar
radiation increased the seriousness of the
drought. The two short deficit periods shown
in Fig. 2.1:3 (November and December) were
particularly critical for soybeans planted at
CPAC because there was not a continuous
supply of moisture to allow the germinating
plants to grow, and replanting was necessary.
Temperature
The highest temperature observed during
the agricultural year 1976-1977 was 32.20C on
March 12 and the lowest was 5.40C on May 18.
During 1977-1978 the highest temperature,
32.50C, was observed September 18 and the
lowest temperature, 10.40C, occurred June 15.
Temperature of Grass-Covered Soil
As shown in Fig. 2.1:4, the soil tempera-
ture in general was lowest in July and highest
in March. At a depth of 5 cm, the difference
between the coldest and hottest month was
70C. At 50 cm, the difference was only 3.40C.
The mean monthly soil temperatures ranged
from a low of 170C to a high of 22.40C, indi-
cating a fairly constant temperature environ-
ment that should not impede biological
activity.
Relative Humidity
The highest monthly mean relative
humidity for 1976-1977 occurred in

November. During the 1976-1977 growing
season the monthly mean relative humidity
varied between 65% and 74% and during 1977-
1978 between 64% and 70%. The average rela-
tive humidity for the 1977 dry season (April to
October) was 49%, and for the 1977-1978
rainy season (November to March) 67%.
The effect of the low relative humidity is
seen in the high Class A pan evaporation rate
(Fig. 2.1:2). In August 1977, the evaporation
sometimes exceeded 10 mm/day. These same
conditions also occur during dry periods in the
rainy season.
The low relative humidity has the positive
aspect of reducing the incidence of fungal
diseases, and therefore, widening the range of
plant species that can be grown in the Cerrado.
Solar Radiation
Monthly solar radiation does not vary
much throughout the year because the higher
radiation expected during the dry season is
counterbalanced by the increased cloud cover.
The 1976-1977 maximum monthly total radia-
tion was observed in August and the minimum
in June.

2.2 RESIDUAL EFFECTS OF LIME RATES
AND DEPTH OF INCORPORATION
Leo Nobre de Miranda
The long-term experiment comparing
depths and rates of lime application was
continued for the sixth and seventh crops.
Cargill-111 hybrid maize was planted on

18

O R I I I I I I I 'l I I I
0 30 60 90 120
I Nov Dec Jan Feb Mar
1977

MONTH

Figure 2.1:3.

150 Days after
Apr planting
1978

AND YEAR

Water balance for a corn crop based on potential
evapotranspiration, rainfall, plant growth stage
and depth of rooting. 1977-1978 growing
season. CPAC.

0100
10
0

E 90
E

2 80
0

| 70

z 60
0

0 50

W
0 40
0
a:
0 30
Q
W
20

I-10
L/ 10
u

25r

20

w.%. m . . .

J ASONDJFMAMJ
1976 1977
5cm

0
o

w

Q-
1-

w
91
0-

w

-J
O3

20

I

a at t is a a II m

JASONDJ FMAMJ
1976 1977
20 cm

25

20

15

25

20

15

25-

SI 11 I I I

JASONDJ FMAMJ
1976 1977
10 cm

a a1111113111

JASONDJFMAMJ
1976 1977
25cm

20-

15 1 1 1
JASONDJFMAMJ
1976 1977
15 cm

25r

20

i .

allah Ill III

JASONDJFMAMJ
1976 1977
50cm

SOIL DEPTHS (cm), UNDER SOD

Figure 2.1:4. Mean monthly soil temperature at six depths under sod, for 1976-77.
Source: Relatorio Tkcnico Anual, CPAC, 1978.

25r

(o

-----------

--

4
I

SI I I I I I I I I I I

I

I v

I

20

LIME APPLIED IN 1972 (t/ha)

Figure 2.2:1.

Yield of
function
in 1972.

Cargill-Ill maize grain (1976-77) as a
of lime rates and depths of incorporation
Dark Red Latosol. CPAC.

UJ
w
5:
Z

w
N

21

November 6, 1976 and harvested on April 18,
1977. UFV-1 soybeans were planted November
17, 1977 and harvested April 24, 1978. The
maize crop received broadcast applications of
83 kg K/ha as KCI, 35 kg P/ha as triple super-
phosphate, and 30 kg Mg/ha as MgSO4.
Nitrogen was applied as urea at a rate of 20 kg
N/ha in the furrow at planting and 120 kg N/ha
sidedressed December 13. The soybean crop
received broadcast applications of 26 kg P/ha
as triple superphosphate and 25 kg K/ha as
KCI. Because of poor nodulation, ammonium
sulfate was sidedressed to supply 50 kg N/ha
on January 3, 1978 and 60 kg N/ha on January
19, 1978.
1976-1977 Maize
The 1976-1977 maize grain production is
shown in Fig. 2.2:1. In spite of a 41-day
drought beginning in February, there was a
significant residual effect of the lime applied in
1972. Where 8 t lime/ha had been incorporated
deeply, the yield was 2.6 times that which re-
ceived no lime. The application of 2 t lime/ha
to 30 cm in 1972 approximately doubled pro-
duction. The relationship between grain pro-
duction and Al saturation in the 0-15 cm layer
of the soil was linear (Fig. 2.2:2), showing the
very strong effect of Al toxicity in reducing
yields.
The relationship between yield and Ca con-
tent of the ear leaves is shown in Fig. 2.2:3.

Presumably decreasing the Al saturation of the
soil, i.e., increasing the base saturation, in-
creases the level of Ca in the plant; Fig. 2.2:4
shows this relationship. Calcium is an impor-
tant structural component in the plant. As the
level of Ca found in the ear leaf increased, the
number of broken and lodged corn plants de-
creased (Fig. 2.2:5).
1977-1978 Soybeans
Table 2.2:1 shows the 1977-1978 soybean
yield as a function of lime applied in 1972. The
application of 8 t lime/ha incorporated to
30 cm more than doubled yields. The relation-
ship between yield and Al saturation in the
0-15 cm layer of the soil (Fig. 2.2:6) suggests
that the soybean yields tended to fall off more
rapidly at Al saturation values above 40%. The
overall response of soybeans to higher levels of
lime is less than that for maize, as can be seen
by comparing Fig. 2.2:1 with Table 2.2:1. As
can be seen in Table 2.2:1, there was no signif-
icant yield response among lime levels at 2 t/ha
and above, whereas for maize there was an
increase in yield for each increase in amount of
lime added. Because the soybeans did not
nodulate and N fertilizer was added, there was
no opportunity for a lime-related response in
N-fixation to become evident.
The soybean stand measured at harvest was
non-uniform and the crop was seriously
attacked by nematodes. Square roots of counts

22

0 10 20 30 40 50 60 70 80 90

ALUMINUM

SATURATION
0-15 cm

Figure 2.2:2.

Cargill-Ill maize grain yield as a
soil Al saturation measured in the
after harvest. Dark Red Latosol.

function of
0-15 cm layer
CPAC.

-c

-J
w

z

z
0

(%)

23

0.20 0.30 0.40 0.50 0.60
Co IN EAR LEAF (%)

Figure 2.2:3.

0.70 080

Cargill-11l maize grain production as related to
ear leaf Ca content. Dark Red Latosol. CPAC.

10 20 30 40 50 60 70 80
Al SATURATION (%)

Figure 2.2:4.

Relationship between Al saturation in the 0-15 cm
soil layer after harvest and the content of Ca in
Cargill-lll maize ear leaves. Dark Red Latosol.
CPAC.

W

Z
n-
(3

Z
0
LJ

5:

0

3.C

2.C

1.5

1.0

)
5 6
r 0
0 S
) 0 S *
0
0 00
5 0
0 0
) So

S

0.80

LL
W

-d

z

O

0.701

0.601

0.50t

0.40

0.30[

0

Depth of Lime
Incorporation

0 3 0 0-15cm.
o 0-30 cm

0

0
o

90 100

24

Depth of Lime
Incorporation
0 0-15 cm
O 0-30cm

0

0
0

0
0

0.6

0

0.8

1.0

(%)

EAR LEAF Ca

Figure 2.2:5.

Relationship between broken and lodged Cargill-
111 maize plants and the content of Ca found in
the ear leaf. Dark Red Latosol. CPAC.

100

0

751-

0

50 -

C)

a.

0

w

Q

0
Y.

251

0

0.2

0.4

I

I

I

I

I

25

Table 2.2:1.

Soil pH and Al saturation measured in July 1978 and

soybean grain yields as a function of levels of lime

applied and incorporated to two depths in 1972. Dark

Red Latosol, CPAC.

Soybean 1,
Lime Applied pH Al Saturation Grain Yield/-
in 1972 0-15 cm 15-30 cm 0-15 cm 15-30 cm 1977-78

t/ha ------- % -------- ------ % ------- kg/ha

Incorporated to 15 cm depth

1 4.2 4.3 61 54 1966 b
2 4.3 4.3 46 53 1862 b
4 4.8 4.5 15 32 1807 b
8 5.2 4.8 2 12 2113 b

Incorporated to 30 cm depth

0 3.9 4.2 80 67 1055 a
1 4.0 4.2 77 65 1304 a
2 4.3 4.4 53 44 2054 b
4 4.7 4.6 19 32 2248 b
8 4.8 5.0 6 2 2254 b

-/Means followed by the same letter are not significantly different

at the 5 per cent level by Duncan's Multiple Range Test.

26

) 10 20 30 40 50 60 70 80
Al SATURATION BEFORE PLANTING,O-15cm

Figure 2.2:6.

325

UFV-1 Soybean grain yield as a function of soil Al
saturation measured before planting (0-15 cm depth).
Dark Red Latosol. CPAC.

0

Stand (plants/ha)= 302,843 -882.24 (% Al
r =-n 8QQ8***

D3
Depth of L
0 Incorporation
0 0 0-15
CD 0 O 0-30

0 N

Saturation)

ime
-1972
cm
cm

10 20 30 40 50 60 70 80 90 100
Al SATURATION AT PLANTING, O-15 cm

Figure 2.2:7.

(%)

UFV-1 soybean stand in harvest area as related to
soil Al saturation measured in the 0-15 cm depth at
planting. Soybeans were initially thinned to a uni-
form stand.

2.

U U

-r

W
T-
0
-J

z

0
Cl)

2.21

2.0k

1.8r

1 .6

1.41

1.21-

S0 Depth of Lime
3 Incorporation
00 0 -15cm
0 0-30cm

0

I I I I I 1 0

1.0
C

90 100
(%)

0
-c

o
0.

az

O)
a
2

300

2751

250

-l I

0

3

I' I I

JI

27

0-s%

o* 0.55 r
+ *0
ogo
Q50-
w
w
C,)
z 0.45
w

S040;

0 I1

Figure 2.2:8.

300(

0
-o
S250(

O 200(
-4
w

Z 150(

z 100(
w
in
0 50C
C/)

Al SATURATION AT 0-15cm (%)

Content of Ca plus Mg in UFV-1 soybean seeds as a function
of soil Al saturation at 0-15 cm. Dark Red Latosol. CPAC.

0 0.40 0.45 0.50 0.55
SOYBEAN SEED Ca+Mg

0.60
(%)

Figure 2.2:9. UFV-1 soybean yield as related to seed con-
tent of Ca + Mg. Dark Red Latosol. CPAC.

( meq / IOOg)
2.0

Figure 2.3:1.

Effect of 600 kg/ha P as
or triple superphosphate
structed virgin Dark Red
1200 mm rainfall.

simple superphosphate (SSP) (contains calcium sulfate)
(TSP) applied to the lime 0-15 cm section of a recon-
Latosol profile after leaching with the equivalent of

Co + Mg

1.0

15-20
20-25
25-30

30-45h

3.0

4.0

TSP

SSP

E

IL
a-
0

45-60

60- 75F

75-90

I

I

I

29

of Meloidogyne javanicum nematodes, made by
Dr. R. D. Sharma*, were negatively correlated
with yield corrected by lime (correlation coef-
ficient of -0.43, significant at 0.3%); this indi-
cated that yields were probably being reduced
by these nematodes. As shown in Fig. 2.2:7,
there was a marked stand reduction at high
levels of Al saturation. Seed yield per plant was
also somewhat reduced at the zero lime level,
but one of the main effects of acidity in reduc-
ing soybean yields this year was the decrease in
the number of plants. The count of Pratylen-
chus brachyurus nematodes was greatly de-
creased in the plots were Al content was low
(r = 0.44, significant at 0.2%); this suggested
that liming may be an effective way of reduc-
ing damage from this nematode.
The effect of Al saturation on the plant's
ability to take up nutrients is reflected in the
decrease in the content of Ca plus Mg in the
soybean grain at harvest (Fig. 2.2:8). The posi-
tive correlation (0.54, significant at 0.001) be-
tween yield and Ca plus Mg content in the seed
is shown in Fig. 2.2:9.

2.3 MOVEMENT OF Ca AND Mg
K. Dale Ritchey, Djalma de Souza, Edson
Lobato and Osni Correa
Studies to investigate the downward move-
ment of Ca and Mg were continued in the
laboratory and in the field.

*CPAC staff member.

A soil profile 0 to 135 cm in depth was
reconstructed in 15 cm increments by filling 5
or 15 cm sections of 9.7 cm diameter heavy
PVC irrigation pipe with air-dried virgin Dark
Red Latosol soil excavated from near the fer-
tility experiments at CPAC.The 0-15 cm sec-
tion of soil had been previously treated with
8 t dolomitic lime/ha and 600 kg P/ha as either
simple superphosphate (SSP) or triple super-
phosphate (TSP), and allowed to incubate for
three weeks. A total of 1200 mm of distilled
water (equivalent to about 75% of the annual
rainfall) was added to the column over a period
of 31 days. Fig. 2.3:1 shows the distribution of
Ca + Mg from the two P sources. The CaSO4
contained in the SSP acts as a source of mobile
Ca which can move down through the soil
profile.
Another column experiment was carried
out to compare the effects of the accompany-
ing anion on the rate of Ca descent and on the
soil. The distribution of Ca shown in Fig. 2.3:2
reveals that Cl was more effective than SO4 in
rapidly moving Ca downwards, whereas CO3
was rather ineffective. Although Cl moved
rapidly, the chemical effects of the Cl on soil
properties were not as desirable as those ob-
tained with SO4. At the depth of maximum
concentration of Ca in the CaSO4 treatment,
the soil pH increased 0.4 units compared to the
check treatment. But at the depth of maximum

30

Co (meq/100g)

I-&

Ca(kg/ha) Paired Anion
o---o 0
o---o 800 CO-
a- 2000 CO3
A---o 800 SO04
- 2000 S0o4
o- 2000 C I
9-* Initial value

Figure 2.3:2.

Effects of various anions on the distribution of Ca
after leaching with the equivalent of 1200 mm rainfall
in a reconstructed virgin Dark Red Latosol profile 0-135
cm.

0- 15
15-20'
20-25'
25-30
30-45-

45-60-

60-75-

75-90-

90-105-

120-1

E
a.
o

I
1-
0U
r0

135-11

150-11

165-11

180-1

31

pH
4.0 4.5 5.0

Vir in 70 280 875
0-15

LSD.05
15-30-

E 30-45 -

I-
,- 45-60-
w

60-75- --

75-90-

Figure 2.3:3. Effect of varying rates of P as SSP on soil pH at various
depths in the Residual Effects of P Rate, Placement and
Time of Application experiment as sampled Aug. 1976.
Dark Red Latosol, CPAC.

Ca + Mg
I

0-

15-30

30-451

(meq/IOOg)
2

Figure 2.3:4.

The distribution of Ca+Mg in the Phosphorus Sources on Pastures experiment planted
to Brachiaria decumbens. The experiment was sampled in June 1977, about 3.5 years
after application of the fertilizers. Gafsa rock ("hiperfosfato") is a treated
Moroccan rock phosphate with high citrate solubility; Araxa rock is a Brazilian
rock phosphate of low citrate solubility. Dark Red Latosol, CPAC.

2

a-
0
w
i--
Q.

45-60

60-75

75-

90-1

105-1

h3

I

I

33

Ca concentration in the CaCI treatment, the pH
was decreased by 0.3 to 0.6 units. When SO4
is sorbed by soil constituents, hydroxyl ions
are released which increase the pH and reduce
the Al concentration, thus providing a better
environment for root growth.
Fig. 2.3:3 shows the increase in pH of soils
sampled in August 1976 from the "Residual
Effects of P Rate, Placement and Time of
Application" experiment (Section 2.7). The
surface incorporation of 875 kg P/ha as SSP
resulted in a substantial increase in soil pH
from 30 cm to at least 90 cm depth.
Since SO4 sorption is decreased as soil
acidity decreases, liming should help the
descent of CaSO4. Evidence supporting this
hypothesis is presented in Fig. 2.3:4 which
shows the distribution of Ca + Mg in the soil
for several P sources as a function of lime level
in the "Effects of P Sources on Pastures" ex-
periment (Section 2.8). The simple superphos-
phate treatment caused the greatest Ca move-
ment. Increasing the amount of lime applied
increased the amount of Ca leached.

2.4 K AND Mg FERTILIZATION
K. Dale Ritchey, Djalma de Souza and Edson
Lobato
The K-Mg experiment, begun in 1975-
1976, was continued by planting Cargill-111
maize as the test crop the second year and
Santa Rosa soybeans the third year.

In the second year, the K study plots were
subdivided. One-half of each plot received a
maintenance application of 83 kg K/ha and
the other half received nothing (residual).
Fig. 2.4:1 shows the relative yield of each
experimental plot as a function of the extract-
able K measured before planting the second
crop. The relative yield was calculated by divid-
ing the production in the residual sub-plot by
the production obtained on the corresponding
maintenance sub-plot. Using the Cate-Nelson
technique for separating the "responding soils"
from the "non-responding soils", it was found
that there was little response to added K for
soil K levels above 41 ppm (0.10 meq/100 cc
soil) in the double acid extractant. For the
plots with soil K levels of 16 to 28 ppm, the
average relative yield was 63% of the K suffi-
cient sub-plots.
Fig. 2.4:2 shows the grain yields of corn
and soybeans as a function of the residual K
applied the first year. Corn grain yields con-
tinued to respond to residual K up to 500
kg/ha, albeit at a reduced level at the higher K
rates. The economic value of the extra corn
grain obtained from the application of 62 kg
K/ha the first year was nine times the cost of
the K fertilizer added. When the increases
obtained in the second year from the
residual effects of this addition were included,
the net return to the investment in K was
18 to 1.

34

150r

0

0
0

0

0

SOIL K (ppm)

Figure 2.4:1.

The 1976-77 relative Cargill-111 maize yield as a function
of the double-acid extractable K measured after the first
crop. Dark Red Latosol, CPAC.

100

J

>-

0

-4

U
0o

0

n

0

0

w

I-
-J
w
ir

0

501-

0

0

41 50

100

150

I I

I

I

I

35

Maize 1976-77

ILSD.os

ILSDo5

0 100 200 300 400 500
0 100 200 300 400 500

K APPLIED

Figure 2.4:2.

1975

(kg/ha)

Cargill-11l maize grain yield at 15.5% moisture
harvested in 1976-77 and Santa Rosa soybean grain
yield at 13% moisture harvested in 1977-78 as a
function of the among ot K fertilizer applied in
1975-76. Dark Red Latosol, CPAC.

0

0 0
0
0- 0
0

0 0
1 I I I I

40
SOIL K

Figure 2.4:3.

Santa Rosa soybean grain production on the residual
subplots in 1977-78 as a function of double-acid
extractable K measured at planting. Potassium rates
from zero to 500 kg K/ha were applied in 1975-76.
Dark Red Latosol, CPAC.

0
"-C
-.
V

-J
w
-1
)-

Z
Q
n-
C.

3000r

25001-

z
0

0

0
L3
(D

Z

0
>-
U)

0

00

0

0
0
00
o
oo

0
-c

200C

150C

0

20

60

80

(ppm)

100

E
a.
a.

w
-J
m

0
I
X
LU

Figure 2.4:4.

GO
0)

DEPTH (cm)
Exchangeable K as a function of depth in Dark Red Labosol after the first crop,
for five levels of broadcast KC1 fertilization. Dark Red Latosol, CPAC.

37

In the third year of the experiment, none
of the plots or sub-plots received potassium.
Soybean yields reached maximum at about
140 kg residual K/ha (Fig. 2.4:2). The produc-
tion of Santa Rosa soybean grain as a function
of soil K measured in each residual sub-plot is
shown in Fig. 2.4:3. Where the soil K values
were from 20 to 29 ppm, the average yield was
71% of the K-sufficient treatments.
Exchangeable potassium as a function of
depth was measured after the first crop
(Fig. 2.4:4). Where economical rates of K were
applied there was little or no leaching loss of
potassium. At the 250 kg K/ha rate, 15 ppm
K were detected in the 45-60 cm layer and for
the 500 kg K/ha rate, 22 ppm K had leached to
the 60-75 cm layer.
The difficulties of accurately sampling for
K are demonstrated in Fig. 2.4:5. Potassium
contents of soil samples taken from within the
maize row and between the maize rows after
the second year harvest are shown. Composites
of two subsamples were collected at 0.5 m
intervals beginning 2 m within a plot excluded
from K fertilization for two years and continu-
ing across the plot boundary to 3 m within an
adjacent plot receiving 250 kg K/ha the first
year and 83 kg K/ha the second year. Where a
total of 333 kg K/ha had been applied over the
two-year period, the within-row samples had
up to 350 ppm K compared to less than 60
ppm K found between the rows. Apparently K

washed from the dead maize plants by rain was
concentrated within the rows, and gave an
unusually high soil test value. It is important
that this spatial non-uniformity be taken into
account when interpreting soil sampling results.
In the third year of the experiment a signif-
icant response to Mg was obtained. In the first
two years of maize production, increasingly
severe Mg deficiency symptoms appeared but
the low Mg plots still yielded well. In the third
year the soybean production on the 97 kg
Mg/ha treatment was 37% greater than that on
the 7 kg Mg/ha treatment and 19% greater than
on the 27 kg Mg/ha plots. The prescribed rates
of Mg were obtained by varying the proportion
of dolomitic and calcitic limestone making up
the 3 tons of lime/ha applied in 1975 to raise
the soil pH to 5.

2.5 K MOVEMENT IN A CLAYEY RED-
YELLOW LATOSOL
T. Jot Smyth, Djalma de Souza, K. Dale
Ritchey
The pronounced movement of K in the
"K and Mg Fertilization" experiment (Section
2.4) led us to consider whether similar results
would be obtained in the clayey Red-Yellow
Latosol. Measurements were made during the
dry season after the first soybean crop in the
"Sources, Rates and Placement of P Fertilizers
on the Clayey Red-Yellow Latosol" experi-
ment (Section 2.6). Physical and chemical

38

A

350

300 A
A Within Rows
0 Between Rows

250-
E
Q.
Q.

200

O
a) 150-

100-

A 0
50- 0

SI *
0

3 2 I *-0--0 I 2 3
DISTANCE FROM BORDER (m)
---0+0 kg K/ha --- -- 250+83 kg K/ha--

2 YEAR FERTILIZER APPLICATION

Figure 2.4:5. Double-acid extractable K at 0-15 cm depth measured
within the maize row and half-way between the maize
rows after harvest of the second year crop.

39

characteristics for this soil as presented in
Table 2.6:1 indicate that the native clayey
Red-Yellow Latosol has 25% more clay and
half the effective CEC of the topsoil for the
Dark Red Latosol.
The soil was sampled to a 90 cm depth in
the treatments receiving 44 and 352 kg P/ha as
broadcast SSP. Potassium was broadcast before
planting at the rate of 125 kg K/ha as KCI. Soil
K status before and after the first crop are pre-
sented in Table 2.5.1. Grain yields were 1885
and 1031 kg/ha and K in the soybean tops
were 28 and 14 kg K/ha, respectively for the
high ard low P treatments. Excess K was de-
tected in the 44 kg P/ha treatment at a depth
of 75-90 cm. The higher K uptake in the high P
treatment may have reduced the amount of K
moved below 60-75 cm.
The pattern of K accumulation for soy-
beans, variety Parana, is presented in Fig.
2.5:1. Over 60% of the plant K was taken up
by the full pod stage. Although significant
amounts of K should remain available in the
subsoil, these results suggest that adequate
levels of topsoil K are important for plant
uptake during the early stages of growth.

2.6 SOURCES, RATES AND PLACEMENT
OF P FERTILIZERS ON A CLAYEY RED-
YELLOW LATOSOL
T. Jot Smyth, Pedro A. Sanchez, Edson Lobato
Direct applications of rock phosphate is
one alternative for reducing the high costs

related to the large initial P fertilizer require-
ments of annual crops on high P fixing Oxisols.
However, information is needed on the extent
to which soluble P fertilizers may be substi-
tuted for the relatively insoluble rock phos-
phates without drastically reducing crop yields.
A field experiment was established during
the 1976-1977 rainy season with the following
objectives: 1) Evaluate over time the relative
plant availability of Patos de Minas rock phos-
phate and simple superphosphate (SSP) on a
clayey Oxisol, 2) study the influence of soil
pH, rates of application and combined methods
of placement on the availability of the P
sources.
Patos de Minas rock phosphate is a low
reactivity rock phosphate found near Patos de
Minas, Minas Gerais containing nearly 10%
total P of which 3% is soluble in citric acid.
The material used in this study was finely
ground with 85% finer than 200 mesh. On the
basis of total P this material sold for approxi-
mately one-fourth the price of SSP in Brasilia,
at the time of planting the first crop.
The experiment is located on a clayey Red-
Yellow Latosol. Soil properties are shown in
Table 2.6:1. Phosphorus sorption studies
(Fig. 2.6:1) indicate that this soil sorbs 70-90
ppm more P than the Dark Red Latosol at
identical soil solution levels. The Dark Red
Latosol is where most other phosphorus
research has been conducted at the Cerrado
Center. Four rates of SSP and three rates of

40

Table 2.5:1. Potassium distribution as a function of depth in a

clayey Red-Yellow Latosol before and after the first

crop. Potassium fertilization was 125 kg K/ha as KC1.

Extractable K (Double Acid Method) at Treatment
Depth Native soil 44 kg P/ha 352 kg P/ha Means

cm -------------------- K, ppm ------------------------

0-15 29 25 28 28

15-30 13 20 17 17

30-45 6 13 18 12

45-60 4 12 16 11

60-75 3 11 14 9

75-90 2 11 6 6

Means 10 15 16

LSD.05 Depth = 4

Treatment = 3

Depth x treatment = 7

41

IOOr

0

4-
0

oz
I-

a(
0.

vZ

D
y:

0 44

80

100

120

DAYS AFTER EMERGENCE

Figure 2.5:1.

Potassium accumulation pattern by soybean (variety
Parana) tops for two P fertilizer treatments with
a broadcast application of 125 kg K/ha as KC1.

80-

60-

Applied P
(kg/ha)
0 352

40[-

201-

0

20

40

60

I

I

I

I

I

I

Table 2.6:1.

Depth

Sand

Chemical and physical prope

source, rates and placement

Silt

Clay

O.M

!rties of the clayey Red Yellow Latosol site used for the P

; experiment. Typic Acrustox, clayey, oxidic, isothermic.

pH in Exchangeable* Eff. Al Fxt.
S H0 Al Ca + Mq K** CEC Satn. P**

%

--------------

17 13

14 11

13 8

11

7

-----------

70

75 2

79 2

-- 1

82

.65 4.55

.72 4.86

.13 4.93

.85 4.94

-- 5.03

-5.11

.37

.11

.03

0

0

0

me/100

.31

.24

.25

.23

.23

.23

cc--------

.08 .76

.03 .38

.02 .30

.01 .24

.01 .24

.01 .24

*Exchangeable cations by N KC1 extraction.
** P and K by the double acid extraction.
P and K by the double acid extraction.

- cm -

0-15

15-30

30-45

45-60

60-75

75-90

-%-

49

29

10

0

0

0

ppm

tr

tr

tr

tr

tr

tr

4N,
p
Mo

L

-

43

A Clayey Red Yellow Latosol

O Dark Red Latosol

Figure 2.6:1.

.005 .01 .05 .10 .50
P IN SOLUTION (ppm)

P retention curves for the clayey Red Yellow and Dark Red
Latosols. Equilibration in .001M CaC12 for 6 days.

E
.400
.400

m

a.

44

Table 2.6:2. Grain yields (13% moisture) of two consecutive soybean
crops (variety Parana) as affected by rate, placement,
source and residual effectsof P applications.

Broadcast Annually Absolute & Relative Grain Yield Cummulative
P Banded P Crop 1 Crop 2 Yield

kg P/ha

22
44

t/ha % t/ha

0.71 38
1.15 61

%0
/0

0.82 48
1.20 71

t/ha

1.54
2.35

Simple Superphosphate:

0
22
44

0
22
44

0
22
44

0

1.03
1.45
1.55

1.48
1.49
1.69

1.64
1.76
1.88

55
77
82

78
79
90

87
93
100

1.89 100

0.60
1.22
1.47

1.02
1.44
1.72

1.43
1.58
1.83

35
72
86

60
84
101

84
93
107

1.70 100

1.63
2.67
3.02

2.50
2.92
3.41

3.07
3.34
3.71

3.59

Patos de Minas Rock Phosphate:

0
22
44

0
22
44

0.40
1.04
1.39

0.86
1.08
1.55

21
55
74

46
57
82

0.38
1.11
1.50

0.90
1.26
1.64

22
65
88

53
74
96

1.02 54 1.20 71

0.79
2.15
2.89

1.76
2.34
3.20

2.22

0

44

88

176

352

88

352

704 0

45

Patos de Minas rock phosphate were broadcast
and incorporated to a 20 cm depth before the
first planting. Two maintenance rates of SSP
were banded below the seed in a factorial com-
bination with all but the highest broadcast rate
of both P sources. Dolomitic lime was applied
to these treatments at the rate of 3 tons
CaCO3-equivalent/ha. Four additional treat-
ments of two broadcast rock P rates with and
without a maintenance P band were limed at
the rate of 0.5 t/ha/year. This lime rate pro-
vided a moderate amount of Ca and Mg but
maintained the soil pH and exchangeable
acidity at approximately the original levels.
Treatments were arranged in a randomized
complete block design with three replications.
A blanket broadcast application of 125 kg
K/ha as KCI, 9 kg Zn/ha and 0.2 kg Mo/ha was
made before the first crop. The soil was again
rotovated before the second crop, banded P
treatments were repeated and a blanket appli-
cation of 83 kg K/ha as KCI was made.
Soybeans, variety Parana, were planted
both wet seasons in 40 cm rows at an approxi-
mate population of 500,000 plants/ha. Planting
the first year was on November 9, 1976 with
harvest on March 5, 1977. In the second year,
planting was on November 2, 1977 with har-
vest on February 21, 1978.
Poor nodulation occurred in both crops.
Nitrogen was sidedressed during the first crop
at 27 and 46 days after emergence at the

respective rates of 80 kg N/ha as urea and
120 kg N/ha as (NH4)2SO4. Sidedress applica-
tions of 60 kg N/ha as urea were made in the
second crop at 31 and 71 days after emergence.
Yields of the first and second crops are
shown in Table 2.6.2 and Fig. 2.6:2. A
maximum yield of 1.89 t/ha was obtained in
the first crop with the broadcast rate of 352 kg
P/ha as SSP, despite 27 days without rain
during pod filling and maturation. Two dry
spells occurred during the second crop; seven
days immediately following germination and
15 days during pod filling.
The importance of placement in reducing
the P fertilizer requirement is illustrated in the
first crop by the 176 kg P/ha broadcast SSP
rate with a 44 kg P/ha band application. Grain
yield in this treatment where only 220 kg P/ha
had been applied was the same as for the
352 kg P/ha broadcast SSP rate.
In the second crop, the 176 kg P/ha broad-
cast SSP treatment with a second band applica-
tion of 44 kg P/ha produced the maximum
yield. There was a tendency for yields of the
broadcast SSP treatments to decrease in the
second crop while yields of broadcast rock P
treatments were maintained or increased.
Results shown in Fig. 2.6:2 indicate that
yield responses to banded P applications
decreased as the level of broadcast P was in-
creased. Response to banded SSP among the
broadcast rock P treatments is small at rock P

CROP 2

A

iool-

S ._ -_ -.-.

80-

I-
Cf)

0
0
M
0-
m

U)
C/)
C,
N
ro

0
w

-J
w
W
WJ

I o- 0 ----1-- -

Anni

0P Source Banc
SSP
--- Rock P

0 44 88

U

176

4

ual SSP
- kg P/ha
22
44
0

352

201-

- -

I,-
-C

c~~0

P Source
-- SSP
--- Rock P
I I I

0 44 88

Annual SSP
Band- kg P/ha
a 22
A 44
0 O

352

176

Figure 2.6:2.

BROADCAST P (kg/ha) BROADCAST P (kg/ha)

Effects of rate, placement and source of applied P on two consecutive soybean crops.
Banded P curves represent the combined effects of 22 and 44 kg P/ha as SSP with the
broadcast rates of SSP or Patos de Minas rock phosphate.

Cf)
0)
Q
-J
wJ

4

100

80

4

2<

4:

L

120r

CROP I

Z#-

I

I

47

EMBRAPA CPACerrados
MANEJO DE FdSFORO E SISTER
MAS DE CULTURAL EM LVm

Photo 2.6:1. Mr. Jot Smyth after harvest of soybeans in the phosphorus sources and placement
fertilization experiment on a clayey Red-Yellow Latosol. CPAC, 1977.

48

Table 2.6:3.

Grain yields of two consecutive soybean crops (variety

Parana) as affected by rates, placement and source of

applied P at two lime rates.

Grain Yields
Lime Broadcast Annually Grain Yields
rate Rock P Banded P 1 2

t/ha ------- kg P/ha -------- -------- t/ha --------

3.0 88 0 0.40 0.38

22 1.04 1.11

352 0 0.86 0.90

22 1.08 1.26

0.5/crop 88 0 0.57 0.48

22 0.91 1.03

352 0 0.81 0.91

22 0.92 1.30

49

rates greater than 88 kg P/ha. However, it is of
interest to note that in the second crop, the
broadcast rate of 352 kg P/ha as rock P with an
annual maintenance band of 44 kg P/ha pro-
duced a yield almost identical to the 352 kg
P/ha broadcast SSP treatment. Based on fer-
tilizer prices at the time the broadcast P was
applied, this combined rock P and SSP treat-
ment had a cost equivalent to 176 kg P/ha of
SSP.
Reduction of the lime rate from one appli-
cation of 3 t/ha to applications of 0.5 t/ha/
crop had little effect on soybean yields (Table
2.6:3). After three months of equilibration,
soil pH was 5.1 and 4.8, respectively, for the
high and low lime rates. Aluminum saturation
in all treatments never exceeded 25%. Estima-
tion of available P by the Bray I method (Table
2.6:4) indicated that the rock P treatments
maintained higher soil test values throughout
the experiment at the low lime rate than at the
high lime rate. The lack of yield response at the
low lime rate may have been due to deleterious
effects of the low soil pH or the small magni-
tude of increase in available P.
The general trends in soil test P values over
time are a decrease for the broadcast SSP rates
and an increase for the broadcast rock P treat-
ments. The 704 kg P/ha broadcast rock rate
initially had 1/3 the soil test value for the
176 kg P/ha broadcast rate of SSP. However,

during the second crop these treatments
had identical soil test values. A continua-
tion of this trend suggests that the residual
effects of the high rate of rock P will con-
tinue for several more years. This experiment
will be continued by CPAC personnel in order
to fully evaluate the relative availability among
the two P sources.

2.7 RESIDUAL EFFECTS OF P RATE,
PLACEMENT AND TIME OF APPLICATION
Edson Lobato
The 1976-1977 and 1977-1978 Cargill-111
maize crops provided the seventh and eighth
consecutive harvests in the ongoing study of
residual effects of P as simple superphosphate
(SSP).
Phosphorus additions were suspended after
the fourth crop in all but the treatment which
received 35 kg P/ha broadcast before the first
crop and 35 kg P/ha banded for each crop. This
strategy provided an interesting comparison of
total yields for the first seven harvests among
four treatments, all of which received a total of
280 kg P/ha applied in various ways. The mean
production per crop among these four treat-
ments ranged from 4.93 t/ha for the applica-
tion of 35 kg P/ha banded before each planting
up to 5.16 t/ha for the initial application of
140 broadcast plus four banded applications of
35 kg P/ha each.

50

Table 2.6:4.

Soil test P values (Bray I method) as a function of time

for the broadcast P treatments.

Months after fertilization
Lime Rate Broadcast P 1 2 3 6 13 15 19

kg P/ha

---------- Extractable P (ppm) --------

Superphosphate

3.0

3.0

3.0

3.0

Patos de Minas

3.0

3.0

3.0

0.5/crop

0.5/crop

44

88

176

352

Rock

88

352

704

3.8

6.1

10.4

28.0

3.5

6.9

14.8

34.5

2.6

5.1

9.6

29.2

2.1

5.0

11.6

27.7

1.8

3.2

7.6

16.0

2.1

3.6

7.4

15.5

1.4

3.0

7.2

18.9

Phosphate

1.3

2.2

3.6

1.7

4.5

88

352

1.9

3.3

4.6

2.2

4.9

1.8

3.6

6.2

2.4

5.3

1.6

3.1

6.0

2.1

5.7

-76/77 Rainy season -

2.2

4.7

7.5

3.4

1.6

4.5

7.2

2.3

1.6

4.0

7.0

2.2

9.3 6.0 5.5

77-78 Rainy season

t/ha

51

Some of the beneficial effects of the high
levels of SSP fertilizer applied in this experi-
ment may be caused by the significant amounts
of gypsum (CaSO4) which the fertilizer con-
tains. As shown in Section 2.3 on "Movement
of Ca and Mg", gypsum allows for relatively
rapid movement of cations and substantially
improves conditions for root growth in the sub-
soil. Because the rooting potential in the high P
treatments is therefore greater, one would
expect the low P treatments to be affected
more adversely by dry periods than the high P
treatments, especially in the sixth, seventh and
eighth crops which received no supplemental
irrigation during periods of water stress. Dry
periods sufficient to impose serious yield losses
were observed (see Section 2.1 "Crop
Weather") for the sixth and seventh, but not
the eighth crop, and this may partially explain
why relative yields in the low P treatments for
the eighth crop tended to be slightly better
than in the previous two crops (Fig. 2.7:1).
Fig. 2.7:2 shows the changes in the double-
acid extractable soil P with time. Because
835 kg of the 875 kg P/ha treatment were
applied before the second rather than the first
crop, the soil P line for this rate is displaced
one crop to the left. By the sixth crop, all of
the broadcast treatments, with the exception
of the 875 kg/ha rate had fallen to less than
20 ppm; the critical level for corn using this
extractant is considered to be 18 ppm P. The

eighth harvest yields accordingly would be
expected to be and are somewhat lower than
the maximum (Fig. 2.7:3 and Table 2.7:1).
The economical optimum level of P fertili-
zation based on the cumulative yields of the
eight harvests is about 600 kg P/ha using a
maize grain/P ratio of 16.6 (Fig. 2.7:4). The
current price of P is somewhat lower than this
so the optimum P rate would be even higher.
The optimum P rate based on the first two
crops was about 220 kg/ha, and if a farmer had
applied only this amount, total production
would have been 18 t/ha less over the eight
crop period.

2.8 EFFECTS OF P SOURCES ON PAS-
TURES
Edson Lobato and Claudio Sanzonowicz
The experiment initiated in February 1974
to study the effect of P rates and sources of P
fertilizer at various levels of liming on produc-
tion of Brachiaria decumbens forage grass was
continued.
As mentioned in the 1975 Annual Report,
the initial production from the slightly soluble
Araxa rock phosphate was low (Fig. 2.8.1).
The application of limestone initially tended to
reduce Brachiaria production somewhat in the
medium and high Araxa treatments, supporting
the hypothesis that the rate of dissolution of
the slowly soluble rock phosphate is more
rapid where the soil acidity is left uncorrected.

n-

0

a-

0

LU
-cJ
>

w

-J

52

Broadcast in 1972

0---0o
os0, '%
f.' *r

P Rate Z Yield
kg/ha (t/ha)

"-.280 (38)
0---0

S140 (26)
-_0--

- I I I I I I I

- Broadcast + Banded
before each crop Bcst.+35 nd.
35 Bcst.+35 Band.
(39)

-\ 140 Bcst.+ 140 Band.
s_--.A(39)

Successive applications
--- No additional P applications

I I

I I I I I I

I 2 3 4 5 6 7 8

CONSECUTIVE CORN CROPS (1972-1978)

Figure 2.7:1. Effects of rate, placement and time of simple
superphosphate applications to eight continuous
corn crops. Grain yields expressed as percent
of the 560 kg P/ha broadcast treatments. Adja-
cent numbers are P rates in kg P/ha. Numbers in
parentheses are the cumulative grain yield in
t/ha. CPAC, 1972-1978.

I I I I I I I

100

80

60

40

20 -

0

100

60

401-

20

0

I

80

53

100

90

80

E
CL
a.

w
Q-

-J

H
CD
M

H
w
0:

0
w
L-
I
0W

0

Figure 2.7:2.

I 2 3 4 5 6 7 8
CONSECUTIVE MAIZE CROPS
Changes in the level of North Carolina double-acid extract-
able soil P for five rates of broadcast P for eight consecu-
tive maize crops. The second and fourth crops (irrigated)
were grown during the dry season. Dark Red Latosol. CPAC.

8750D
I
\
l

I

O
\
I
I
t
l
I
I
I
I
I
I
t
I
I
I
I
\
\

560 \
\
0

60

50

40

30

\0

bN
N%\1 1

201

10

b,
\\

54

Table 2.7:1. Cargill-Ill maize yields by crop as a function of rates,

placement and time of simple superphosphate applications.

CPAC, 1972-1978.

1/
Applied Number of Total P Grain Yield by Crop-
P Applications Applied 1 2 3 4 5 6 7 8

kg/ha kg/ha -------------- t/ha ---------------

Broadcast

70 1 70 5.2 3.3 .9 1.8 1.7 .7 .5 .7
140 1 140 6.3 5.7 2.2 3.3 3.0 1.9 .9 1.2
280 1 280 6.8 7.5 3.0 6.4 4.8 3.9 2.6 2.6
560 1 560 8.0 8.5 3.9 9.1 6.3 6.2 4.7 4.9
8752/ 1 875 2.3 9.6 4.6 9.0 6.6 7.0 5.3 6.4

Banded
35 4 140 2.4 5.1 3.1 6.0 4.5 2.3 1.3 1.5
70 4 280 3.9 6.6 3.4 8.1 5.9 4.7 2.7 2.9
140 4 560 4.8 8.4 4.2 9.0 6.9 6.9 4.9 5.1

Broadcasted once
+ Banded

35 + 35 8 315 4.6 6.0 2.6 6.5 5.8 4.9 4.2 4.7
140 + 35 4 280 6.7 7.3 3.3 7.2 5.4 3.7 2.5 2.7

1-Crops 2 and 4 were
irrigation.

grown during the dry seasons with supplemental

7/Received 35 kg P/ha the first crop and 840 kg P/ha the second crop.

55

71

-J
>-
Z

4-

0

Figure 2.7:3.

Z

0
W
N

-J

0

0
4-

Figure 2.7:

5 10 15 20
SOIL P (ppm)
Cargill-Ill maize grain yield obtained in the eighth
harvest as a function of North Carolina double-acid
extractable soil P sampled May 1978, for five broad-
cast rates of P. Dark Red Latosol. CPAC.

0 70 140 280 560
TOTAL P APPLIED (kg/ha)
4. Cumulative Cargill-1ll maize grain yield for eight
crops as a function of total P applied. The straight
line shows the number of kg of corn grain needed to
pay for the P applied, using a ratio of 16.6 kg grain
per kg P.

6-

5_

4-

31-

2[

I

I

56

--- SSP
Araxa Rock
-- SSP Surface Applied
37 kg P/ha /yr

P
Rate
(kg/ha)
0112
S600
o 150
0 37

Method
of Applic.

surface
broadcast
broadcast
broadcast

100 -A---------A-----

73-74

74-75 75-76

AGRICULTURAL YEAR

76-77

Figure 2.8:1.

Yield of Brachiaria decumbens forage dry matter relative
to the 600 kg P/ha (SSP) rate on an agricultural year
basis for Araxa rock phosphate, SSP incorporated Feb.
1974, and SSP surface-applied annually, beginning Oct.
1974. Dark Red Latosol. CPAC.

120 -

w

w

w
IJ

n-

60

40 -

20 -

0

80-

cr/

57

Since large quantities of rock phosphate are
easily available in Brazil, there is great interest
in learning how to use it efficiently. For the
soluble P forms, a slight response to liming is
evident (Table 2.8:1); but the response is very
much less than that obtained with other crop
species, indicating that Brachiaria decumbens is
quite tolerant to acidity. It is Brachiaria
decumbens' tolerance to acidity that makes
feasible the strategy of growing this pasture
grass while using soil acidity to increase the
solubility of rock phosphate.
After the fourth cut, the beneficial effect
of the zero lime rate for the Araxa rock seems
to have become less important, perhaps due to
depletion of soil Mg in the unlimed treatments.
However, the overall productivity of the Araxa
treatments began to increase in relation to the
highest yielding treatments (SSP).
Phosphorus in the form of Araxa rock is
currently only 40% of the cost of P in the form
of simple superphosphate. By the third year
the Araxa treatments were yielding equal to
the corresponding SSP treatments (Fig. 2.8:1),
and the relative yields (Araxa vs. SSP) for the
cumulative production obtained in 10 cuts was
67% at 150 kg P/ha and 85% at 600 kg P/ha.
Clearly there is a long range economic advan-
tage to using rock phosphate.
These data suggest the possibility of apply-
ing both a soluble P source and an inexpensive
rock phosphate to allow for a high level of
initial production as well as a good long-term

residual effect. Termofosfato, a thermally
altered mixture of Brazilian rock phosphate
and magnesium silicate, produced Brachiaria
decumbens yields equal to those produced by
SSP (Table 2.8:1) and could be a viable source
of easily available P. Termofosfato also has a
good liming effect.
The annual surface application of 37 kg/ha
P as SSP beginning the second year gave us an
unusually good response (Fig. 2.8:1 and Table
2.8:1) with the significant advantage that
pasture P levels could be increased without dis-
turbing the soil.
Soil K levels, even with relatively high
K fertilization rates, can be reduced below
the critical level with high forage produc-
tion. The relationship between soil K remain-
ing and the total dry matter removed from
the field in the first ten cuts is shown in Fig.
2.8:2. There is a highly significant decrease
in soil K for the more productive plots.
Where Brachiaria yield exceeded 26 t/ha,
the residual soil K values ranged from 17
to 59 ppm averaging about 40 ppm, which
is below the critical K level for maize. In
contrast, where production was 25 t/ha or
lower, the soil K ranged from 63 to 119
ppm, averaging well over the critical level of
41 ppm. During the 10 cuts a total of 540
kg K/ha was added to the plots. This was
probably insufficient for the high producing
treatments and slightly more than necessary for
the medium and low producing treatments. If

Means

58

Table 2.8:1.

Cumulative dry matter Brachiaria decumbens forage yield
for 10 cuts (to June 1977) as a function of source and
level of P for 3 lime rates. The soil Al saturation is

shown as sampled June 1977. Dark Red Latosol. CPAC.

Soil Al Saturation Forage Dry Matter
Phosphorus Lime (t/ha) Lime (t/ha) Means
Source Rate 0 3.0 4.5 0 3.0 4.5

kg/ha

Super-
phosphate

37
150
600

Means

Termo- 37
fosfato* 50
150
600
Means

Araxa Rock

Hiper-
fosfato**

Means

P 37
150
600

37
150
600

N.C. Rock P 150

SSP
Control

112+
0

-------- t/ha --------

77
83
80

57
42
32

76
65
30

31
25
0

78
82
82

50
38
28

78
79
52

82

77
79

48
37
11

27

51
42

21
26
15

17
7
0

22
21
12

18
13
5

7

23
16

18.1
37.9
43.8
33.3

11.1
37.6
47.8
32.2

4.3
26.1
37.8
22.7

14.2
35.7
40.0
30.6

14.8
38.8
46.0
33.2

13.0
35.5
46.0
31.5

9.3
26.1
39.8
25.1

15.6
39.1
46.7
33.8

17.8
37.4
48.4
34.6

24.4
38.7
47.7
37.0

11.4
23.9
40.2
25.1

14.9
38.4
44.0
32.5

16.9
38.0
46.1

16.2
37.3
47.2

8.3
25.4
39.3

14.9
37.7
44.2

35.6 36.8 39.0 37.1

30.6
7.5

35.1
6.3

35.6
9.4

33.8
7.7

*

Rock P fused with MgSiO4.
**
Treated Moroccan Rock P.
+Surface applied as 37 kg P/ha

each year, beginning Oct. 74.

-- % -------

o
400- o
. o o o o

08 0
0 0 .00
0" O0 0 0

030- 0

o 0
5" 0 0

-I 0
S10 D
0
0) O

20 30 40 50 60 70 80 90 100 I I0 120

RESIDUAL SOIL K (ppm)

Fig. 2.8:2. Soil K level found in the 0-15 cm layer June 1977 as related to total Brachiaria dry matter
removed in 10 harvests. Dark Red Latosol, CPAC.

Table 2.8:2.

Forage dry matter yield of Brachiaria decumbens, per harvest and total, as a function of P
fertilizer source and level, applied Feb. 1974. Averaged over 3 lime rates. Dark Red
Latosol. CPAC.

Phosphorus Cutting Dates by Rainy Season
Source Rate 1974-1975 1975-1976 1976-1977 1977-1978
May Dec. Mar. Aug. Dec. Mar. May Nov. Feb. Jun.++ Jan. Jun. Total

kg/ha --------------------------------- t/ha -------------------------------------
SSP 37 .2 1.8 2.6 1.6 .4 1.9 2.6 1.4 1.6 2.8 1.4 .4 18.7
150 .4 4.7 7.6 3.3 1.3 4.6 5.3 1.7 4.6 4.4 3.5 .8 42.2
600 .9 6.3 8.4 4.2 2.0 4.9 6.7 1.4 6.6 3.8 5.5 2.4 53.1

Termofosfato* 37 .2 2.1 2.3 1.1 .4 1.6 2.1 1.4 1.3 2.4 1.1 .4 16.4
150 .5 4.9 7.3 2.7 1.2 6.0 4.8 2.0 3.8 4.1 3.3 .7 41.3
600 1.3 6.8 8.8 3.2 2.5 6.3 6.7 2.0 6.4 3.3 6.6 1.8 55.7

Araxg Rock P 37 .1 .3 .5 .8 .1 .9 1.7 .9 .8 2.1 .9 .6 9.7
150 .2 1.4 3.1 2.4 .6 3.5 4.4 2.0 3.3 4.5 2.6 .5 28.5
600 .2 2.3 7.0 3.8 1.3 5.7 6.6 2.1 5.8 4.4 5.1 1.4 45.7

Hiperfosfato** 37 .2 1.4 2.3 1.7 .4 1.7 2.0 1.5 1.1 2.6 1.2 .4 16.5

150 .4 3.8 7.0 2.8 1.0 6.2 5.5 1.9 4.6 4.5 3.2 .8 41.7
600 .6 6.5 7.1 3.3 1.5 6.2 6.6 1.9 6.9 3.6 6.4 1.8 52.4

N.C. Rock P 150 .4 4.0 7.0 3.1 .9 5.6 5.0 1.8 4.3 5.0 3.2 .7 41.0

SPP 185+ .1 .9 5.0 1.3 .6 7.1 6.0 2.7 7.3 2.9 6.2 1.4 41.5

Control 0 .1 .3 .4 .6 .2 .9 1.5 1.0 .7 2.0 .9 .5 9.1
*Ro **;
Rock P fused with MgSi04; Treated Moroccan rock P.

+Surface applied as 37 kg of P/ha on Oct. 1974, Nov. 1975, Nov.

1976, Nov. 1977, and Oct.

1978.

+Because of severe insect and disease damage the entire experiment was replanted after the 10th. cut.

0)
0

61

the average value of 1.5% K found in the forage
in the first two cuts is considered typical of all
ten cuts, the amount of K removed from the
medium level treatments would be about
225 kg/ha. In order to avoid depleting the
small soil K reserve during the first three years
of production at medium P investment rates, at
least 225 kg K/ha would have to be added as
fertilizer. In order to correct and maintain a
reasonable K level in the soil, probably at least
100 kg K/ha should be applied annually.
In the unlimed treatments, the total Mg
applied was 50 kg/ha, whereas the forage may
have removed as much as 150 kg Mg/ha. In
some of the rock phosphate treatments, S may
also become limiting after continued cropping.
By the time of the tenth cut, disease and
insect ("cigarrinha", Deois flavopicta) attacks
seriously reduced stands in the low P treat-
ments, and the entire experiment was re-
planted. The 11th and 12th cuttings shown in
Table 2.8:2 represent the production obtained
from the new plantings.

2.9 N FERTILIZATION
K. Dale Ritchey
Dark Red Latosol
The fifth and sixth consecutive rainy
season maize crops (Cargill-111 hybrid) on the
clayey Dark Red Latosol were planted Novem-
ber 6, 1976 and October 19, 1977, respec-
tively. In the fifth year all the N treatments

received 30 kg N/ha 17 days after planting and
the rest at 35 days. In the sixth year N was
similarly applied as urea at 25 and 50 days.
The general yield responses to N were
similar to that of previous years except that the
break in the yield curve at about 90 kg N/ha,
which in previous years was quite obvious, was
rather indistinct (Table 2.9:1). In the fifth and
sixth years the yield with 60 kg N/ha was 75%
and 78%, respectively, of the yield at the
highest level of N (200 kg N/ha).
Treatment 2, which never received any N,
except 20 kg applied the first year, produced
3.45 t maize grain/ha the sixth year. Average
production of this treatment for the six-year
period was 3.40 t/ha. Production the fifth year
was only 1.85 t/ha. The sudden yield reduction
may have been related to the fact that 314 mm
of rain fell during the month before the maize
was 20 days old. This rainfall exceeded by 50%
that of the next highest rainfall observed for
the same month during the six years of the
experiment. This precipitation may have
caused losses of soluble N by leaching N below
the root zone or by denitrification. In the sixth
year, maize was planted before the rainy season
began so as to maximize utilization of the
"N flush" released by wetting the soil after the
dry season.
In the fifth year, all of the treatments
received a maintenance application of 100 kg
Mg/ha and 3 kg Zn/ha except treatments 1, 4,

62

Table 2.9:1. Cargill-1ll maize grain production (15.5% moisture) for

various levels of N applied as urea on the clayey Dark

Red Latosol. CPAC, 1976-77 and 1977-78.

1976-77 1977-78

Other Grain Grain
Tmt. N applied Tmt.* Yield** N applied Yield**

----- kg/ha ------

2+

3

6

5

9

0

60

100

140

200

6

7

1

4

100

100

100

100

8++

10+++

0

0

t/ha

1.85

4.59

5.36

5.76

6.11

-Zn -Mg

+Cu

-Zn

5.36

5.27

5.56

5.58

4.45

3.28

kg/ha

d

b

ab

a

a

ab

ab

0

60

100

140

200

t/ha

3.45

4.31

5.16

5.21

5.53

100

100

30

30

a

a

b

c

5.16

4.98

4.01

4.10

4.09

3.36

0

30

cd

b

b

a

a

a

a

bcd

bc

bc

d

Unless otherwise indicated all treatments received a supplementary
application of 100 kg Mg/ha and 3 kg Zn/ha. In addition, treatment
1 received 4 kg Cu/ha.
**
Values followed by the same letter within the same year were not
significantly different at the 5% level (Duncan).

+Received 20 kg N/ha the first year (1972-73) and none thereafter.

++Received lime-coated ammonium nitrate the first three years of the
experiment and 200 kg N/ha as urea the fourth.
++ Received sulfur-coated urea the first three years of the experiment
and zero N thereafter.

63

and 7 which were modified as shown in Table
2.9.1. The yields from these treatments can be
compared with the yield from treatment 6,
since they all received 100 kg N/ha. There was
no significant difference among these treat-
ments, indicating that the 150 kg Mg/ha and
11 kg N/ha applied in 1972 were still supplying
the needs of the plants. This sufficiency con-
tinued into the sixth year as shown by compar-
ing the yields of treatment 7 (no Mg or Zn)
with treatment 6. The Mg content in the stover
reflected the nutrient additions as shown in
Table 2.9:2.
Red-Yellow Latosol
A supplementary experiment was planted
to Cargill-111 maize hybrid November 5, 1976
on a clayey Red-Yellow Latosol (LVA) at
CPAC in order to evaluate the N response of
another important Cerrado soil. The Cerrado
vegetation was cleared mechanically and
2.6 t/ha of dolomitic lime (76% relative neu-
tralizing power), 176 kg P/ha, 125 kg K/ha,
7.5 kg Zn/ha, 1.1 kg B/ha and 0.2 kg Mo/ha
were broadcast and incorporated.
Urea was applied as shown in Table 2.9:3.
The maximum yield was considerably lower
than that obtained for the fifth year of the
Dark Red Latosol N experiment, which was

planted one day later. The P fertilization was
insufficient, as shown by yields of 2.77 t/ha in
auxiliary plots receiving twice the P applied to
the main experiment. Acidity as suggested by
the low soil pH (5.0) probably limited yields as
did a 40-day dry period. The maize plants
developed more slowly and flowered ten days
later than in the Dark Red Latosol experiment.
The maize was still in the late flowering stage
when the drought began. In addition, rooting
depth was less at the Red-Yellow site.
The yield from the 50 kg N/ha treatment
was 68% of the highest yielding treatment
(150 kg N/ha). The grain production from the
zero N treatment was only 0.33 t/ha or 13% of
that obtained form the highest yielding treat-
ment. In the Dark Red Latosol the zero N
treatment yielded 30% of the highest treatment
the fifth year after cultivation began; the
average over six years was about 57% of the
annual high yield. The relatively lower produc-
tion on the zero N treatment in the Red-
Yellow Latosol may have been partially caused
by lower N-supplying power of the soil plus the
presence of a large amount of organic residues
and debris left from the land clearing opera-
tions which would have resulted in additional
microbial immobilization of some fertilizer N.

Table 2.9:2. Mineral content of Cargill-lll maize hybrid grain and stover as affected by pre-plant

incorporation of supplementary amounts of sulfates of Mg, Zn, and Cu. CPAC.

1976-77.

Mineral Content

Tmt. Supplement N P K Ca Mg Mn Cu Zn Fe

----------- ppm-------------

Stover

Mg

Mg, Zn

Mg, Zn, Cu

Mg

Mg, Zn

Mg, Zn, Cu

0.67

0.64

0.69

0.67

0.04

0.04

0.05

0.06

0.05

0.07

0.06

0.08

0.54

0.45

0.50

0.45

0.07

0.11

0.10

0.12

31

29

24

24

5

5

5

7

34

31

36

33

O)
4S

Grain

1.59

1.58

1.58

1.45

.23

.27

.25

.24

.32

.37

.34

.34

.5

.5

.6

.5

.09

.10

.10

.09

5

7

6

5

2

2

2

2

15

18

17

16

28

32

33

26

7

4

6

1

7

4

6

1

------------- % ---------------

65

Table 2.9:3. Cargill-1ll maize grain production (15.5% moisture) for

various levels of N applied as urea on clayey Red-Yellow

Latosol. CPAC. 1976-77.

Total N N applied
applied at planting Grain Yield

--------- kg/ha ---------- t/ha

0 0 0.33 d

50 20 1.71 c

100 20 1.93 bc

150 20 2.50 a

200 20 2.29 ab

100 0 1.67 c

100 100 1.59 c

*Values followed by the same letter were
the 5% level (Duncan).

not significantly different at

66

Sr. Luis Souza-Lima, a prominent Cerrado farmer,
explains his hopes for the future of agriculture in
the Brazilian Cerrado based on the cooperative
EMBRAPA-CPAC/Cornell/NCSU research. Unai,
Minas Gerais. June, 1978.

67

AMAZON JUNGLE OF PERU

Sr. Luis Gonzalez, a Peruvian small farmer, and Ing. Ruben Mesia,
Peruvian head of the Yurimaguas Station and Extrapolation Program,
discuss the transferability of the improved continuous cropping tech-
nology for peanuts. Km. 22, Yurimaguas-Tarapoto road. July, 1979.

68

Aerial view of physical plant (upper left) and some field trials at Yurimaguas Experiment Station.
Part of Chacra I is in upper right. July, 1979.

69

During the biennial period of 1978-1979,
soil-crop management research was conducted
on both the Yurimaguas experiment station
and on surrounding farms. Although the on-
farm trials (reported in Section 4.1) are having
a significant impact on the extension of the im-
proved agronomic technologies developed over
7 years on the experiment station to the area's
small farmers and their acceptance of these, the
agronomic technologies are in continual need
of refinement-especially as second and third
generation problems manifest themselves and
as crop varieties better adapted to the local
conditions are found. Accordingly, the con-
tinuous cropping experiments, adaptation trials
with annual crops and forages, and other soil-
crop management refinement trials have con-
tinued on the experiment station.
Reported in this section are the annual
crops and forage adaptation experiments, the
management of annual crops study, the deep
lime trial, the K and Mg fertilization experi-
ment, the multiple cropping-N experiment, the
study on the effect of clearing and continuous
cultivation on soil physical properties, the con-
tinuous cropping experiment, residual lime
effects trial, minimum inputs and kudzu fallow
studies and mulching and composting trials.
All trials on the experiment station were
conducted on the Yurimaguas soil series which
is representative of a large portion of soils in
the Peruvian Amazon Jungle basin. The soil is a
Typic Paleudult, fine loamy, siliceous, isohy-
perthermic. It is well-drained, quite acid, with
high percentage Al saturation, low in organic
matter, deficient in N, P, K, Ca, Mg and, in
some cases, S, B and Mo. The sandy texture of

the topsoil prevents any major P fixation pro-
blem but contributes to the susceptibility of
this soil to compaction-be it by rain, animals
or machinery.

3.1 CROP WEATHER
D. E. Bandy
Rainfall was the most dominant climatic
factor influencing crop growth and production
in Yurimaguas during 1978 and 1979. The rain-
fall and other climatic data for Yurimaguas
during 1978. and 1979 are presented in
Tables 3.1:1 and 3.1:2, respectively.
In general, 1978 had a more stable cli-
matic pattern than 1979 with only the June-
July period having water stress conditions suf-
ficient to reduce crop yields (Figure 3.1:1). In
1979, rainfall distribution was very erratic,
sometimes being excessive and at other times
very limiting (Figure 3.1:2); this reduced yields
by an average of 30-40%.
Air temperatures were normal for the area
with night temperatures showing the most fluc-
tuation. June through September usually had
the coolest night temperatures during the year.
Due to the more uneven distribution of
rainfall in 1979, solar radiation was slightly
higher than in 1978, thus higher daytime tem-
peratures and pan evaporation rates were re-
corded. Bare soil temperatures at the 2 cm
depth were nearly 1.50 C higher for 1979.
Normally, it is not very windy in Yuri-
maguas; 1978 and 1979 were no exceptions.
The prevailing winds were from the NE (Figure
3.1:3) at an average velocity of 3-4 m/sec
(Figure 3.1:4).
Advective rains caused by air masses

70

Map of Yurimaguas Experiment Station. July, 1979.

1141.30

Table 3.1:1. Summary for climatic data for 1978. Yurimaguas Experiment Station.

Air Temperature Total Total Solar Min. Wind Soil Temperature
Month Max. Min. Mean Rainfall Evapor. Radiat. Rel Velocity 2cm 5cm 10cm 50cm 100cm
Hum.

----- C ------

31.4

31.8

30.6

31.0

30.9

31.1

31.1

29.9

32.0

31.8

31.7

29.4

21.6

21.8

22.9

21.7

21.5

19.1

19.5

17.6

19.7

20.5

21.1

20.4

26.5

26.8

26.8

26.4

26.2

25.1

25.3

23.8

25.9

26.2

26.4

24.9

(mm)

198.1

121.7

262.8

358.5

230.4

171.6

111.5

165.0

163.1

263.8

211.9

239.8

(mm)

97.0

84.8

66.2

58.1

68.4

79.6

85.4

87.1

111.1

113.5

109.8

74.9

lang/day

344

318

269

318

307

327

329

333

404

384

329

297

58

58
58

62

61

44

35

33

53

54

56

58

61

m/sec/day

5.3

4.3

4.4

5.7

5.0

5.4

5.3

6.0

6.0

6.7

6.3

7.0

------------ oC

34

33

32

32

31

33

34

33

35

35

36

32

33

32

31

31

32

32

32

32

34

33

34

31

31

31

31

29

29

30

30

30

32

31

32

29

---------------

27 27

26 26

28 28

26 26

25 25

25 25

26 25

25 25

26 26

26 26

26 26

25 25

Year 31.0 20.6 25.8 2082

January

February

March

April

May

June

July

August

September

October

November

December

1036 330 53 5.6

33.3 32.3 30.4 25.9 25.8

Table 3.1:2. Summary for climatic data for 1979. Yurimaguas Experiment Station.

Air Temperature Total

Month

January

February

March

April

May

June

July

August

September

October

November

December

Total Solar Min. Wind Soil Temperature
1 Evapor. Radiat. Rel. Velocity 2cm 5cm 10cm 50cm 100cm
Hum.
(mm) lang/day % m/sec/day -------------C ---------------

Max. Min. Mean Rainfal
------ ------ (mm)

32.7 21.1 26.9 188.2

31.4 21.5 26.5 59.1

30.6 21.1 25.9 450.7

31.1 20.7 25.9 191.6

31.2 21.1 26.2 178.5

30.1 18.4 24.3 85.7

30.8 19.4 25.1 120.3

32.4 20.0 26.2 113.0

31.8 20.3 26.1 160.0

31.9 21.0 26.5 188.6

30.9 21.3 26.1 328.8

31.2 21.7 26.5 149.2

390

319

303

300

318

296

306

387

413

395

317

354

50

61

63

62

62

50

48

43

42

44

62

63

6.7

5.5

5.5

5.7

4.5

5.3

4.3

5.0

5.7

6.4

5.8

5.7

Year 31.3 20.6 26.0 2214

36 34 31 26 26

35 33 31 26 25

32 30 29 26 25

34 32 30 26 26

34 32 30 26 26

33 31 29 25 25

34 32 30 26 25

36 34 32 26 26

37 33 32 27 26

37 33 32 27 26

34 32 31 -- 26

34 31 30 -- 26

34.7 32.3 30.6 26.1 25.7

N,

120.5

88.6

62.8

95.9

87.5

88.3

88.1

99.5

113.5

105.6

74.7

80.4

1085 342 54 5.5

.1

I 1 1 "" "" I I i J I
JAN IFEB I MAR APRIL MAY J UNE I JULY AUG SEPT IOCT I NOV [ DEC I

.I I

Figure 3.1:1. Rainfall at Yurimaguas Agricultural Experiment Station, 1978, Yurimaguas, Peru.

r

120

100

E
E
-J
-J
LL
Z
nr

80

60

40

20

0

-1
CO)

I

,I

I

I

I

I

I

I

lI

SI

I

I

I, II

Figure 3.1:2. Rainfall at Yurimaguas Agricultural Experiment Station, 1979, Yurimaguas, Peru.

120

100

E

EJ

n-

80

60

40

20

II

0

JAN FEB MAR APRIL MAY JUNE JULY AUG SEPT I OCT I NOV I DEC

I

I

I1

I

I

I

75

NW /

/
/
/

/
I
I
WI.

N

I --- '
I

I1-
I

N

SW
N

S
S

Figure 3.1:3. Prevailing wind direction. Yurimaguas, Peru, 1978, 1979.

E

E

"o. *-*'

76

N
--4----

\
6m/sec
4- E

S

Figure 3.1:4. Average velocities of winds by direction. Yurimaguas, Peru, 1978, 1979.

W

77

N

NW / \ NE

/ --- ~ \

/N

\

20%
4-E
I
I
/
/
/
/

SE

S

Figure 3.1:5. Direction of rain movement. Yurimaguas, Peru, 1978, 1979.

/

I
7
I

\

-\

/
/
I/
I
I
I
\
\
\

Ns

SW \

78

moving westward from the Amazon Jungle
Basin occurred more than 70% of the time
(Figure (3.1.5). These rains usually are of low
intensity (<10 mm/hr), may last for one-two
days and result in very little crop lodging or
soil erosion. But, approximately 10% of the
time, convective rains developed NW and SW of
Yurimaguas in the high jungle areas of the
Andes. These are cold rains of high intensity
and usually are accompanied by strong winds,
which can result in severe lodging of corn, soy-
bean, cassava and tall varieties of upland rice.
In addition, soil erosion could become a pro-
blem on large areas of recently cleared land
that does not yet have a crop established. Un-
fortunately, these convective rains have not
shown any sharply defined seasonal patterns;
thus, changing the crop or its planting date
would not be of much benefit.
Soil management in relation to the Yuri-
maguas climate is essentially managing soil
moisture and soil temperature regimes. The re-
presentative soil of the region is an Ultisol,
classified as Typic Paleudult, fine loamy,
siliceous, isohyperthermic; it is well drained,
quite acid, low in organic matter, deficient in
most nutrients, with high amounts of ex-
changeable aluminum which increase with
depth. Because of these acid and aluminum
saturated subsoil conditions, plant roots us-
ually are restricted to the top 10-15 cm of
the soil where the exchangeable aluminum has
been neutralized somewhat either by the ash
from a recent burn or the addition of lime. In
addition, due to their physical properties, these
soils have very low water retention capacities.

Thus, a few days without rain with high air
temperature and intense solar radiation can
cause most plants to suffer from internal water
stress. Only moderate water stress for a short
duration is needed during a plant's critical
growth stage to severely reduce yields.
It will be demonstrated later in this re-
port that certain crop and soil management
practices can reduce plant water stress; these
practices include 1) deep placement of lime
and fertilizer, 2) mulching, and 3) proper plant-
ing date. In terms of excessive rainfall or
moisture, such crop and soil management prac-
tices as 1) larger doses of N and K with two to
three split applications to compensate for
leaching and/or the low solar radiation effect,
2) use of systemic (and not contact) herbicides,
3) proper planting date, and 4) crop species
and variety selection can help overcome the
detrimental effects of excessive rainfall.

3.2 ANNUAL CROP VARIETAL
ADAPTATION EXPERIMENTS
D. E. Bandy
This program essentially started in 1975
with the testing of 16 maize varieties from the
National Corn Program in Peru and 22 varieties
from CIMMYT (Annual Report 1976-1977).
The program has since expanded to include
soybeans, cowpeas, rice, mung bean and sugar
cane. The main cooperators or germplasm
sources are the Peruvian National Programs for
corn, rice, soybeans and sugar cane, IITA,
ICRISAT, AVRDC and INTSOY.
The approach of this program is a logical
continuation of the soil management program.

79

Due to genetic variability, certain species and
cultivars will be more adapted to the udic soil
moisture regime than others. In some cases, the
environment can be modified to meet the
plants' needs, e.g., soil management, agronomic
practices, etc. In other instances, it is almost
impossible to change the environment, e.g., air
temperature, solar radiation, etc. For these rea-
sons it is an economic necessity to find the best
crop species and cultivars which are adapted to
the existing environment and/or require the
least amount of environmental change or in-
puts. This is especially true for the resource-
limited small farmer of the humid tropics.
Objectives
The program is not designed to do large-
scale germplasm screening; that is what the
cooperating national and international insti-
tutes are doing. We accept lines or cultivars
which the national and international institutes
believe have the most potential for the humid
tropics. Our objectives are more site specific.
The exact objective for each crop might be
different but overall the program looks for cul-
tivars which have. (1) the highest yield poten-
tial for either monocultural or polycultural sys-
tems, for high input or low input systems (2)
tolerance to adverse soil conditions, e.g.,
acidity or Al toxicity, (3) disease and insect
resistance (4) in the case of legumes, N fixation
and residual or carryover capabilities (5) adap-
tability to small farmer's conditions and (6)
consumer acceptance.

3.2A MAIZE (Zea mays)
The maize adaptation trials are a
continuation of Ing. Jose Benites' work in

cooperation with the Corn Program of the
National Agraria University in Peru (Annual
Report, 1976-1977).
The objectives are the same as when the
trials began in 1976 and are:
1. Selection of varieties with most pro-
mising yield potential for maximum input and
low input systems.
2. Tolerance to the prevalent maize
diseases and insects in the area; leaf blight
(Helminthosporium sp.), kernel dry rot
(Diplodia sp.), and European corn borer
(Ostrinia nuveladis).
3. Short statured plants for better resist-
tance to lodging.
4. Selection of populations or composites
(and not hybrids) so the small farmer can select
and produce his own seed. (Hybrid seed pro-
duction facilities are non-existent in the selva
of Peru.)
Methodology
The experiments consisted of 20 tropical
varieties selected by the National Corn Program
from its own germplasm bank and from
CIMMYT's tropical maize selections. The local
check was Mez. Amarillo Planta Baja selected
by Benites in the 1976 trials but with seed pro-
duced and selected in Yurimaguas. Two fer-
tility levels were used.
Trial No. 1 had fertility levels of 120-35-
66-18 kg/ha of N-P-K-Mg respectively, and
3-3-2 kg/ha of Zn, Cu and B respectively. Only
500 kg/ha of lime was applied since the area
was previously cropped and still had a residual
lime effect. All of the lime, P, Mg, Cu, B and
30% of the K were incorporated into the soil
before planting. N and remaining K were split

Performance of 21 maize genotypes in a uniform trial at Yurimaguas, 1978.

Dry Rot
Entry Grain Stover Harvest Ears Unfilled Infected Plant
Cultivar No. Yield Yield Index* Harvested Ears Ear Sterility Ht.

t/ha t/ha No. No. No. % cm

Mez. Am. P. Baja 1 1.99 3.38 34 21 3 8 34 212
PMC 747 2 2.61 4.61 33 20 2 8 50 225
SA 12 3 1.21 4.61 20 19 4 10 59 175
Tuxp. Br. Blanco 4 0.92 3.87 17 17 9 4 65 210
Lote 8 5 2.00 5.37 26 18 1 4 30 205
Pob. I Am. Dent 6 2.86 6.47 29 26 6 5 26 225
Pob. II Am. Dent Selec. Flint 7 2.19 5.64 26 22 3 5 59 178
PM 701 Hib. Trop. 8 1.80 5.03 24 26 6 8 29 175
Pob. III Comp. Mez. A.P.B. 9 2.53 3.38 39 23 2 7 29 205
(C10xC7) x (108.90) 10 1.49 4.64 23 25 4 13 37 203
(C xC 1) x (108.90) 11 1.74 4.68 25 20 2 9 63 198
(C1 C ) x (108.90) 12 2.09 6.92 22 26 5 9 19 240
PM 2A 13 1.90 4.35 28 24 9 9 51 203
PM 210 14 1.34 5.67 17 29 9 13 63 178
PM 212 15 1.47 4.62 22 22 8 7 40 238
PMS 264 16 1.64 6.27 20 25 8 7 17 218
PMC 2 17 2.15 3.80 34 22 4 6 62 178
POEY T 66 18 1.79 5.31 24 18 3 8 45 200
PMC 5 19 1.17 5.86 15 21 9 5 16 218
PMC 6 20 1.38 5.31 19 21 7 3 62 178
Am. P. Baja (Local check) 21 2.31 3.89 34 28 4 6 27 195

00
0

Grain wt.
* Harvest Index = grain yield divided by total amount of plant dry matter production eg. for corn: Gai 100= H.I.
Grain+stover+cob"

Table 3.2A:1.

81

applied in two equal portions at 15 days and
30 days after planting. The trials were planted
April 16, 1978 and harvested after 114 days.
At harvest time, the soil for trial No. 1 had the
following chemical characteristics: pH 5.2,
0.80 and 3.03 meq/100 cc of exchangeable Al
and Ca+Mg, respectively, 21% Al saturation
and 15 ppm P.
Trial No. 2 had no lime or fertilizer ap-
plied. It was planted on a two-year old
abandoned field which had received a low level
of nutrients at that time. The soil had the
following chemical characteristics at harvest:
pH 4.6, 2.40 and 1.65 meq/100 cc of exch. Al
and Ca+Mg, respectively, 59% Al saturation
and 7 ppm P.
Results and Discussion
Various growth and yield characteristics
from the trial which received lime and ferti-
lizers are presented in Table 3.2A:1. Two of
the three varieties that outyielded the local
check were composites; one of them (entry 9)
had Mez. Amarillo Planta Baja, which is the
check, as part of its pedigree. In addition to
being the highest grain yielder, Pob. I Am.
Dent (entry 6) also had low plant sterility
(26%) and less problems with kernel dry rot,
Diplodia sp. Its main drawback is its plant
height which could lead to serious lodging
problems.
PMB 747 (entry 2), which has been
designated the corn hybrid for the selva of Peru
by the National Corn Program, demonstrated
good yield potential, good photosynthate dis-
tribution and excellent ear filling qualities. Its
main drawbacks are the high number of plants
without ears (50%) and plant height.

Previous experience at Yurimaguas has
shown that plants over 2 m tall have a serious
tendency to lodge. Eight of the 21 cultivars
tested were less than 2 m tall. The three com-
posites or populations with Am. Planta Baja
(entries 1, 9, 21) were among the most physio-
logically adapted to the environment in terms
of photosynthate distribution, indicated by the
harvest index (Fig. 3.2A:1).
PM-701 or Hibrido Tropical (entry 8) pro-
duced 20% less grain than the check even
though it displayed good growth character-
istics. The main yield limiting factor was pro-
bably its very poor harvest index.
It is interesting to note the yield dif-
ferences between entries 1 and 21. They are
both Am. Planta Baja composites but entry 1
had seed produced under Peruvian coastal con-
ditions, while entry 21 (local check) had seed
produced and selected under the environmental
conditions at Yurimaguas.
As had been demonstrated in the 1976
results, Tuxpeno Braquitico Blanco (entry 4)
was the lowest grain yielder, had an extremely
low harvest index, and had a very high number
of plants without ears (65%).
Yield results and responses to lime and
fertilizers for the 21 corn cultivars are shown in
Figs. 3.2A:2 and 3.2A:3, respectively. As
would be expected, almost all of the entries
gave over 100% increases in grain production
when limed and fertilized. In selecting cultivars
for response to maximum input systems, en-
tries 2, 6, 9, 17 and 21 showed the most pro-
mising potential. They gave the best yields,
showed good response to applied soil amend-
ments and had the best harvest indices (Fig.
3.2A:1).

82

40 With Fert.and Lime

35 W/0 Fert. and Lime

30

S25 -

w 15 -
20

5

0
S2 3 4 5 6 8 9 I 11121314151617 18192021

CORN CULTIVAR ENTRY NO.

Figure 3.2A: 1. Harvest Index for 21 corn varieties in relation to the application of lime and fertilizer.
Yurimaguas, Peru, 1978.

[ ]With Fert. and Lime
W/O Fert. and Lime

3.0
3.0--------------------- 20% Increase n
3 2.5
S------------ Equiv. Yield ,
S2.0,
o -- --20% Decrease 0
0
w 1.5 o

F-

w
0 5

23 4 56 7 8 9O11 12 1314 151617 1819 2021
_J
w
CORN CULTIVAR ENTRY NO.

Figure 3.2A:2. Grain yield results from 21 corn varieties in relation to the application of lime and
fertilizer. Yurimaguas, Peru, 1978.

83

a:
w
IL

i,
O

0
a-
(I)

O0
0J

w
>-

CORN CULTIVAR ENTRY NO.

Fig. 3.2A:3. Yield response (%) of 21 corn varieties to lime and fertilizer.
Yurimaguas, Peru, 1978. *No yield was realized without lime
and fertilizer.

500

300

250

200

150

100

50

0

10

I 2 3 4 5 6 7 8 9 10 11 12 1314 15 1617 18 19 2021

,i

u)
C-
z

cr_
0)

0

84

Table 3.2A:2.

Climatic data for the 28 day period (June 7-July 4)
coinciding with the pollination and silking periods for
maize. Yurimaguas, 1978.

Max. Precip- PAN Relative Solar Soil
Temp. itation ET Humidity Radiation Temp.

0C mm mm % cal/cm2/day 0C

31.6 74.5 78.4 33 325 34.2

Photo 3.2B 1. Soybean varietal adaptation trial, 1979, Yurimaguas, Peru.

85

On the other hand, if cultivars are being
selected for low input systems, the criteria are
not maximum yields but their tolerance to
acid, Al saturated, low base status soils which
are characteristic of the Ultisols in the region.
In addition, the cultivars should be ones which
are physiologically most efficient. Entries 5, 8
and 16 fit those criteria.
Fortunately, in addition to hybrids there
are population and composite entries surviving
the selection criteria, thus they can quickly
adapt to the small farmer's needs. Cultivars
completely unadapted were entries 3, 4, 10,
11, 14, 15, 18, 19 and 20.
It should be noted that the low overall
yields were probably due to climatic factors. A
28-day hot, dry period coincided with the pol-
lination stage (Table 3.2A:2). The pollination
period usually lasts for 14 days or more but if
the climate is hot and dry pollination could be
shortened to only a few days. A delay in silking
can also be induced by moisture stress. Thus, a
shortened pollination period with a delay in
silking could account for the unusually high
percent of plant sterility. During the three-
week period following silking kernel size and
number are determined. The large number of
unfilled ears for some of the cultivars could
also be related to this low rainfall period.
Benites also showed significantly reduced corn
yields for an April planting with rainfall being
one of the main causal factors.

3.2B SOYBEAN (Glycine max)
Soybeans were first introduced to the
Yurimaguas area by the Tropical Soils Program

sequence of rice-corn-soybeans. The soybean
has shown good adaptability to the area and
has a grain yield average of 2.5-3.5 t/ha when
optimum soil moisture and fertility require-
ments were met. Improved Pelican, National,
and Jupiter have been the three main cultivars
grown in the past. Of the three varieties,
Jupiter has been shown to be the most tolerant
to aluminum toxicity and low base status soils.
Improved Pelican has yielded well, but does
not have a strong determinate growth habit,
thus maturity of the pods are not uniform.
This makes harvesting difficult and subjects the
pods and seed to fungal attack.
In 1978, the program introduced soybean
production to the neighboring farmers. They
have readily adopted the soybean for both
home consumption and the local market. Soy
milk, flour, and cheese have become very
popular and are in constant demand. In ad-
dition, the National Marketing Board, ENCI,
has a standing order to buy all the soybeans
produced in the area for oil extraction
processing.
Objectives
1) Select soybean cultivars that have more
tolerance to leaf, pod, and seed fungal diseases
which at present prohibit the growing of soy-
beans during a prolonged rainy period. The
diseases frog eye spot (Cercospora sojina), pod
and stem blight (Diaporthe phaseolorum var.
sojae) and purple stain (Cersospora kikuchii)
can reduce yields and/or reduce seed quality
drastically when cloudy, humid conditions
exist during the pod filling stage.
2) To select soybean genotypes which

Performance of 16 soybean cultivars in a uniform trial at Yurimaguas, Peru, 1979.

100
Entry Grain Seed Seed1 2 Shat- Nodule'/ Plant Pod Pods 50% Growth
No. Variety Yield Wt. Quality Lodging- tering- Abundance Activity Ht. Ht. Plant Flowering Duration Germ.

4 Pod
t/ha g Wks Fill % cm cm # Days Days %
-------------- 1 5 scale --------
1 CH-3 1.42 16.1 3 4 2 3.2 3.3 58 106 15 49 37 102 95
2 UFV-1 1.70 16.5 3 1 1 3.6 3.4 66 60 7 55 37 90 94
3 SJ-2 1.90 15.1 3 4 2 3.4 2.3 49 94 13 55 41 90 75
4 Hardee LS 1.94 15.0 2 1 1 2.9 3.1 45 92 10 91 45 124 22
5 Orba 0.87 11.4 4 4 4 3.3 2.9 8 68 13 52 35 77 84
6 IAC-2 1.66 18.4 3 4 2 3.4 2.6 80 94 12 88 37 91 95
7 Tunia 1.68 18.2 3 1 3 2.5 2.8 60 77 12 73 35 95 56
8 Caribe 0.63 9.3 3 4 3 2.9 2.9 24 110 11 89 38 124 60
9 Jupiter 1.33 15.5 3 2 1 3.7 3.1 34 87 15 58 46 92 79
10 Improved
Pelican 1.88 15.5 2 3 2 2.1 1.8 75 89 13 72 38 85 80
11 Acc.2120 1.93 7.0 1 4 1 3.2 3.2 23 119 12 72 47 83 89
12 Rillito 0.98 17.4 5 1 2 3.8 3.9 39 63 8 48 33 90 52
13 Williams 1.33 18.6 4 1 2 3.2 4.0 14 62 10 50 31 78 43
14 Davis 2.00 17.8 2 1 2 3.3 3.3 40 62 8 63 31 84 52
15 Bossier 1.28 15.5 3 1 1 3.6 3.1 19 75 9 56 37 91 73
16 Ransom 1.21 20.0 3 1 2 3.1 3.5 15 55 6 56 31 98 55

1/ Seed Quality Scale: (1) Very good, (2) good,
2/ Lodging Scale: (1) >95% erect, (2) 75-95% er
3/ Shattering Scale: (1) No shattering, (2) <10:
(5) >50% shattered.
4/ Nodule Scale: (1) Many nodules on main root;
few nodules on lateral roots, (3) few nodules
on main root; few nodules on lateral roots, (

(3) fair, ('
ect, (3) 50-
% shattered,

4) poor, (5) very poor.
75% erect, (4) 25-50% erect (5) <25% erect.
(3) 10-25% shattered, (4) 25-50% shattered,

many nodules on lateral roots, (2) many nodules on main root;
on main root; many nodules on lateral roots, (4) few nodules
5) no nodules on main root; no nodules on lateral roots.

o0
0)

Table 3.2B:1.

87

have good seed viability or storability. Cli-
matic conditions are such in Yurimaguas that
it is impossible to keep soybean seed viable
from one planting season to the next.
3) Nodule activity and abundance in re-
lation to the native soil conditions is very
important for the development of low input
systems utilizing soybeans. Good nodulation
and persistence would not only eliminate the
necessity of N fertilizer for the soybean crop
itself but also eliminate or reduce N fertiliza-
tion requirements for the crop to follow.
4) Yield potential and lodging and shat-
tering resistance are standard essential
characteristics for the selected cultivar(s).
Methodology
In cooperation with INTSOY a uniform
adaptation trial was established in September
1979. The trial consisted of 16 soybean geno-
types (selected for the tropics by INTSOY)
arranged in a randomized complete block
design with four replications. The experiment
was conducted in an area that was previously
planted to soybeans and peanuts. The soil had
the following chemical characteristics: pH
5.5, 0.10 and 3.44 meq/100 cc of exch. Al
and Ca+Mg, 3% Al saturation and 15-20 ppm
P. Fertilizer N-P-K-Mg rates of 25-25-25-18
kg/ha were broadcast and incorporated into
the soil before planting. Potassium was split-
applied in 3 equal portions: before planting,
15 days and 30 days after planting. A granular
inoculant was applied in the seed row during
planting.
Results and Discussion
The performance of a number of cul-
tivars were equal or superior to the local

check Jupiter (Table 3.2B:1), with five culti-
vars outyielding Jupiter by more than 40%
(Fig. 3.2B:1). In relation to yield and growth
parameters, the cultivars Davis, Improved
Pelican, and Hardee performed very well. SJ-2
and Acc. 2120 also demonstrated excellent
yield potential, but had serious lodging pro-
blems. Nodulation abundance criteria showed
the cultivars CH-3, SJ-2, Hardee, Orba, IAC-2,
Caribe, Tunia, Improved Pelican, Davis and
Acc. 2120 to have superior nodulation on
either the main root or the lateral roots. All
of these varieties outyielded Jupiter except
Caribe and Orba. Their reduced yields were
due to serious problems with lodging and pod
shattering.
Cultivars with the most nodule activity
during late pod filling stage were CH-3,
UFV-1, IAC-2, Tunia and Improved Pelican.
Seed viability tested 60 days after har-
vest showed the following cultivars to have at
least 80% germination: CH-3, UFV-1, Orba,
IAC-2 and Improved Pelican and Acc. 2120,
Jupiter had 79% germination.
Conclusion
The only cultivar to meet the selection
criteria in all four categories (yield, nodula-
tion abundance, nodulation activity, and seed
viability) was Improved Pelican. Thus, it is
adaptable and would give an acceptable per-
formance in both high input and low input
systems. In the selection of cultivars for high
input systems, only Hardee and Davis were
superior; they demonstrated both high yield
potential and excellent plant growth charac-
teristics. For low input systems, where good
nodule abundance and late season activity are

88

2.00

1,80

1.60

1.40

1.20

1.00

0.80

0.60

0.40

0.20

0

?>
ILL

1,
-)

w
w
0
4

4
0
0

C4
I
0

4
z
I-

w

o
M
cc

I-

-j
w
0.
q:
z

0
N
F4

<

I-

-J
-J

u)

-1

U)

w

C,
0
0

La
V')

LU
(C)
z

0
-J
LJ

2
0
U)
z
4

SOYBEAN CULTIVAR

Fig. 3.2B:1. Yield performance of 16 soybean cultivars in a uniform trial
at Yurimaguas, Peru, 1979. Jupiter is the local check.

^-

w
5I

z
Z

0,
w

0.

-
L
h-

89

important for residual-N effect, the cultivars
Tunia and Improved Pelican were superior.
The second stage of the varietal selection
process is currently being conducted. Two
weeks after the soybeans were harvested corn
was planted, without any applied N, to test
the hypothesis that there is a difference be-
tween the soybean cultivars in the amount of
residual-N available to the following corn
crop.
The third stage of the experiment will be
to quantify aluminum tolerance and nutrient
requirements for the most promising soybean
cultivars.

3.2C RICE (Oryza sativa)
The principal limiting factor to upland
rice production at Yurimaguas is blast,
Pyricularia oryza. No matter how well adap-
ted a rice variety may be to upland soil
conditions, if it doesn't have excellent
horizontal resistance to blast, it will fail in
Yurimaguas. For the last 10 years, hundreds
of rice lines have been tested yearly at Yuri-
maguas in cooperation with the National Rice
Program in Peru. Only one introduction,
IR 4-2, has survived for more than two years.
This variety was introduced into the area by
Dr. Sanchez and his Peruvian co-workers in
1969 and has continued to perform well.
Blast infestation on IR 4-2 seldom passed the
2% mark even when severe water stress
conditions existed. When potassium defi-
ciency and/or soil moisture was limiting,
IR 4-2 has been severely attacked by brown
leaf spot, Helminthisporium oryza. Under
optimum nutrient and soil moisture

conditions IR 4-2 has yielded 4.5 t/ha on the
Ultisols of Yurimaguas.
The problem with IR 4-2 is acceptance.
The farmer does not like it because its short
stature (70-80 cm) and upright leaves create
more problems with weed control. The rice
mills do not want to receive it since its small
grain size creates milling problems and it also
has a low milling percentage (62%). The con-
sumer doesn't like it because of its small grain
size and "white belly" (high amylose con-
tent). Rice breeders at IITA have had some
success in breeding for blast resistance using
African germplasm sources. In August 1979,
an introductory rice nursery consisting of
nine upland rice lines was planted at Yuri-
maguas on an alluvial soil beside the Shanusi
River. The semi-dwarf IR 4-2 and Carolina,
the traditional tall leafy upland rice variety
commonly grown in the area, were the two
checks used. No fertilizer or lime was applied
since this was the first crop after the slash and
burn system of land clearing.
All nine introductions from IITA show-
ed good to excellent resistance to blast (Table
3.2C:1). Entry 1 had 100% resistance to blast,
displayed good yield potential, no lodging
problems, short growth duration, and no
brown leaf spot (H. oryza) or leaf scald (R.
oryza) symptoms. In addition, grain quality
for TOX 340-1-1-1-1 (entry 1) was superior to
IR 4-2 in that grain size is larger and it does
not have "white belly." Entries 4 and 9 also
showed good yield potential and disease re-
sistance. The low yields for entries 5 and 8
were due to plant sterility.
The need for an improved upland rice

Table 3.2C:1. Performance of 9 rice genotypes and two local checks in a uniform trial at Yurimaguas, 1979.

Entry Grain Blasti/ Growth Helm. Rhyn.
Cultivar No. Yield Resistance Lodging- Duration oryza oryza

kg/ha 1-10 scale 1-5 scale Days Infestation -

TOX 340-1-1-1-1 1 3447 1 1 110 No No
TOX 737-1-1 2 ---
TOX 475-1-2-1 3 2504 4 1 115 No Yes
TOX 86-1-1 1 4 2454 4 1 115 No No
TOX 515-22-107-1-1 5 1190 2 1 120 Yes Yes
TOX 516-28-10-103-5-5 6 -
TOX 515-38-101-1-1-7 7 1481 3 1 120 Yes Yes
TOX 95-8-1-1-LS3 8 550 4 1 100 No No
TOX 514-16-1-1-1 9 2838 3 1 130 No No
IR 42 (Local) 10 3594 4 1 130 Yes Yes
Carolina (Local) 11 2281 2 2 140 No Yes
Carolina (Local) 12 1471 2 4 140 No Yes

* Did not germinate.
1/ Blast resistance scale: 0, 1, 2 = very resistant; 3 = moderately resistant; 4 = moderately
susceptible; 6 = 25% susceptible; 7 = 50% susceptible; 8 = 95% susceptible; 10 100% susceptible.

75-95% erect, (3) 50-75% erect; (4) 25-50% erect (5)<25% erect.

(0
0

2/ Lodging scale: (1)>95% erect, (2)

91

variety is evidenced by Carolina's performance.
Not only does it not have the genetic potential
to produce high yields, its plant height (150
cms) favors lodging which can further decrease
yields by 50%. The average yields for farmers
using Carolina are 1.0 to 1.5 t/ha.
The IITA selections were replanted in
January, 1980 to test their yield potential and
disease resistance during the rainy season. In
addition, 25 more upland rice selections from
I ITA were introduced for the first time.

3.2D SUGARCANE (Saccharum officinarum)
Sugarcane is primarily a tropical crop that is
better adapted to ustic soil moisture regimes in
terms of harvesting because the dry period helps
to physiologically mature the plant. The cultiva-
tion of sugarcane in Yurimaguas is nothing new.
During the 1940's, semi-refined sugar was
exported from Yurimaguas to Europe and the
U.S., along with latex and rotenone from the
plantations of rubber (H. brasiliensis), and
barbasco (Lonchocarpus nicau). During that
time the sugarcane clone POJ 2878 (cross made
in 1921) was introduced into the area and is
still the major clone cultivated. This same clone
is the check in this experiment.
The Peruvian government has a strong desire
to study the possibilities of expanding the sugar-
cane industry from the coast to the selva of
Peru.
Methodology
A sugarcane adaptation trial was initiated in
September 1978 in cooperation with Ing.
Hernan Tello and the Instituto Central de Inves-
tigaci6n Azucaral. The trial consists of 18 entries
selected by the Sugarcane Institute plus one

local check (POJ 2878) planted in a randomized
complete block experimental design with four
replications and a plot size of 12 m x 6 m. The
trial was planted on a 20-year old abandoned
pasture field. Samples for soil analysis were
taken after the crop was harvested in November
1979 but were not yet analyzed at time of
writing.
Only 500 kg/ha of lime and 22 kg/ha of P
were applied since one of the first basic objec-
tives of the trial was to study the clonal response
to native soil conditions. The lime applied was
enough to supply Ca as a nutrient but not to
markedly change percent Al saturation or soil
pH.
Results and Discussion
Some growth and yield parameters after a
13-month growing period are shown in Table
3.2D:1. Seven cultivars produced over 50 t/ha of
cane which is a fairly good yield considering the
very low amount of fertilizers applied. All cul-
tivars except one showed good to excellent
disease and insect resistance. Only one applica-
tion of an insecticide was used to control sugar-
cane borer (Diatraca saccharalis).
In the selection of adapted cultivars, cane
and juice quality (Table 3.2D:2) are just as
important as cane yield. The varieties which pre-
sent the best cane and juice qualities in order of
merit are: NCO-310, CAC 57-11, NCO-419, and
PHIL 56-95. NCO-310 is the best of the 19
varieties because it has the following character-
istics: Low fiber content, excellent Pol, low per-
centage of reductors, very good Brix, optimum
purity, and low cane moisture at harvest. The
other three varieties are also very good but their
percent reductors are a little higher. The next

Growth and yield results for 19 sugar cane cultivars grown at Yurimaguas, 1978-1979.

Entry Fresh Fresh Plant Stalks2 Flowering Insectl/ Diseasel/
Cultivar No. Cane Wt. Leaf Wt. Ht. per 3Qm @ 9 months Resistance Resistance

CO 622
H32 8560
Q88
PHIL 56 95
H57 5174
H37- 1933
H39 5803
PHIL 53 33
POJ 2878
NCO 310
NCO 419
CAC 57 11
849 119
PR 980
H50 7209
L60 25
LAR52 604
B47 161
H50 2036

No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19

t/ha
52.8
35.1
46.3
43.3
44.5
51.0
57.4
39.6
29.6
49.7
40.6
44.6
38.4
57.8
43.6
51.9
41.7
59.3
56.3

t/ha
65.5
30.8
34.1
31.1
26.6
24.0
57.5
40.3
22.8
65.0
18.8
47.0
33.6
37.5
21.3
31.8
22.4
32.8
32.0

cm
269
241
224
255
304
262
238
225
237
231
239
234
234
275
238
253
230
290
301

No.
205
112
126
151
103
107
116
194
97
192
108
210
163
137
97
139
108
131
159

%
6
50
8
2
7
2
0
76
3
77
0
0
64
0
4
5
2
0
38

----- 1-10 scale -----
8 7
7 9
8 7
8 7
9 8
10 9
9 9
10 9
8 7
9 9
9 9
9 8
8 7
8 7
9 8
8 6
9 9
9 8
8 7

1/ Rating Scale: 9-10, excellent; 7-8, good; 5-6, average; 3-4, poor; 1-2, very poor.

Co
NJ

Table 3.2D:1

93

Table 3.2D:2.

Cane and juice quality for 19 sugar cane varieties grown at
at Yurimaguas, 1978, 1979.

Cane Juice Cane
Fiber Reduc-
Cultivar Content Pol* tors** Brix+ Purity Pol Rec++ Moisture

CO-622
H32-8560
Q88
PHIL56-95
H57-5174
H37-1933
H39-5803
PHIL-53-33
POJ-2878
NCO-310
NCO-419
CAC57-11
B49-119
PR-980
H50-7209
L60-25
LAR52-604
B47-161
H50-2036

16.73
15.08
15.23
18.42
16.87
16.38
14.95
15.36
13.05
13.63
16.80
15.82
15.40
16.38
16.73
13.56
15.43
16.73
15.09

14.03
13.69
13.05
15.21
12.72
14.97
12.86
14.45
12.97
15.67
15.12
15.60
13.68
14.04
13.26
11.97
13.48
11.33
13.76

0.719
0.927
0.937
0.825
0.949
0.924
0.971
1.003
1.027
0.776
0.866
0.905
1.060
0.910
0.963
1.214
1.472
0.976
0.937

19.53
19.91
18.83
21.29
18.07
21.18
18.16
20.12
18.91
20.79
21.26
21.35
19.14
19.75
18.81
17.39
20.17
16.53
19.09

86.27
80.99
81.76
87.59
84.70
84.55
83.29
84.87
78.89
87.26
85.49
86.80
84.50
85.05
84.68
79.64
79.01
82.32
84.91

12.53
12.22
11.62
13.60
11.26
13.42
11.47
12.98
11.60
14.26
13.56
14.08
12.24
12.54
11.78
10.65
11.97
9.92
12.34

67.0
68.0
68.8
64.2
68.0
65.8
69.6
67.6
70.5
68.3
65.5
66.1
68.4
67.0
67.6
71.4
67.5
69.5
68.7

* Pol = Polariscropic reading of t

he amount of sucrose.

** Reductors = amount (%) of reducing sugars.
+ Brix = Hydrometer reading of the total amount of soluble solids in
the juice.
++ Pol Rec = amount (%) of sucrose reclaimable or convertible to 960 raw
sugar.

% -----

%

----- -------

Table 3.2E:1. Performance of 21 indeterminate and one determinate cowpea genotypes in a uniform trial at Yurimaguas, 1979.

Genotypes Entry Grain Harvest Pods/ Pod-/ Seed-/ 100 Seed Plant Plant 50% Disease-/ Nodules-
Entry No. Yield Index Plant Quality Quality Seed Wt /Pod Ht. Width Flowering Resistance @ 4 wks.

kg/ha No. -- 1-5 scale -- g No. cm cm Days ---- 1-5 scale ----
VITA-3 1 435 82 7 1.5 3 14.3 8.1 58 60 47 2 3.9
VITA-4 2 237 78 6 1.8 2 8.5 9.6 63 64 49 2 3.4
VITA-5 3 573 82 16 2.3 2 9.1 7.4 52 51 45 3 3.6
TVX 66-2H 4 316 86 7 1.5 2 10.1 7.6 65 55 48 3 3.9
TVX 289-4G 5 333 76 10 1.6 3 10.0 7.4 57 68 46 1 3.5
Ife Brown 6 484 85 14 2.3 3 12.0 5.4 54 61 44 3 3.2
TVX 33-1J 7 548 76 12 2.2 2 9.2 8.7 68 56 48 2 3.3
TVX 1850-01E 8 568 82 10 2.2 2 11.6 8.4 64 66 42 2 3.9
TVX 1948-01E 9 479 78 9 2.0 2 9.7 8.9 77 51 49 3 3.8
TVX 1952-01E 10 442 80 9 2.2 2 10.3 7.7 59 59 46 2 4.0
TVX 1999-02E 11 443 84 11 1.7 2 9.4 7.5 60 58 48 2 3.2
TVX 1999-01F 12 613 87 12 1.9 2 10.5 9.1 62 58 48 2 3.3
TVX 2907-02D 13 672 78 18 2.4 2 12.3 5.7 57 57 46 3 3.4
TVX 2912-011D 14 391 78 10 1.8 2 12.4 7.0 63 59 47 1 3.5
TVX 2939-09D 15 476 82 11 1.8 2 9.9 8.2 63 51 48 2 3.4
TVX 2949-01D 16 469 79 11 2.1 2 13.1 5.6 65 61 44 2 3.2
TVX 2949-03D 17 458 81 9 1.6 3 10.2 9.3 69 64 45 2 3.1
TVX 3048-02D 18 442 85 11 1.8 2 9.4 8.0 73 65 46 2 3.5
TVX 3218-02D 19 514 83 17 1.6 3 7.4 8.4 63 64 46 2 4.0
Local Check 20 370 79 10 1.5 3 10.7 6.7 68 69 45 1 3.2
Mean 463 81 11 1.9 2 10.5 7.7 63 60 46 2 3.5

1/ Pod Quality: (1) very good, (2) good, (3) fair, (4) poor, (5) very poor.
2/ Seed Quality: (1) very good, (2) good, (3) fair, (4) poor, (5) very poor.
3/ Disease Resistance: (1) highly resistant, (2) moderately resistant, (3) intermediate, (4) moderately
susceptible, (5) highly susceptible.
4/ Nodule Scale: (1) Many nodules on main root; many nodules on lateral roots, (2) many nodules on main
root; few nodules on lateral roots, (3) few nodules on main root; many nodules on lateral roots, (4)
few nodules on main root; few nodules on lateral roots, (5) no nodules on main root; no nodules on
lateral roots.

95

four acceptable cultivars in order of merit are:
H37-1933, PHIL 53-33, PR-980, and CO-622.
Their principal limitation is high percent re-
ductors. The following six cultivars which pre-
sented inferior qualities are: H50-2036, H32-
8560, B49-119, LAR 52-604, H50-7209, and
Q88. These varieties have a Pol between
13-13.8%, high percent reductors, Brix of 20 or
less and juice purity of only 79-84%. The re-
maining five cultivars were judged to be com-
pletely unacceptable: POJ-2878, H39-5803,
H57-5174, L60-25, and B47-161.
Conclusions
The 50-year old check, POJ-2878, which is
the predominate cultivar grown in the area, not
only gave very low cane tonnage but ranked 16
out of the 19 cultivars tested in terms of cane
and juice quality.
Some of the newer materials have performed
extremely well and have demonstrated very
clearly that the potential for sugarcane produc-
tion in the selva of Peru is very promising. The
second stage in the testing of the sugarcane
material is to study their performance under
optimum soil fertility conditions. The first
ratoon crop has been fertilized at rates suffi-
ciently high enough to give maximum yields.
When these results are in, a preliminary choice
of cultivars can be made as to which cultivars are
best suited for high input and/or low input
systems.
Sugarcane farmers from the surrounding area
who have seen the trial and its results are very
anxious to change their cane plantings to some
of the newer, more promising clones.

3.2E COWPEA (Vigna unguiculata)
Cowpea is very important for the humid
tropics as a source of plant protein since

common bean (P. vulgaris) is not well adapted to
this ecological zone. In addition, cowpea is
already an accepted basic food staple and has a
readily available local and national market. For
example, consumer demand has given cowpea a
market value five to six times greater than that
for corn.
Methodology
Twenty-eight lines, nineteen indeterminate,
semi-erect or spreading, and nine determinate
lines were compared to the local determinate
cultivar. The experiments are a component of
the International Cowpea Trials of IITA in
cooperation with Dr. P. R. Goldsworthy. A ran-
domized complete block experimental design
with four replications was used. Each plot con-
sisted of four rows, 0.4 m x 4.0 m. The experi-
ments were established in an area that was pre-
viously planted to soybeans and peanuts and had
the following soil chemical characteristics:
pH 5.5, 3.44 and 0.10 meq/100 cc of Ca + Mg
and exch. Al, respectively, 3% Al saturation and
15-20 ppm P. Fertilizers were applied at a
moderate rate: N-P-K-Mg rates of 0-26-75-18
kg/ha, respectively. K was split-applied in three
equal parts, before planting, 15 days and 30
days after planting. Neither nitrogen nor inocu-
lant was applied since the rhizobia for cowpea
nodulation is omnipresent in these soils.
Results and Discussion
The determinate and non-determinate cow-
pea lines were planted in two separate trials.
Results for the indeterminate lines are shown in
Table 3.2E:1. All of the yields were low due to
the late planting date (September 20, 1979),
thus the grain filling stage coincided with the
onset of the rainy season. Heavy infestations of
fungal diseases occurred on the leaves, pods and
seeds allowing only two harvests (four are usual)

	Front Cover
	Title Page
	Dedication
	Acknowledgement
	Personnel
	Table of Contents
	Introduction
	Cerrado of Brazil (Joint embrapa...
	Amazon jungle of Peru (Joint MAA...
	Extrapolation activities
	Intercropping
	Economic interpretation
	Communication of results

MILLISECOND	CLASS.METHOD	MESSAGE
0	sobekcm_page_globals.constructor
0	sobekcm_page_globals.constructor	Application State validated or built
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.constructor	Navigation Object created from URI query string
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.display_item	Retrieving item or group information
0	sobekcm_page_globals.get_entire_collection_hierarchy	Retrieving hierarchy information
0	sobekcm_assistant.get_entire_collection_hierarchy
0	cached_data_manager.retrieve_item_aggregation
0	cached_data_manager.retrieve_item_aggregation	Found item aggregation on local cache
0	item_aggregation_builder.get_item_aggregation	Found 'all' item aggregation in cache
0	system.web.ui.page.page_load (ufdc.page_load)
0	sobekcm_page_globals.constructor.on_page_load
0	html_echo_mainwriter.add_style_references	Adding style references to HTML
0	html_echo_mainwriter.add_text_to_page	Reading the text from the file and echoing back to the output stream
50	html_echo_mainwriter.add_text_to_page	Finished reading and writing the file