February 1957
UNIVERSITY OF FLORIDA
AGRICULTURAL EXPERIMENT STATIONS
JOSEPH R. BECKENBACH, Director
GAINESVILLE, FLORIDA
Design, Analysis and Results of an Experi
ment on Response of Pangolagrass and
Pensacola Bahiagrass to Time, Rate
and Source of Nitrogen
A. T. WALLACE, G. B. KILLINGER,
R. W. BLEDSOE, and D. B. DUNCAN
TECHNICAL BULLETIN
Single copies free to Florida residents upon request to
AGRICULTURAL EXPERIMENT STATION
GAINESVILLE, FLORIDA
Bulletin 581
CONTENTS
Page
IN TRODUCTION . ... ... ........ ... .. ..................... 3
EXPERIMENTAL PROCEDURE .... ........  ... . ................................ 3
STATISTICAL ANALYSIS .......... ............................. 4
A analysis by Y ears ....  .............. ........................ ....... ..... 4
Analysis Over All Years .............. .... ............. .... ....... ... 5
Special F Tests .......... .. ................. .................................... 5
Fitting and Testing of Trends ................. .................. ... ... 6
RESULTS AND DISCUSSION ..............~..................... 8
Phosphorus and Potash Levels ........ ............................ .. ... 8
Timesequence of Nitrogen Application ................ ... ................. 11
Source of Nitrogen ................... .. ..................................... 12
Rate of Nitrogen Application .... ................................ .. 14
Protein Analysis ........... ...... ............................. ...... 19
SUMMARY AND CONCLUSIONS .................................... ... 21
LITERATURE CITED ............... .............. ..... .. ...... .. 22
A PPEN DIX ........... ................ .......... ..... .. ... .. ... ... 24
Design, Analysis and Results of an Experi
ment on Response of Pangolagrass and
Pensacola Bahiagrass to Time, Rate
and Source of Nitrogen
A. T. WALLACE, G. B. KILLINGER,
R. W. BLEDSOE, and D. B. DUNCAN 1
INTRODUCTION
Pangolagrass (Digitaria decumbens) and Pensacola Bahia
grass (Paspalum notatum va.) were introduced into Florida dur
ing the 1930's and were released to farmers in the state in 1943.
These two relatively new pasture grasses immediately became
popular with cattlemen, resulting in a rapid increase in acreage.
By 1955 there were approximately 500,000 acres of Pangolagrass
and 750,000 acres of Pensacola Bahiagrass growing in the state.
Although several research workers (2, 7, 8) have conducted
fertility experiments with Pangolagrass and Pensacola Bahia
grass, the production capacity of these two grasses has never
been fully determined. It was with this in mind that a detailed
experiment involving rates of phosphorus and potassium and
rates, sources and time of application of nitrogen was planned.
EXPERIMENTAL PROCEDURE
The experiments were conducted on uniform fields of Pango
lagrass and Pensacola Bahiagrass at the Beef Research Unit
near Gainesville, Florida. When the experiments were started
in 1950 the sods were two years old. The soil type was Leon
fine sand. Uniform applications of dolomitic limestone, two tons
per acre, and fertilizer, 575 at 500 pounds per acre, were made
to the soil before the grasses were planted, and minor elements
were applied after the sods had become established.
The design used was a splitsplitsplitsplit plot repeated over
three years. A summary of the treatments and the notation
used for each is as follows:
SAssociate Agronomist, Agronomist, Associate Director, and Statisti
cian, respectively.
Design, Analysis and Results of an Experi
ment on Response of Pangolagrass and
Pensacola Bahiagrass to Time, Rate
and Source of Nitrogen
A. T. WALLACE, G. B. KILLINGER,
R. W. BLEDSOE, and D. B. DUNCAN 1
INTRODUCTION
Pangolagrass (Digitaria decumbens) and Pensacola Bahia
grass (Paspalum notatum va.) were introduced into Florida dur
ing the 1930's and were released to farmers in the state in 1943.
These two relatively new pasture grasses immediately became
popular with cattlemen, resulting in a rapid increase in acreage.
By 1955 there were approximately 500,000 acres of Pangolagrass
and 750,000 acres of Pensacola Bahiagrass growing in the state.
Although several research workers (2, 7, 8) have conducted
fertility experiments with Pangolagrass and Pensacola Bahia
grass, the production capacity of these two grasses has never
been fully determined. It was with this in mind that a detailed
experiment involving rates of phosphorus and potassium and
rates, sources and time of application of nitrogen was planned.
EXPERIMENTAL PROCEDURE
The experiments were conducted on uniform fields of Pango
lagrass and Pensacola Bahiagrass at the Beef Research Unit
near Gainesville, Florida. When the experiments were started
in 1950 the sods were two years old. The soil type was Leon
fine sand. Uniform applications of dolomitic limestone, two tons
per acre, and fertilizer, 575 at 500 pounds per acre, were made
to the soil before the grasses were planted, and minor elements
were applied after the sods had become established.
The design used was a splitsplitsplitsplit plot repeated over
three years. A summary of the treatments and the notation
used for each is as follows:
SAssociate Agronomist, Agronomist, Associate Director, and Statisti
cian, respectively.
Florida Agricultural Experiment Stations
Main plots 2 Rates of PK PK1 = 06060
PK, = 0120120
Sub plots 2 Sources of S = NaNO3
nitrogen S2 = (NHi)2 SOI
Subsub plots 5 Timesequence of T1 = March and June
nitrogen applica T2 = March and August
tion T3 = March, June and August
T, = March, June and October
T5 = January, March, May,
July, September and
November
Subsubsub plots 5 Rates of N R1 = 30 lbs. per acre per year
Rs = 60 lbs. per acre per year
Ra = 120 lbs. per acre per year
Ra = 240 lbs. per acre per year
R5 = 480 lbs. per acre per year
The phosphorus, as 20 percent superphosphate, and the pot
ash, as 60 percent muriate of potash, were applied broadcast in
January of each year.
The experiment was conducted for three years1950, 1951
and 1952. All plots were harvested at the same time, except in
1950, when the first harvest was restricted to certain plots
which had made exceptionally early growth. All plots were in
cluded in subsequent harvests. A sickle type mower was em
ployed.
Plot size was 5 x 15 feet. Eighteen inches were cut off both
ends of each plot and discarded at each harvest. Twelve feet
of plot remained for harvest, and a 36inch strip was cut length
wise on each plot. The remaining vegetation was mowed and
removed from the plots. When harvested, all grass had begun
to produce seedheads with an estimated 10 to 20 percent seed
head population.
All samples of grass were dried in a heated forced air drier
at 130 to 1400 F., and yields were expressed as weight of dry
material per acre. Total moisture remaining in samples was
between 5 and 6 percent.
In 1950 certain samples were saved for crude protein analysis.
Hereafter, Pangolagrass and Pensacola Bahiagrass will be
referred to as Pangola and Bahia, respectively.
STATISTICAL ANALYSIS
Analysis by Years.The type of analysis used with each
year's data is presented in Table 1 in the Appendix. The ob
served mean squares are denoted by M1, M, ... M21, and their
expectations (the "true" mean squares) are listed under the
Response of Pangola and Bahia Grass to Nitrogen 5
column "Expected M.S." in terms of the variance components2
involved. The expectations for the mean squares have been
computed with all effects considered fixed except those for repli
cations and errors (9).
The F ratio computed for each row in the analysis is selected,
tested and interpreted in the usual way. For example, the F
ratio for source of nitrogen is selected as MT/Mio, because the
expectation of M7 differs from that of Mo1 by the addition of
400 Vs only. If Vs, the variance component for sources, is zero,
Mr and M,o are estimating the same true variance, and the ratio
M,/MIo should have the probability distribution of an F ratio
with 1 and 14 degrees of freedom, respectively. Thus, the ratio
will give "significant" or "highly significant" evidence that Vs
is not zero, if it exceeds the 5 percent or 1 percent value of F
for 1 and 14 degrees of freedom, respectively.
Analysis Over All Years.In addition to the yearly an
alyses, a pooled analysis was made over all years for each grass.
Appendix Table 2 shows the type of analysis used. The addi
tional year effects included were assumed to be random. Thus
conclusions drawn from these analyses may be projected over
all years for which the years under consideration can be re
garded as a random sample.
Special F Tests.Because of the mixedmodel (presence of
both fixed and random effects), many of the F ratios have to be
obtained as the ratios of the sums of mean squares taken in
pairs. For example, to test nitrogen rates (i.e. the hypothesis
Vn = 0), it is necessary to take the ratio of M,, + M,7 to M,5
+ M,1 (the ratio of Ml to M1, + M1 M7a could have been taken
for the same purposeAnderson and Bancroft, 1, p. 350). The
important point is that the expectations of M11 + M_7 and MI.,
+ M17 differ only by the presence of 480 Vi in the former, and
are identical if VR = 0.
In testing ratios of this form, the effective degrees of free
dom for the numerators and denominators are obtained (1)
from
f (M. M,,) 2/ [ (Ma/f) (M (M,/f,)2]
where M,, and M, are the mean squares being added to give the
numerator or denominator, as the case may be, and fa and fb
Each variance component is denoted by the symbol, V, with an ap
propriate subscript, e.g. VR denotes the variance component for replica
tions. It should also be noted that the term, variance component, is being
used in the broad sense as a measure of variation between fixed effects as
well as between random effects.
Florida Agricultural Experiment Stations
are their respective degrees of freedom. To illustrate, for test
ing nitrogen rates, the term used in the denominator is M15 +
Mi7, as mentioned above. The effective degrees of freedom for
this denominator are
f = (M1i + M7)2/[(M1/132)2 + (M7/8)2]
which works out to be 98.
The effective degrees of freedom are used for entering the
F tables in the usual way.
Fitting and Testing of Trends.In cases in which the F tests
indicated the presence of appreciable differences, further analy
ses and tests were conducted. Where meaningful polynomial
trends could be fitted, this was done; otherwise multiple range
tests (4) were applied.
For instance, in the analysis over all years for Pangola, the
mean square for rates of nitrogen was highly significant, and
an analysis of variance of the polynomial trends was made as
indicated in the following table.
TABLE 1.ANALYSIS OF VARIANCE FOR TRENDS.
Source d.f. S.S. M.S. F
Rate of N application 4 SN Mn (M1l + M27)/(Mi, + Mi7)
Linear ........................... 1 SL Mul,r (Mii,L + M2,)/(MI, + M1)
Quadratic .................... 1 SQ MI,Q (Mn,o + M)/(M, + M)
Residual ....................... 2 SR Mil,. (M ,IR + M, ,)/(M,. + M1)
The sum of squares for rates of nitrogen, SN, was obtained
from the overall analysis, and the remaining terms represent a
decomposition of this as discussed by Cochran and Cox (3, p. 58).
In getting the sums of squares SL and SQ for the linear and
quadratic trends, respectively, two special sets of polynomial
values, z, and z2, were uesd.
Letting x denote the nitrogen rates, the corresponding values
of x, zi and za were as shown in Table 2.
The linear orthogonal polynomial z, was obtained from the
relation z, = k(x x), k being a purely arbitrary number se
lected to make z1 as simple a set of numbers as possible. In
doing this, the (x x) values were first observed to be 156,
Response of Pangola and Bahia Grass to Nitrogen 7
126, 66, 54, 294; then since these had a highest common
factor of 6, k was chosen as 1/6 to give the zl values shown.
TABLE 2.ORTHOGONAL POLYNOMIALS USED FOR OBTAINING LINEAR AND
QUADRATIC TRENDS.
X Zi Z2
30 26 30
60 21 11
120 11 19
240 9 47
480 49 25
The quadratic orthogonal polynomial z. was chosen to be as
simple a set of numbers as possible satisfying the conditions
S Z = 0, S ziza = 0 as well as being a quadratic function of x
(or z1).
If z', = kz2 = a + bzz + z21, the conditions give the equa
tions
Sz' = 5. + b Zi + y Z21 = 0
S zlz' = a zi + b S z21 + Z31 = 0
Since S zl = O, it follows that a = S z21/5 and b =  z31/
z21. Then Y z21 = 3,720 and 5 z31 = 90,210 from which a = 744
and b = 24.25. The z'. values are next found to be 562.50,
206.25, 356.25, 881.25, 468.75, and the given z, values are
obtained from these using z2 = z'2/18.75, the denominator 18.75
being the highest common factor of the z'2 values.
At this stage, the regression curve of yield on nitrogen rates
can be determined as y = y + b1z, + b2z2, where bl = 2
yzl/2 z27, and b0 = S yz.,/S z22, care being taken to sum over
all results in the experiment involved. For example, for Pangola
over all years b, = ( 26Ti 21T, + 49Ts)/480(3720)
where T1, T2, T5 are the yield totals for the five nitrogen
rates, 3,720 is the sum of the squares of the five different z,
values and the factor of 480 is introduced to allow for the fact
that each zl value occurs 480 times.
The significance of each b is tested by calculating SL
b, S yzi and SQ = b2S yz> and computing an F ratio as shown in
Florida Agricultural Experiment Stations
the analysis of variance, Table 1. Each mean square in this
table has to be tested in the same way as was explained for
testing the M11 term in the overall analysis of variance. Thus,
the F ratio for testing M11,L, for example, is (M11,L + M27)/(Mi5
+ M17). The effective degrees of freedom have to be obtained
for the numerator by the method explained. The F ratio for
the residual variation provides a basis for testing goodness of fit.
In passing, it should be noted that regression could have
been fitted on the orthogonal polynomials 2, 1, 0, 1, 2 for zL
and 2, 1, 2, 1, 2 for z2 given by Fisher and Yates (5).
These are the values which would normally be used if the x
values were evenly spaced. Their use in these data would be
equivalent to fitting a linear and quadratic regression on the
logarithms of nitrogen rates applied, since the latter rates are
evenly spaced on a logarithmic scale.
Where no convenient regression variable can be associated
with treatments, such as time of nitrogen application, then
multiple range tests (4) may be used for testing the differences
between means. For example, multiple range tests were used
for testing the differences between the mean yields of the five
different time sequences of nitrogen application for each grass
and in those years in which significant F ratios were obtained.
RESULTS AND DISCUSSION
In this experiment, Pangola showed more variation from
treatment to treatment and a larger coefficient of error variation
than Bahia. These differences between the two grasses are in
dicated in Tables 3 and 4, which present the F ratios from the
analyses of variance for both grasses for each of the three years
in which this experiment was conducted and for the pooled
analysis over all years, respectively.
It should be noted that the 240 and 480pound rates of nitro
gen caused a reduction in stand of Pangola during the 195051,
195152 and 195253 winter months, as evidenced by a reduction
in stand on these plots in the spring of 1951, 1952 and 1953.
Stands of Pangola rapidly reestablished on these plots and were
ready to harvest with the remaining plots. This variation in
stand may help explain the larger coefficient of error variation,
in Pangola.
Phosphorus and Potash Levels.The F ratios for the 06060
and 0120120 levels of PK are presented in Tables 3 and 4.
These F ratios were significant only in 1951 for Pangola and in
TABLE 3.THE F RATIOS FOR ALL TREATMENTS AND TREATMENT INTERACTIONS FROM THE ANALYSES OF VARIANCE FOR BOTH
GRASSES FOR EACH YEAR.
Source of Variation
Replications ................................
PK levels ...................... ..............
Error (a) ..................................
Timesequences of N application.
Timesequence x PK ...................
Error (b) ................ ...... 
Source of N ....................... ...........
Source x Timesequence ............
Source x PK ...............................
E rror (c) ............... ......... ......
Rate of N application ..... ...
Rate x Source ..... .........
Rate x Timesequence ...... .......
Rate x PK ...................................
Error (d) ........................... ....
H harvests ................... .............
Harvests x Rate ........... ....
Harvests x Source ................... ..
Harvests x Timesequence ..........
Harvests x PK .................. .........
E rror (e) .......................................
C.V.
d.f. Pangola Bahia
1950 1951 1952 1950 1951 1952 '3
... 1 18.40 435.98** 9.48 363.33* 1.29 0.16
.........  .0
......... 4 20.34** 5.84* 22.22** 12.13** 1.16 3.10 "
.. 4 0.22 0.76 1.42 1.92 1.06 0.98
8 Co
......... I 8
.... 1 122.51** 15.18** 0.93 0.70 15.20** 0.11 P
....... 4 5.93** 0.74 1.14 1.32 0.48 1.05
......... 1 0.08 0.05 .81 0.31 3.59 2.55
......... 14
........ 4 622.83** 183.71** 143.03** 200.38** 305.36** 94.13**
... 4 6.09** 3.39* 2.60* 2.50* 25.55** 1.50
....... 16 2.81** 3.16** 3.34** 0.90 1.14 2.84** "
....... 4 2.15 1.49 2.67* 0.44 2.20 2.20
........ 132
... 3 124.94** 9.33** 304.55** 217.95** 264.70** 558.90**:
.... 12 19.06** 6.47** 19.20** 27.29** 36.16** 1584**
....... 3 0.23 0.34 0.73 3.20** 3.42* 2.44*
..... 12 66.62** 30.54** 28.22** 26.77** 14.23** 6.63**
. 3 0.97 3.98** 7.06** 1.13 0.60 2.16
.... 567
I CQ
9.2 11.4 11.0 6.8 4.9 6.4
percent percent percent percent percent percent
10 Florida Agricultural Experiment Stations
1950 for Bahia. The existence of larger F ratios appearing
only in 1951 for Pangola and in 1950 for Bahia is difficult to
interpret, especially with Pangola, since it is known (6) that
this grass requires rather high quantities of potash for maxi
mum growth. Gammon and Blue (6) have reported that Pan
gola will store potassium in excess of its growth requirements
under conditions where potassium is present in large amounts.
Keeping this in mind and recalling that the PK applications
were made at the beginning of each growing season and noting
that the harvests x PK level F ratios are significant in 1951
and 1952 (Table 3), it can be assumed that after the first harvest
the residual amount of potash from the higher (0120120) level
was not much larger than that from the lower level of PK. Fur
TABLE 4.THE F RATIOS FOR ALL TREATMENTS AND TREATMENT INTER
ACTIONS FROM THE ANALYSES OF VARIANCE FOR BOTH GRASSES
OVER ALL YEARS.
Source of Variation d.f. Pangola Bahia
Over Years Over Years
Replications ....................... 1
PK levels .. .. ................. 1 13.19* 0.11
Error (a) .......... .. 1
Timesequences of N application. 4 1.15 2.34
Timesequence x PK............. 4 0.63 1.29
Error (b) ...... ........ 8
Source of N ................................ 1 2.08 0.41
Source x Timesequence.................. 4 0.78 0.50
Source x PK ...................... ........... 1 0.95 1.27
Error (c) ..... ....... .... ............... 14
Rate of N application ... ...... 4 16.67** 9.69**
Rate x Source ............... .. 4 0.62 0.89
Rate x Timesequence..................... 16 0.90 0.84
Rate x PK .............................. 4 1.28 1.53
Error (d) ....................................... 132
Years ............. 2 86.57** 17.18**
Years x Rate ..... ... 8 8.52** 57.45**
Years x Source ........................... 2 15.78** 8.51**
Years x Timesequence ...... 8 14.40** 3.43**
Years x PK ............................... 2 0.81 2.30
Years x Rate x PK........................... 8 1.55 1.09
Years x Rate x Timesequence .... 32 3.36** 2.08**
Years x Rate x Source ..... 8 4.01** 10.41**
Years x Source x PK ............. 2 0.20 1.54
Years x Source x Timesequence. 8 1.54 1.55
Years x Timesequence x PK .. ..... 8 1.45 1.01
Error (e) ....... .......... 312
C.V. 9.2 7.0
I percent percent
Response of Pangola and Bahia Grass to Nitrogen 11
thermore, the means of individual harvests show that the 0120
120 level was superior at the first harvest only. (This can
be seen in Appendix Table 3, which presents all of the data.)
The superiority of the 0120120 level at the first harvest only
leads one to conclude, as other workers have before, that potash
should be applied throughout the growing season.
TABLE 5.MEAN PRODUCTION OF DRY FORAGE PER ACRE FOR THE FIVE TIME
SEQUENCE OF NITROGEN APPLICATION TREATMENTS FOR THE FOUR CASES
IN WHICH THERE WERE SIGNIFICANT F RATIOS.
Grass Pangola
Year 1950
Treatments T:. Ti T, T1 T,
Means 7197 6949 6096 6032 5675
Grass Pangola
Year 1951
Treatments i T T, T1 T, T,
Means 8716 7950 7598 7579 6045
Grass Pangola
Year 1952
Treatments T T i T. T, T,
Means 9513 8292 8195 8067 6560
Grass Bahia
Year 1950
Treatments Ti T T T. T, T
Means 6757 6232 5847 5662 5649
Note: Any two means underscored by the same line are not significantly
different.
Any two means not underscored by the same line within one year
are significantly different.
Timesequence of Nitrogen Application.The significant F
ratios for timesequence of nitrogen application in 1950, 1951
and 1952 for Pangola and in 1950 for Bahia (Table 3) indicate
that in these years some of the time sequences for the nitrogen
application were better for forage production than others. The
means and the results of multiple range tests for the four cases
which had significant F ratios are presented in Table 5. The
significant interactions of harvest x time sequence for all three
years for each of the grasses indicates that the influence of time
12 Florida Agricultural Experiment Stations
sequence of nitrogen applications on yield differs somewhat
from one harvest to another. The nonsignificance of the F
ratios for time sequence of nitrogen application indicates that
over the three years the time sequences balanced out so that no
one was significantly better over all three years. This is more
vividly illustrated in Figs. 1 and 2. For instance, with Pangola
(Fig. 1), time sequences 1 and 3 were superior in 1950, time
sequence 2 superior in 1951, and time sequence 3 was superior
in 1952. The superiority of certain time sequences in some years
and other time sequences in other years probably depends on
the rainfall distribution of that particular year.
There were no significant time sequence x PK interactions
(Tables 3 and 4).
50
45
4.0
0
S2.5
L.
"o
1950
PANGOLA
Fig. 1.The production of Pangola for each time
application for each year.
sequence of nitrogen
Source of Nitrogen.There were significant F ratios for
source of nitrogen for Pangola in 1950 and 1951 and for Bahia
7
IIN
t tstut I:, t 4tts tI \It3tst
t, ttl tI It, Il'lt`t, It, tIlt
,.,. ~~~', .,N: 7
/ \ .'" " 
1901
Oz1
Response of Pangola and Bahia Grass to Nitrogen 13
in 1951 (Table 3). Figs. 3 and 4 show that for Pangola, source
1, NaNO3, was superior in 1950 and 1951; while for Bahia, source
2, (NH4)_SO4, was superior in 1951. The reason for the superi
ority of NaN03 on Pangola is probably that this plant is able
to make use of sodium in the absence of potash. This fact has
been demonstrated by Gammon (6). However, the nonsignifi
cant source F ratios in Table 4 for the pooled analysis indicate
that over a period of years, under conditions similar to the ones
in which this experiment was conducted, there would be no
differences in plant responses to the two sources of nitrogen.
45
o0
43.0
1.3.5
S2b
S2.0
> 1.5
S1.0
0.5
0
1950
1951
BAHIA
1952
Fig. 2.The production of Bahia for each time
application for each year.
sequence of nitrogen
The significant interaction of source x time sequence in 1950
for Pangola indicates that the plant's response to source of
nitrogen varied with the time at which the nitrogen was ap
plied. This is indicated in Appendix Table 3. However, the
failure of this interaction to be significant in any other year
makes it unimportant in the nutrition of these grasses, insofar
as this experiment demonstrates.
/ I
St t t t
13M 11 t1t 1 S 41 2 1214 415
r / II
/11 .. 11111 ..
Florida Agricultural Experiment Stations
It is interesting to note that with Bahia, but not with Pan
gola, there were significant harvests x source interactions (Table
3) for all three years. These interactions indicate that the in
fluence of source differs from one harvest to another.
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
1950
1951
1952
PANGOLA
Fig. 3.The production of Pangola for each source of nitrogen for
each year.
Rate of Nitrogen Application.By far the largest sources
of variation between treatments in this experiment came from
the rates of nitrogen applied. Although it can be seen that the
F ratios (Table 3) are larger in some years than in others for
the nitrogen rates, they are nevertheless quite large each year.
The large variation due to nitrogen rates (Tables 3 and 4)
was examined for polynomial trends. Highly significant linear
and quadratic trends were found. The F ratios and regression
coefficients for these linear and quadratic effects are presented
in Table 6. The production curves of the grasses based on the
linear and quadratic trends with different rates of nitrogen are
8, 8, 8 S,
'N\1
Response of Pangola and Bahia Grass to Nitrogen 15
presented for each year in Figs. 5 and 6. In general there is an
increase in production of forage for each additional increment
of nitrogen, except that at the higher rates the added incre
ments are less efficient in increasing forage production. This
fact is most striking with Bahia.
4.0
o 3.5 \
1 3.0
` 2.5.
0
'2.0 s8 8; S, 6, 8, 63
IJO\
005 L ,_ _
1950 1951 1952
BAHIA
Fig. 4.The production of Bahia for each source of Nitrogen for each year.
The fitted curves based on the significant polynomial trends
from the pooled analysis (over all years) are presented in Fig. 7.
Here, as with the individual year analysis, the significant linear
trend shows that over all years for each added increment of
nitrogen there was an increase in forage production. The signifi
cant quadratic trend indicates that at the higher rates the in
crements were less effective in stimulating forage production
than at the lower rates. The nonsignificant residuals in Table
6 for the pooled analysis indicate that the curves in Fig. 7 are
as reliable a fit as can be made.
The curves in Fig. 7 also indicate that at the lower rate (30
pounds) of nitrogen, Bahia was more efficient in nitrogen utili
zation, while at the higher rate (480 pounds) of nitrogen, Pan
16 Florida Agricultural Experiment Stations
gola was the more efficient. This result is better illustrated in
Fig. 8, which presents curves showing the production of dry
forage per pound of nitrogen applied for both grasses. At the
60, 120 and 240pound rates there were no differences in the
efficiency of nitrogen utilization between the two grasses.
TABLE 6.ERROR TERMS, F RATIOS AND REGRESSION COEFFICIENTS FOR THE
LINEAR AND QUADRATIC EFFECTS OF THE RATE OF NITROGEN
TREATMENT FOR EACH YEAR AND OVER ALL YEARS.
I Over
Source of Variation d.f. 1950 1951 1952 3 Years
Pangola
Rate of N application .. 4 4
Linear ............................ 1 2,289.28** 476.50** 410.89** 207.39**
Quadratic ... .. 1 192.95** 256.86** 144.76** 57.05**
Residual ............ .. 2 4.54* 0.74 8.23** 1.23
Error ........................ 132 (1,688) (5,613) (6,821) (44,700)
Linear regression: bl .. 25.23 21.20 21.70 22.79
Quadratic regression: b.. .371 .780 .645 .598
Bahia
Rate of N application...... 4
Linear .............. .... 1 7,335.14** 958.63** 38.77** 25.68**
Quadratic ................. 1 669.15** 260.89** 336.36** 13.04**
Residual ..... ............. 2 0.56 0.96 0.68 0.09
Error ............................ 132 (3,098) (2,514) (3,730) (149,859)
Linear regression: bli 19.53 18.00 4.41 13.55
Quadratic regression: b, .296 .471 .651 .485
The significant F ratios, Table 3, for the interactions, rate x
source and rate x time sequence, indicate that the grasses did
not respond alike to all rates of nitrogen for both sources of
nitrogen or to all times of nitrogen application. In fact, for the
most economic nitrogen utilization, these ratios would indicate
that each rate had an optimum source and time of application.
However, when the data were analyzed in a combined analysis
(Table 4), the interactions averaged out so that there were no
significant rate x source or rate x time sequence interactions.
The insignificance of these interactions in the combined analysis
points out that on the average at all rates of nitrogen applica
tion, the plants responded alike to both sources of nitrogen at
all five times of application.
Response of Pangola and Bahia Grass to Nitrogen 17
6.0
5.0 195
4.5 ____
4.0
315
2.5 /
1.5
1.0
0.5
0
30 60 120 240 480
PANGOLA Ibs. N per acre
. 5.Fitted curves showing yield of Pangola for each year, based on
significant linear and quadratic trends.
30 60 120 240 480
BAHIA Ibs. N per acre
Fig. 6.Fitted curves showing yield of Bahia for each year, based on
significant linear and quadratic trends.
Florida Agricultural Experiment Stations
6.0 
5.5 _PANGOLA
5.0
4.5 . _________ BAHIA
4 .5
V 1.0 
to
3 60s 120 240 480
Three year average lbs. N per acre
?ig. 7.Fitted curves showing the average yield of Pangola and Bahia,
based on significant linear and quadratic trends.
145
140
I130 __________________________________
110_____
7
o 4o
2.0
CP
490
** '
0.5  
_J 10
30 60 120 240 480
Lbs.Three yr vere Ibs. gen per acre
ig.Fitted curves showing the average yield of Pangoladry forage per pound Bahia,of nitrogen for
based on significant linear and quadratic trends.
140
120 \
 7o _
Response of Pangola and Bahia Grass to Nitrogen 19
Protein Analysis.In 1950, samples of forage from selected
rate treatments (R,, R3, R5) and selected time sequence treat
ments (T1, T3, Ts) were analyzed for their crude protein content.
Results from these chemical analyses averaged over all harvests
and replications are presented in Tables 7 and 8. Table 7 pre
sents the average percent of crude protein for each treatment
combination, and Table 8 presents average yields of crude pro
tein in pounds per acre for the selected treatment combination.
TABLE 7.AVERAGE PERCENT OF CRUDE PROTEIN IN CLIPPINGS FROM
SELECTED PLOTS OF PANGOLA AND BAHIA GRASS FOR THE YEAR 1950.
R,
Pangola
Ti ................ ........ 5.63
Ta ............................ 5.85
Ts ............................ 5.75
PK1 ......... ............. 5.83
PK2 ........... ...... 5.65
S1 ............................ 5.50
S ............................ 5.98
A ve. ..... .... .... .... 5.74
14.20
11.73
10.29
12.01
12.10
11.67
12.48
12.06
Ave.
9.38
8.12
7.53
8.27
8.19
8.08
8.40
8.23
Bahia
6.93
. 7.80
7.03
. 7.20
. 7.30
7.05
. 7.45
S 7.25
A ve. .......
7.58
7.73
7.83
7.78
10.63
11.00
9.73
10.31
10.58
10.45
10.45
10.45
8.44
9.07
7.99
8.50
8.49
8.41
8.58
8.49
The F ratios from the analyses of variance computed from
the yields of crude protein on the selected plots are presented in
Table 9. It can be seen that the factors, time sequence and
rate, influenced protein production in both grasses, and that the
factor, source, influenced protein production in Pangola. There
were also significant interactions between rate x source, rate x
time sequence, and rate x PK.
20 Florida Agricultural Experiment Stations
TABLE 8.AVERAGE YIELDS OF CRUDE PROTEIN IN POUNDS PER ACRE FROM
SELECTED PLOTS OF PANGOLA AND BAHIA FOR THE YEAR 1950.
R, R3 R. Ave.
Pangola
Ti .......... ....... 171 515 1615 767
T3 .......................... 166 479 1359 668
T5 ....... ...... ........ 142 344 1082 523
PK1 ......... ......... 165 434 1282 627
PK ........... ...... 154 458 1421 677
S1 ........................... 161 I 491 1417 690
S2 .. ................... 159 402 1287 616
Ave. .. .. ........ 160 447 1352 653
Bahia
Ti ............. ..... .... 191 374 686 417
T, ...... .. ......... 190 386 760 445
T ........................... 159 256 540 318
PK1 ........... ......... 177 348 687 404
PK ........ 184 329 637 383
S ........ .. 178 345 601 375
S ................... .. 183 333 723 413
Ave. ................... 181 339 662 394
If Tables 7 and 8 are examined, a comparison of the responses
of Pangola and Bahia to the treatments can be made. With
Pangola, T1 was superior in the production of crude protein
(both percent and total yield), whereas T3 was superior with
Bahia. There were no differences in response to the two levels
of PK, but S2 was superior in percent of crude protein in Pangola.
The superior forage production of S1 in Pangola (Fig. 3) was
sufficient to more than make up for the lower percent of crude
protein, thus making S, the higher of the two sources for total
production of crude protein.
Bahia had the higher percentage of crude protein at the R,
and R3 nitrogen treatments, while Pangola had the higher per
cent at the R5 treatment. Pangola was superior in the produc
tion of crude protein at the Rs and R5 treatments. At the R5
treatment it produced more than twice the total protein of
Response of Pangola and Bahia Grass to Nitrogen 21
Bahia. This result would lead one to conclude that at the higher
levels of nitrogen application Pangola is the best grass to choose,
while at the low nitrogen level Bahia is the best performer. A
better illustration of this may be noted in Table 10, which com
pares the two grasses in their average production of protein and
forage for each pound of nitrogen applied. As was pointed out
in Fig. 8, the grasses were more efficient in utilization of nitro
gen for the production of protein as well as for forage at the
lower levels, and Bahia was found to be the more efficient at
the low level and Pangola the more efficient at the higher levels.
It should be remembered that the protein analysis is based
on one year's data. If the same relation exists with protein as
with forage production, the significant time sequence and source
F ratios would average out over the years and be nonsignificant.
TABLE 9.THE F RATIOS FOR ALL TREATMENTS AND TREATMENT INTERAC
TIONS FROM THE ANALYSES OF VARIANCE FOR PROTEIN PRODUCTION IN 1950.
Source of Variation d.f. Pangola Bahia
Replications ......... ...... ........ .. ...... 1
PK levels ....... ................................... 1 4.73 13.40
Error (a) ..... ......................... .......... 1
Timesequences of N application ......... 2 24.99** 43.60**
Timesequence x PK ............................ 2 0.75 0.75
Error (b) ........................ 4
Source of N  ............ ........ ........... .. 1 38.53** 4.63
Source x Timesequence ........................... 2 0.74 0.77
Source x PK ......................................... 1 1.02 0.05
E rror (c) .......... ........ ..... .............. .. 8
Rate of N application .................... 2 933.39** 1516.46**
Rate x Source . ............. ........... ......... 2 5.17** 102.42**
Rate x Timesequence ....... ....................... 4 10.35** 18.54**
Rate x PK ............................. .... .. ..... 2 1.69 6.48**
Error (d) ............................................. 38
SUMMARY AND CONCLUSIONS
An experiment with a series of treatments involving two
rates of phosphorus and potash, two sources of nitrogen, five
rates of nitrogen, five times for nitrogen application, and two
grasses, Pangola and Pensacola Bahia, was conducted for three
years on a Leon fine sand. The experimental design used was
a split plot with four splits. The plots were harvested four
and five times each year and yields expressed as weight of dry
forage per acre. In 1950 the crude protein content was deter
Florida Agricultural Experiment Stations
mined on selected treatments. Analyses of variance were com
puted on an annual basis and on the pooled data over all three
years. Linear and quadratic trends were computed on the rate
of nitrogen treatment.
Although the F ratios for the PK, time sequence and source
treatments were significant in certain years, they failed to be
significant in the pooled analysis. This failure indicates, insofar
as this experiment demonstrates, that averaged over the three
years there were no differences in the response of the two grasses
to these treatments.
TABLE 10.AVERAGE PRODUCTION IN POUNDS OF PROTEIN AND DRY FORAGE
PRODUCED PER POUND OF NITROGEN APPLIED ON THE PLOTS FROM
WHICH PROTEIN ANALYSES WERE DETERMINED IN 1950.
30 lbs./acre 120 Ibs./acre 480 Ibs./acre
Grass Dry I Dry Dry
Protein Forage Protein Forage Protein [ Forage
Pangola 5.33 92.5 3.73 53.9 2.82 23.4
Bahia ... 6.03 1 83.0 2.83 36.1 1.38 13.2
There were significant F ratios for rates of nitrogen applied.
Significant linear and quadratic trends indicated that there was
an increase in forage production for each added increment of
nitrogen, but the higher rates of nitrogen were less efficient in
stimulating forage production than the lower rates. These
trends were about the same for each year, as well as for an aver
age over the three years. There were significant year inter
actions with all treatments except PK levels. All other inter
actions were insignificant in the pooled analysis.
Pangola, on an average, was the more productive of the two
grasses. Bahia was the more efficient in using nitrogen at the
lower level for the production of dry forage and protein, while
Pangola was the more efficient at the higher level.
LITERATURE CITED
1. ANDERSON, R. L., and T. A. BANCROFT. Statistical Theory in Research.
McGrawHill. 1952.
2. BLASER, R. E., W. E. STOKES, J. D. WARNER, G. E. RITCHEY and G. B.
KILLINGER. Pastures for Florida. Univ. of Fla. Agr. Exp. Sta.
Bul. 409. 1945.
3. COCHRAN, W. G., and G. M. Cox. Experimental Designs. John Wiley
and Son, Inc. 1950.
Response of Pangola and Bahia Grass to Nitrogen 23
DUNCAN, D. B. Multiple Range and Multiple FTests. Biometrics
11:1:142. March 1955.
FISHER, R. A., and F. YATES. Statistical Tables for Biological, Agri
cultural, and Medical Research. 3rd Ed. Oliver and Boyd. Edin
burgh. 1948.
GAMMON, N., and W. G. BLUE. Potassium Requirements for Pastures.
Proceedings, Soil Sci. Soc. of Fla. 12:154156. 1952.
HODGES, E. M., D. W. JONES and W. G. KIRK. Grass Pastures in Cen
tral Florida. Univ. of Fla. Agr. Exp. Sta. Bul. 484. 1951.
KILLINGER, G. B., and FRED H. HULL. Florida's Pasture and Forage
Crops. Univ. of Fla. Economic Leaflets 12:8. 1953.
SCHULTZ, E. F., JR. Rules of Thumb for Determining Expectations of
Mean Squares in Analyses of Variance. Biometrics 11:123135. June
1955.
Florida Agricultural Experiment Stations
APPENDIX
TABLE 1.TYPE OF ANALYSIS USED FOR EACH YEAR'S PRODUCTION
OF EACH GRASS.
Observed
M.S. M.S. Expected M.S.
Replications 1 M1 VE(e) + 4 VE(d) + 20 VE(d) + 40 VE(b) + 200 V(a) + 400 VR
PK levels 1 M2 VE(e) + 4 VE(d) + 20 VE(d) + 40 VE(b) + 200 VE(a) + 400 VPK M/M3
Error (a) 1 M3 VE() + 4 VE(d) + 20 VE(d) + 40 VE(b) + 200 V(a)
Timesequence of N
application 4 M4 VE(e) + 4 VE( + 20 VE(c) + 40 VE(b + 160 V M/M
Timesequence x PK 4 M5 VE(e) + 4 VE(d) + 20 VE(c) + 40 VE(b) + 80 VTPK M/M6
Error (b) 8 M6 VE(e) + 4 VE(d) + 20 VE(c) + 40 VE(b)
Source of N 1 M7 VE(e) + 4 VE() + 20 VE(c) + 400 VS M7/MI(
Source x Timesequence 4 M8 VE(e) + 4 VE(d) + 20 VE(c) 80 VST Ms/MI
Source x PK 1 M9 VE(e) + 4 VE(d) + 20 VE(c) + 200 VSPK Mg/MI(
Error (c) 14 M10 VE(e) + 4 + 20 VE(c)
Rate of N application 4 MI VE(e) + 4 VE(d) + 160 VN M/M
Rate x Source 4 M12 VE(e) + 4 VE(d) + 80 VNS MI1/M
Rate x Timesequence 16 M13 VE() + 4 VE() + 32 VNT M3/M
Rate x PK 4 M14 VE(e) + 4 VE(d) + 80 VNPK M14/MI
Error (d) 132 M15 VE(e)+ 4 VE(d)
Harvests 3 M16 VE(e) + 200 VH M16/M:
Harvests x Rate 12 M17 VE(e) + 40 VHR M17/M:
Harvests x Source 3 M18 VE(e) + 100 VHS Ma/M:
Harvests x Timesequence 12 M19 VE(e) + 40 VHT M19/M:
Harvests x PK 3 M20 VE(e) + 100 VHPK M20dM:
Error (e) 567 M21 VE(e)
Response of Pangola and Bahia Grass to Nitrogen 25
TABLE 2.TYPE OF ANALYSIS USED FOR POOLED ANALYSES OVER ALL YEARS.
Replications
PK levels
Error (a)
Timesequence of N
application
Timesequence x PK
Error (b)
Source of N
Source x Timesequence
Source x PK
Error (c)
[ate of N application
tate x Source
atte x Timesequence
tate x PK
iror (d)
fears
fears x Rate
fears x Source
'ears x Timesequence
'ears x PK
'ears x Rate x PK
'ears x Rate x Time
Observed
M.S. M.S. Expected M.S.
1 MI VE() + 4 VE(e) + 12 VE(d) + 60 VE(c) + 120 VE(b)
+ 600 VE(a) + 400 VY, + 1200 VR
1 M2 VE(f) + 4 VE(e) + 12 VE(d) + 60 VE(c) + 120 VE(b)
+ 600 VE(a) + 400 VypK + 1200 VPK
1 M3 VE() + 4 VE(e) + 12 VE(d) + 60 VE(c) + 120 VE(b)
+ 600 VE(a)
4 M4 V'(0 + 4 VE(e) + 12 VE(d) + 60 VE() + 120 VE(b)
+ 160 VyT + 480 VT
4 M5 VE(f) + 4 VE(e) + 12 VE(d) + 60 VE(c) + 120 VE(b)
+ 80 VypK + 240 VTPK
8 M6 VE(f) + 4 V(e) + 12 VE(d) + 60 VE(c) + 120 VE(b)
S M7 VE( + 4 VE(e) + 12 VE(d)+ 60 VE() + 400 VS
+ 1200 VS
4 M VE(f) + 4 VE(e) + 12 VE(d)+ 60 VE(c)+ 80 VyST
+ 240 VST
1 M9 VE() + 4 VE(e) + 12 VE(d) + 60 VE() + 200 VYSPK
+ 600 VSPK
14 M10 VE(f) + 4 VE(e) + 12 VE(d) + 60 VE(c)
4 M11 VE(f) + 4 VE(e) + 12 VE(d) + 160 VR + 480 VR
4 M12 VE(f) + 4 VE(e) + 12 VE(d) + 80 VyRS + 240 VRS
16 M13 VE(f) + 4 VE(e) + 12 VE(d) + 32 VYRT + 96 VRT
4 M14 VE(f) + 4 VE(e) + 12 VE(d) + 80 VRpK + 240 VRPK
132 M15 VE(f) + 4 V(e) + 12 VE(d)
2 M16 VE(f) + 4 VE(e) + 800 V
8 M17 VE() + 4 VE(e) + 160 VYR
2 M18 VE( + 4 VE(e) + 400 VYS
8 M19 VE(f) + 4 VE(e)+ 160 VYT
2 M20 VE(f + 4 VE(e) + 400 VypK
8 M21 VE(f) + 4 VE(e) + 80 VRPK
F
(M2 + M27)/(M20 + M3)
(M4 + M27)/(M6 + M19)
(M5 + M27)/(M6 + M2)
(M7 + M27)/(MIO + MI8)
(Mg+ M27)/(M0 + M25)
(M9 + M27)/(MIO + M24)
(M11 + M27)/(MI5 + M17)
(M2 + M27)/(M5 + M23)
(M13 + M27)/(MI5 + M22)
(M4 + M27)/(M5 + M21)
M16/M27
M17/M27
M18/M27
M19I/M27
M2/M27
M21/M27
sequence 32 M22 VE() + 4 VE(e) + 32 VYRT
'ears x Rate x Source 8 M23 VE(f) + 4 VEe) + 80 VyRS
'ears x Source x PK 2 M24 VE(f) + 4 VE(e) + 200 VSPK
Florida Agricultural Experiment Stations
TABLE 2.Continued.
Years x Source x Time
sequence
Years x Time x PK
Error (e)
Observed
M.S. M.S.
Harvests 3
Harvests x Years 6
Harvests x Rate 12
Harvests x Source 3
Harvests x Timesequence 12
Harvests x PK 3
Years x Harvests x Rate 24
Years x Harvests x Source 6
Years x Harvests x Time
sequence 24
Years x Harvests x PK 6
Error (f) 1701
Expected M. S.
M25 VE( + 4 VE(e) + 80 VST
M26 VE() + 4 VE(e) + 80 VypK
M27 VE() + 4 VE(e)
M28 VE() + 200 Vy + 600 VH
M29 VE(f) + 200 VYH
M30 VE(f) + 40 VYHR + 120 VHR
M31 VE(f) + 100 VYHS + 300 VHS
M32 VE(f + 40 VYHT + 120 VHT
M33 VE(f) + 100 VYHPK + 300 VHPK
M34 VE(f + 40 VYHR
M35 VE() + 100 VYHS
M36 VE() + 40 VyHT
M37 VE(0 + 100 VYHPK
M38 VE(f
F
M2/M27
M2/M27
M28/M29
M29/M38
M3/M34
M31/M35
M32/36
M3/M37
M34/M38
M35/M38
M36/M38
M37/M38
TABLE 3.AVERAGE YIELDS OF AIRDRY FORAGE IN POUNDS PER ACRE OF PANGOLA AND BAHIA GRASS FOR EACH TREATMENT
SHOWING SIGNIFICANT INTERACTION.
1952
SPK, PK,
3,757 3,869
6,236 6,144
8,116 9,237
9,710 11,689
11,074 11,423
7,778 1 8,473
2,664
4,006
5,727
8,374
10,221
6,198
6,687
5,891
7,003
5,594
5,817
2,048
1,307
1,807
1,036
2,677
4,273
6,212
8,533
11,212
6,581
7,211
6,300
7,391
5,756
6,248
2,211
1,464
1,796
1,110
1950
PK, i PK:,
Pangola
1951
Bahia
1951
Ave.
3,306
5,225
7,488
10,051
10,750
7,363
7,613
7,669
8,096
6,595
6,847
2,437
2,030
1,486
1,407
PK_
4,372
5,347
8,078
11,426
11,572
8,159
8,386
8,064
7,980
8,145
8,219
1,563
1,053
1,844
2,252
1,477
1952
PKi PK,
5,482 5,012
6,503 6,132
8,420 8,204
10,325 11,075
7,574 6,563
7,661 7,397
7,746 7,679
8,706 7,667
7,994 7,364
7,377 7,253
6,481 7,023
2,326 2,106
2,453 2,444
913 962
1,368 1,332
601 553
PK,
3,312
5,422
7,711
10,306
10,104
7,371
6,959
8,407
7,882
5,725
7,883
1,942
1,734
1,826
1,849
PK,
4,781
5,731
7,996
10,569
11,265
8,068
8,493
8,278
8,277
7,678
7,615
1,571
966
1,781
2,279
1,472
PK,,
3,559
5,272
7,925
11,697
10,469
7,784
8,238
9,026
7,276
6,365
8,017
2,402
1,811
1,751
1,820
8,715
8,802
10,086
8,111
6,650
3,347
2,932
885
1,309
PK, PK.
3,202 3,257
4,338 4,070
5,639 5,780
7,602 7,765
9,131 9,508
5,982 6,075
6,849 6,665
5,619 6,074
6,418 6,046
5,417 5,881
5,609 5,715
1,731 1,682
1,981 2,030
1,028 1,122
1,241 1,241
7,868
7,589
8,941
8,024
6,470
2,672
2,935
852
1,321
Ave.
4,350
5,354
7,352
9,793
9,268
7,223
7,636
7,401
7,346
7,002
6,777
1,814
1,821
1,275
1,619
1,026
TABLE 3.Continued.
Pangola Bahia
1950 1951 1952 1950 1951 1952
S, S. S1 S SI S SI S, SL S2 S1 Sr
R ........ ... 2,796 2,545 3,487 3,385 3,679 3,947 3,292 3,167 4,532 4,621 5,172 5,323
R ....... 4,503 3,777 5,607 5,087 6,474 5,906 4,342 4,066 5,857 5,221 6,447 6,188
R .............. 6,219 5,720 8,720 6,916 8,991 8,363 6,162 5,258 8,003 8,070 8,196 8,427
R .... ... ...... 9,199 7,708 11,796 10,206 10,427 10,972 7,734 7,633 10,943 11,053 10,465 10,935
RH ...................... 11,583 9,850 11,352 9,221 10,548 11,949 8,989 9,650 9,492 13,345 7,508 6,629
Ave. .............. 6,859 8,192 5,920 6,963 8,024 8,277 6,104 5,955 7,765 8,462 7,557 7,500
T   7,707 6,191 8,462 6,735 8,352 8,232 6,683 6,831 8,004 8,875 7,508 7,918
T, 6,530 5,661 9,246 8,187 7,732 8,659 6,309 5,384 7,782 8,560 8,106 8,267
T3 . ....  7,789 6,605 8,289 6,869 9,347 9,680 6.251 6,213 7,679 8,578 7,906 7,452
Ti ...... ...... 6,106 5,245 6,174 5,916 8,269 7,866 5,726 5,572 7,803 8,020 7,274 7,356
T5 6,166 5,899 8,792 7,108 6,420 6,701 5,551 5,774 7,557 8,277 6,995 6,509
HI ...... ............... 2,236 2,024 2,335 2,010 2,932  3,087 1,795 1,619 1,432 1,702 2,219 2,214
H .......... 1,531 1,240 1,981 1,584 2,877 2,990 1,983 2,029 979 1,041 2,546 2,360
H ....................... 1,919 1,685 1,904 1,673 856 881 1,104 1,047 1,733 1,861 900 975 .
H. .. ........ ..... 1,174 972 1,973 1,696 1,359 1,270 1,223 1,260 2,517 2,374 1,338 1,361
H, i.. ... 1,465 1,483 554 600 c
TABLE 3.Continued.
Pangola
1950 1951 1952
T, T. T, T TI .. T T, TT T:, T, T, T T. T T, T., o
H, ......... 2,338 2,237 1,994 1,835 2,244 2,730 2,779 1,677 1,444 2,231 2,625 3,562 3,225 2,932 i 2,704
H ... 2,050 579 1,861 1,867 571 2,225 1,316 1,896 2,028 1,449 3,986 1,647 3,760 3,568 1,705 0
H:, ..... 2,061 931 1,685 1,581 2,751 1,790 1,189 1,694 1,841 2,428 893 675 877 899 996
H, ..... 500 2,349 1,657 394 467 853 3,432 2,312 733 1,843 787 2,312 1,651 668 1,155
R .......... 3,065 2,768 2,826 2,338 2,354 2,865 3,463 3,643 3,675 3,533 4,175 3,783 4,213 3,856 3,039 o
R 4,965 4,358 4,529 3,756 3,090 5,591 5,451 4,841 4,348 6,504 6,396 6,155 8,314 6,006 4,105 "
R., 6,193 5,453 7,233 4,824 6,145 7,311 9,860 7,716 5,660 8,541 9,356 8,326 11,750 8,241 5,709
R ... 8,970 8,071 9,850 7,593 7,783 11,950 11,818 11,440 8,774 11,024 10,784 10,311 13,008 10,436 8,958
R .... 11,551 9,829 11,546 9,865 10,790 10,274 12,990 10,254 7,768 10,146 10,774 12,401 10,281 11,795 10,990 
Bahia
H, .... 1,995 1,786 1,463 1,401 1,891 1,420 1,745 1,388 1,634 1,649 2,114 2,406 2,233 2,279 2,048
H 2,532 1,500 2,233 2,271 1,494 1,104 1,115 926 929 976 2,876 2,244 2,449 2,565 2,108
H ........ 1,135 963 1,040 1,063 1,175 2,130 1,543 1,947 1,905 1,462 1,038 914 912 930 895 c
H .... 1,095 1,600 1,500 914 1,103 2,507 1,903 2,327 2,287 2,304 1,207 1,756 1,445 1,094 1,248
H.. 1,280 1,866 1,541 1,158 1,526 479 866 640 448 453
R ......... I 3,620 3,271 3,314 2,946 2,996 4,465 4,131 4,786 4,568 4,931 5,481 5,251 5,541 4,881 5,080
R .... 5,485 3,953 3,915 3,951 3,716 6,034 5,209 5,423 5,836 5,193 6,425 5,974 6,980 6,481 5,725
R:, .... 6,195 5,948 5,836 5,043 5,528 8,346 8,204 7,834 7,724 8,075 8,423 9,076 9,350 7,175 7,533
R, ... ..... 8,779 7,390 7,988 6,994 7,266 11,193 11,570 10,946 10,329 10,950 10,518 12,104 11,256 10,441 9,180
R ... ........ 9,705 8,671 10,106 9,310 8,804 12,160 11,741 11,653 11,101 10,435 7,716 8,525 5,265 7,594 6,241
I I
TABLE 3.Concluded.
Pangola
R,
852
458
808
553
1
R2
1,434
810
1,079
817
1950
R.3
2,020
1,196
1,528
1,226
761 1,101 1,419
1,034 1,402 1,849
695 803 1,201
740 898 1,241
1 1951
R, R R1  R, RI R4 
2,523 3,819 810 1,253 1,957 3,672
2,034 2,431 738 1,335 2,043 2,344
2,431 3,162 879 1,282 1,708 2,497
1,467 1,304 1,009 1,476 2,110 2,488
Bahia
2,144
2,616
1,281
1,643
3,110 444 618 1,144
3,128 433 571 933
1,397 1,020 1,299 1,920
1,685 1,710 1,900 [2,438
970 1,156 I1,602
1,710__
2,384
1,445
2,387
2,770
2,012
R
3,169
2,452
2,576
2,089
R1 I R.
1,117 2,106
1,334 2,316
655 746
658 1,024
1952
R 
2,863
3,480
775
1,560
3,244 1,325 1,685 2,364
1,669 1,737 2,124 2,768
2,361 782 856 1,039
2,514 1,041 1,193 1,557
1,631 363 460 585
R5
4,410
4,015
1,229
1,593
2,616
2,072
754
1,045
582
R.
4,552
3,472
936
1,739
3,090
3,542
1,258
1,914
895
