|
ON-FARM
RESEARCH
+&(C~i5x /2
- YR8~~
Todd Johnson
Steven Cook
Farming Systems Research and Extensio Methods
AGG 5813
Spring 1992
Dr. P. Hildebrand
INTRODUCTION
Problem Statement
The Green Revolution brought many improvements in the potential for very high yields to
agriculture with high levels of inputs. Most of this work was carried out at the International
Agricultural Research Centers (IARCs) scattered around the world. The research conducted at these
research stations often produced few benefits for low-resource farmers who could not dominate their
environments. These farmers needed lower input solutions that were more sustainable. Methodologies
for more effective research in sustainable agriculture were needed to replace existing ones borrowed
from on-station trials. Traditional experimental design focused on replications and predictable
environments. Hypothesis testing and means separation were the predominant concerns.
Farming Systems Research and Extension, in contrast, focused on small farm problems by
conducting on-farm research (OFR). OFR involves an entirely different set of criteria from those used
for experiment station work. FSRE on-farm research recognizes the fact that farms and farmers are
highly variable and targets this variability. (Wotoweic, et al.,1988). Research domains ideally contain
a wide range of environments which can be associated with farms, fields, or even portions of fields.
As an exercise in Farming Systems Research Methods two "villages" were selected on which
seven 3-person teams were to establish fertilizer trials on a "farm" in each village. The objectives of
the exercise were: 1) to learn the process of farmer-managed on-farm trials; and 2) to familiarize
ourselves with the different analytical procedures required for OFR.
SITE CHARACTERIZATION
Village Level
Two village sites were established on the University of Florida campus Agronomy Farm: 1)
behind Fifield Hall, on a plot formerly used for perennial peanut research for many years; and 2) near
1
INTRODUCTION
Problem Statement
The Green Revolution brought many improvements in the potential for very high yields to
agriculture with high levels of inputs. Most of this work was carried out at the International
Agricultural Research Centers (IARCs) scattered around the world. The research conducted at these
research stations often produced few benefits for low-resource farmers who could not dominate their
environments. These farmers needed lower input solutions that were more sustainable. Methodologies
for more effective research in sustainable agriculture were needed to replace existing ones borrowed
from on-station trials. Traditional experimental design focused on replications and predictable
environments. Hypothesis testing and means separation were the predominant concerns.
Farming Systems Research and Extension, in contrast, focused on small farm problems by
conducting on-farm research (OFR). OFR involves an entirely different set of criteria from those used
for experiment station work. FSRE on-farm research recognizes the fact that farms and farmers are
highly variable and targets this variability. (Wotoweic, et al.,1988). Research domains ideally contain
a wide range of environments which can be associated with farms, fields, or even portions of fields.
As an exercise in Farming Systems Research Methods two "villages" were selected on which
seven 3-person teams were to establish fertilizer trials on a "farm" in each village. The objectives of
the exercise were: 1) to learn the process of farmer-managed on-farm trials; and 2) to familiarize
ourselves with the different analytical procedures required for OFR.
SITE CHARACTERIZATION
Village Level
Two village sites were established on the University of Florida campus Agronomy Farm: 1)
behind Fifield Hall, on a plot formerly used for perennial peanut research for many years; and 2) near
1
two pecan trees on a field where winter rye had been growing. These villages we called Fifield and
Pecan Grove, respectively. The simulation of on-farm research lacked interaction with farmers/ land
managers. It was, however, useful as an exercise in methods of conducting on-farm trials. The
process is probably best learned as a simulation first, in order for mistakes to be made before the
researchers' credibility with farmers is added as another of the many uncontrollable factors inherent
in on-farm research.
Pecan Grove had a discernable, fairly constant slope toward the southeast corner. A soil
sample revealed slightly more acidic (pH=6.0) conditions than optimum for radish production
(pH =6.5). Fifield was originally to be situated in a fertile area, long used as a experimental garden.
Because of miscommunication however, the area actually donated for our use and tilled was the tractor
entrance road adjacent to the garden. This area therefore had a history of heavy compaction and
exhibited a plow-pan several inches below the surface. The soil sample from this farm had pH of 7.6,
which was significantly higher than the 6.5 considered optimal. The high pH at Fifield may have
contributed to poor results obtained.
Farm Level
Differences in farm environments were more pronounced at Pecan Grove than at Fifield
primarily because the "farms" were neither contiguous nor even in line. The Fifield farms were mostly
in line and, with the exception of two treatment plots, represented the same environment.
We conducted a pre-plowing qualitative assessment of the existing micro-environments at
Pecan Grove. A transect was taken using a meter tape and a Growth Quality Index value
(scale = 1 > > 10) assigned, based on vigor and greenness of the cover crop, in this case winter rye.
(See Figure 1).
, *, ,B -7
a~.
4.
..: >
- c
A-; ~fl)l
,.. ,.. .. -
.- ,- ~~~~~~~~~~~~~~~.: ....! ::.....".-:-".:i-.--.:.... ..i:- .: :.: -'..
-oO
Cr1
-. : : -- .--. -"- "... -" -
-, .. -,
.. .. .: ..~ ~ ~.. -,. .: .- .. .
_. ". .t :. .
.. .- .- : : '....... ." ... ..'
.. ,. ._ . _. : .... ,,
,..~~ ,. ,..
.o : '. : : _' ,.., -- -. :.. .
*'
~~`
- -
'"'
i
~-. _1
-. ~--~I
"' -:'-;;
::
::
'"~-
III. METHODS
On-Farm Trials
Each of the seven farming systems teams was assigned, in each village, a "farm" composed
of five 4 x 5 ft plots, for a total of 14 farms of 100 ft2 each. Teams prepared their plots and sowed
radishes on 29 February, 1992. Each 20 ft2 plot corresponded to one treatment -- rate of ammonium
nitrate fertilizer (33-0-0 NH4NO0) application. The treatments (in Kg of elemental N applied) were: a)
0 Kg N; b) 100 Kg N; c) 200 Kg N; d) 300 Kg N; and e) 400 Kg N. No replicating was done on
individual farms, rather, the farms themselves became replicates, consistent.with OFR practices. All
teams harvested on 4 April, 1992. We weighed radish yields per plot in grams (w/o greens), then
calculated per-acre yields in tons, based on the area harvested.
Description of the 14 farms [see SITE CHARACTERIZATION] gave a research domain.
Biophysical similarities reduced heterogeneity across the research domain to less than would be
expected in an actual OFR project. However, a range of environments emerged, presumedly based
primarily on differences in management practices. Variability of e.g. intensity of land preparation,
seeding rates (plant density), or plot aspect (orientation and slope) was common between teams, but
homogenized within the teams' farms. Also, the watering, weeding, thinning and hilling that teams
practiced varied over a heterogeneous population of homogeneous individual farms. This conforms to
OFR standards. Calculation of the Environmental Index (El) accounts for these management
differences, i.e. they are embedded in Els, the mean response over all treatments for each individual
farm.
Statistical Analysis
Data were analyzed using the Modified Stability Analysis (MSA) procedure. Because of
blocking errors invalidating their results, two farms were removed from the data set. Among the
remaining twelve farms, one exhibited an El several orders of magnitude larger than the other eleven.
Thus, two recommendation domains became obvious, based on high or low El. Since only one farm
fell into the high environment category, further analysis concentrated on the poor-environment farms.
Regression of radish yields per treatment, as a function of environment, gave prediction
equations for response across all environments in the population of interest. Differences in radish yield
within treatments were identified and separated. Diffusion domains corresponding to the two villages
resulted.
Further partitioning of farm by treatment variation proved difficult. An additional parameter
was generated based on the differences in management practices described earlier. This gauge of
input levels we called the Overall Management Index (OMI). El was regressed against water input and
against OMI.
Confidence intervals were calculated for yields per SUS of cash cost input at each treatment
level. From these intervals, we computed an estimation of relative risk in both better and poorer
environments. Results are now presented, and discussed later.
S IV. RESULTS
Appendix I contains summary tables of results on each farm. These are grouped by teams, one
farm per team in each of the two villages. Also included in Appendix I are graphs of results, by farm,
across all treatments. The data are summarized in Table 1. Els are defined as mean responses at each
location across all treatments. Scatter plots of data points at each treatment level across all Els are
given in Appendix II. The first set of scatter plots are for all Els, while a second set shows data spread
for only the poorer environments.
Based on the shape of data spread, orders of parameters were determined for modelling line
4
TOTAL YIELD (TONS/ACRE) PER TREATMENT
Radish Trials; Pecan Grove & Fifield; 1992
LOCAT 0 100 200 300 400 El
111 I II
0.096
0.101
0.096
0.000
0.180
0.000
0.296
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.366
0.568
0.911
0.473
1.162
0.240
4.242
0.425
0.041
0.042
0.000
0.036
0.042
0.545
0.497
1.270
0.473
0.467
0.863
0.485
4.969
0.018
0.286
0.036
0.180
0.012
0.473
1.056
0.294
0.621
0.348
0.444
0.252
0.222
4.689
0.000
0.212
0.024
0.042
0.030
0.515
0.234
0.210
0.633
0.288
0.348
0.252
0.270
5.057
0.024
0.168
0.024
0.030
0.000
0.665
0.072
0.292438
0.638457
0.422994
0.346442
0.541624
0.243252
3.850484
0.093466
0.141350
0.025164
0.050328
0.015578
0.339114
0.381425
Table 1. Yield as tons/acre of all radishes >$0.10 in 14 plots on University of Florida
Agronomy Farm, 2 sites. Five treatments (Kg N / Ac). Environmental Index (El)
calculated as mean yield across all treatments.
Table 1
equations. For the AII-Els graph, linear relationships described the data distribution for all levels of
fertilization. The Low-Els plots, however, differed in their relationships. Data were distributed thus:
a) 0 Kg treatment level -- linear; b) 100 Kg --quadratic; c) 200 Kg -- cubic; d) 300 Kg -- linear; and e)
400 Kg -- linear. Regressions of yield against El, using the appropriate order of parameter (linear,
quadratic or cubic) allowed generation of prediction equations at each treatment level. These lines are
in Figures 2 and 3 (AII-Els and Low-Els).
Characterization of the environments produced a Growth Quality Index (GQI) and an Overall
Ba *
Management Index (OMI) (Table 2) [see SITE CHARACTERIZATION and STATISTICAL ANALYSIS].
Attempts to partition the environmental variability were conducted. A regression of El as a function
of water input (Figure 4) showed little influence. Also, El did not appear to be strongly correlated with
GQI (Figure 5) or the estimate of aggregated management practices, OMI (Figure 6). There remains
a large amount of unexplained variation in El.
V. DISCUSSION
Recommendations based on the All-El Treatment Responses (Figure 2) must be prefaced with
the warning that the vast differences between the lowest eleven environments and the single high one
(far right) distort the scale and shape of the treatment curves. The only point that can be made based
on this view of the results is that the zero nitrogen treatment is clearly the lowest across environments.
The Low-El Treatment Responses (Figure 3) graph disaggregates those eleven lowest
environments. Again, the zero treatment was lowest across all environments. In addition, the
environments begin to separate into three subgroups that can be considered different recommendation
domains; the five clustered to the left side (00.2), the four in the middle (0.20.5), and the
two to the right (0.50.7). These subgroups will be referred to as Groups 1, 2, and 3,
respectively. All groups showed a similar basic trend in which the 200 and 100 treatments appear
5
ALL-E.I. TREATMENT RESPONSES
Radishes; Pecan Grove and Fifield; 1992
5.0
000 Kg
w .0
cc...' 200 Kg
< 0 ------ - - ------- --. ---. --. ---- -- -- -- -- -- --------- ---- ------------------- --- 4 0 0 K
2.0- -------------------------------------------------------------- Kg
2 1 0 .0. . .
40Kg
--------------------------------------------
-III Il I I I
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
ENVIRONMENTAL INDEX
Figure 2
LOW-E.I. TREATMENT RESPONSES
Radishes; Pecan Grove and Fifield; 1992
1.0
200 000 Kg
0.8 -------- -.--------------------------------- ------------- -------- -"-- ------ 100 Kg
100
200 Kg
S 0.6 300--------- --------------- ---
S..-0 300 Kg
)--*-.--- 400 -
S0.4 -4 .t ------------------------ 400 Kg
S.......................................................................................
0 --- ------- -- --: ---- ----- -
^ --"'"^*''"'
-AAA A A A A A A A A
-0.'0 0.2 0.4 0.6
ENVIRONMENTAL INDEX
Figure 3
OVERALL MANAGEMENT INDEX (O.M.I.)
SOIL
LOCAT TEAM VILLAGE PREP
FERT
APPL
WATER
APPL
THIN
FREQ
WEED
FREQ
SOIL
MOUND
0.096
0.096
7.656
0.096
0.096
11.436
83.256
7.656
18.996
22.776
41.676
11.436
7.656
98.376
AREA OF
HARVES
8.00
8.00
8.00
6.65
8.00
8.00
6.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
O.M.I. LOCAT
47.596
74.096
31.656
62.746
89.096
51.436
151.256
55.156
97.996
46.776 -
108.176
110.436
52.656
188.376
LOCAT
TEAM
VILLAGE
SOIL PREP
FERT APPL
WATER APPL
THIN FREQ
WEED FREQ
SOIL MOUND
AREA OF HARVEST =
O.M.I.
location number assigned to each farm
Farming Systems team number
village of farm location-- PG: Pecan Grove; FF: Fifield
pre-planting land preparation; relative intensity (1-5 scale 10)
fertilizer applied as a single band (1) or as double bands (2)
liters of water applied per plot, including rainfall
number of thinnings (5 points assigned per occurrence)
number of weeding treatments (10 points assigned per occurrence)
number of times radishes mounded after thinning/weeding/watering (5 points per occurrence)
average area (sq. ft.) harvested per plot
summation of above factors for each farm
Table 2
KEY:
stated above) the zero treatment gave the lowest yields. Group 1's yields were perhaps slightly better
at the 200 treatment than the 100 treatment although a means separation would have to be performed
to be more certain. In Groups 2 and 3 the results look similar except that the differences between
treatment yields appear more clearly delineated than in Group 1. Responses to 100 Kg of nitrogen per
acre were higher than those for 200 in Group 2. Finally, in Group 3 the responses of the top two
treatments are reversed and 200 is better than 100.
Yield responses are per acre projections based on very small actual-yields per plot. Despite the
fact that droughts are experienced by real farmers, dry periods of the type we experienced are rare in
North Florida in March. Any recommendations based on separation of data means from on-farm trials
with real farms and farmers would probably show real differences in response to similar treatments.
Further partitioning of variation in El was attempted. We tried to explain differences based first
on the most obviously varying management practice: water input. Figure 4 is a curve obtained by
regressing liters H20 applied per plot against El. The model contains a quadratic function (R2=.284),
yet maintains linear shape. This suggests that a simpler model would do as well in explaining the
variation. The relatively flat slope of the line {E(Y) = a +x + 4x2 ; = 0.005973- W = 0.000112} across
a wide range of water application rates shows that El is not very responsive to watering. Another
source of variability might be identified.
Figure 5 indicates that the Growth Quality Index (GQI) also does not explain the prevalence of
low Els. The single high El, while on the highest quality site, is nevertheless an outlier when compared
S to all other environments. The remaining higher growth environments did not exhibit substantially
higher yields or Els. There is a mild trend, however, of slightly increasing Els for successively higher
quality sites. More detailed analysis, using larger sample sizes, would be needed for this observed
trend to be described as significant. The sample size for our trials was not large enough to allow
6
ENVIRONMENTAL CHARACTERIZATION
Radishes; Pecan Grove & Fifield; 1992
A-.
0 30 40 50 60 t0 80
WATER APPLICATION RATE (I/plot)
- PRED.
* ACTUAL
Figure 4
3
2
-----------------------------------------------------------------------------------------------------------------------------
-----------------------------------------------------------------------------------------------------------------------------
- -....................................................................................................................
> - . - . . . . . . . . . .
^^ "7"___
1
"6
n
100
ENVIRONMENTAL CHARACTERIZATION
Radishes; Pecan Grove & Fifield; 1992
GROWTH QUALITY INDEX (GQI)
Figure 5
interaction terms to be included in a model.
The Overall Management Index (OMI) represents another attempt at environmental
characterization or explaining variation among Els (Figure 6). Admittedly based on subjective
observatiori, and in some cases relatively arbitrary assignment of values, OMI is a tool for possibly
quantifying differing levels of management input. We tried to be as objective as we could in assigning
values for e.g. degrees of soil preparation, weeding and thinning, and water input. The scaling for
different parameters included in the OMI was based on general trends of relative effects on plant
growth, as contained in the body of agronomic knowledge. More complete, detailed data on the
various parameters could make OMI a more useful tool.
Application of OMI to our data set, by regressing it against the response variable El, gave
similar results as did the water input curve. The model {E(Y) =a+4x+ Yx2 ; #=0.010183, =-3.9E-
06} explained only 18% of the variation in El when all 14 data points were included. Next, we
considered the relationship between OMI and El at each farm location. As Figure 7 indicates, higher
levels of management input did not necessarily result in a higher El. What about a possible correlation
of GQI, OMI, and El? Since our sample size of 14 farms was only one-third of that needed for the
simplest linear interaction model (E(Y) = a +x + w +A v}, a bar chart (Figure 8) proved the only means
of observing the relationship. The X-axis is a sorted listing of progressively larger GQI values. OMI
values are paired with corresponding Els and scaled on opposite Y-axes. The low Els observed, even
at relatively high OMI levels, indicate very little relationship between El and OMI, or any evidence of
interaction with GQI. We could not explain nor identify factors influencing El.
Appendix III contains confidence bands for each treatment. Unlike the other analyses discussed
so far, which were based on the researchers' criterion (yield per unit land area), the confidence bands
use a criterion more commonly important for farmers: yield per unit cash outlay. From the lower values
7
ENVIRONMENTAL CHARACTERIZATION
Radishes; Pecan Grove & Fifield; 1992
A-
34----------------------------------------------------------------
2 -----------------------------------------------------------------------------------------------------------------
140
1i6o 1i
200
* ACTUAL
"aI
-- - - - - - -- - - ---------------
---------
---- ---- ---- ---- ---- ----
n
hI
100
120
OVERALL MANAGEMENT INDEX
- PRED.
Figure 6
- --
20
ENVIRONMENTAL CHARACTERIZATION
Radishes; Pecan Grove & Fifield; 1992
LOCATION NUMBER
O.M.I. E.I.
Figure 7
ENVIRONMENTAL CHARACTERIZATION
Radishes; Pecan Grove & Fifield; 1992
200 X
LU
180
160 z
I-
120
0
100 <
z
80 <
60 _.
-J
-40
-20
0
GROWTH QUALITY INDEX
E.I. O.M.I.
Figure 8
X
Q
H-
z
Z
LU
0
LU
usa .In crlifiiiu inure cominiily iniapuatfIrnf^ rmerg yield pnr uinit cach outatfy. Fruiim tlia Inwa~i V 5rp
of the confidence intervals, estimations of risk were calculated. These are given in Figures 9 and 10,
for the poorer and better environments, respectively. The better environments are found in Pecan
Grove, while Fifield had the poorer environments.
Looking at Figure 9, we see that farmers requiring yields of 0.2 Kg for every $US spent could
choose to apply either 100 or 200 Kg N/ac, depending on the level of acceptable risk. Application of
200 Kg entails a slightly lower risk of not obtaining the required yield. Conversely, if a different farmer
was seeking to maximize yield per dollar spent, pnd was willing to accept that 25% of the time she
may not reach the peak yield, then 100 Kg of N/ac would be a more logical choice. In the better
environments of Pecan Grove* village (Figure 10), 100 kilos of nitrogen applied per acre is optimum
at all levels of acceptable risk.
8
8
RISK ESTIMATION, RADISH
POORER ENVIRONMENTS (FIFIELD)
0.6
sa -m--
100 Kg
0.5..----------------------------------..------...--------------- 200 Kg
-A-
---- 300 Kg
Cf)0.4 ------------------------------------------------ ---------------- ----------------
I I I I I
5 10 15 20 25 Kg
.3-------E ---------------------------- ------------------------
-r ....X.-
0 0 ..........................................-..............-- - ..... ..-...........----......
-/ ----
0----------------------------------- ----
-. .-: -.......x.
0 5 10 15 20 25
PERCENT OF TIME BELOW VALUE
Figure 9
RISK ESTIMATION, RADISH
BETTER ENVIRONMENTS (PECAN GROVE)
6
100 Kg
--,.-4--.
5- ------------------------------------------------------...............----------....................... ------------------------------------------------------- 200 Kg
at-A-
300 Kg
-- 4- ....300 Kg........-.. -.................................................................................... ....
0 400 Kg
CO ) 3- ----- --- -- --- ------------ ----------- --.----- --------..------------- ----------: ......------------... ...........
< /- ---------------- -
0) / - ---
A -----------~
1-- "" ........-- ... ... -.-----.-
.- ......-...................--................... ......................
0 5 10 15 20 25
PERCENT OF TIME BELOW VALUE
Figure 10
APPENDIX I
Summary tables of results, by team
Response curves across treatments, by farm
FSR ON-FARM TRIALS: RADISH RESPONSE TO FERTILIZATION
TEAM 1 TEAM 1
PECAN GROVE
PECAN GROVE
RADISH #
TRTMNT > $.05
Kg/ha N
0 2
100 46
200 43.
300 31
400 46
TOTALS
# TRTs
TOTAL
NO.
RADISH WT
grams
15
501
614
570
598
28 196
2298
grams
22
30
8
17
35
TOTAL AREA OF
WT. HARVEST
grams ft*ft
37 6
531 6
622 6
587 6
633 6
2410
YIELD BY SIZE
> $.05
tons / acre
0.119828
4.002267
4.904974
4.553477
4.777157
30 18.357702
> $.10
tons / acre
0.175748
0.239657
0.063908
0.135805
0.279599
0.894718
GRAND TOTAL YIELD (GTY)
ENVIRONMENTAL INDEX (El)
FIFIELD
FIFIELD
RADISH # TOTAL RADISH WT
TRTMINT > $.05 > $.10 NO. > $.05
grams
0
75
159
21
0
grams
0
16
34
TOTAL
WT.
grams
0
91
193
AREA OF
HARVEST
ft*ft
8
8
8
80 335
YIELD BY SIZE
> $.05
tons / acre
0.000000
0.449356.
0.952635
0.125820
0.000000
1.527811
GRAND TOTAL YIELD (GTY)
ENVIRONMENTAL INDEX (El) 0.401425
TOTAL
YIELD
tons/ acre
0.295577
4.241923
4.968882
4.689283
5.056756
19.252421
3.850484
Kg/ha N
0
100
200
300
400
TOTALS
# TRTs
> $.10
tons / acre
0.000000
0.095863
0.203708
0.107846
0.071897
0.479313
TOTAL
YIELD
tons / acre
0.000000
0.545219
1.156344
0.233665
0.071897
2.007125
PECAN GROVE: TEAM I (LOCAT 7)
6
large
5 -------------------------------------------------------- ----------------......................................................
total
4 ---------------------- -- -------------------------------------------------------------------------------------
c I
I4-
C
I
2$------------------------------------
C
0 100 200 300 400
NITROGEN APPLICATION RATE (Kg/ac)
FIFIELD: TEAM I (LOCAT 14)
1.25
large
1 total
LU
S 0.75 ------------------------------------ ------- ----------------- ---- ---------------------------------------------
0
S0.5 -------------------------.....---- ------------------------------------------------------------------.....--.....................---------------
0.25------------- -------- -------------------------------------- ----------------- -------------------------------
100 200 300
NITROGEN APPLICATION RATE (Kg/ac)
FSR ON-FARM TRIALS: RADISH RESPONSE TO FERTILIZATION
TEAM 2 TEAM 2
PECAN GROVE
PECAN GROVE
RADISH # TOTAL RADISH WT
TRTMNT > $.05 > $.10 NO. > $.05 > $.10
Kg/ha N
0
100
200
300
400
TOTALS
# TRTs
grams
0
0
29
10
10
grams
0
40
52
27
35
49
TOTAL AREA OF
WT. HARVEST
grams ft*ft
0 8
40 8
81 8
203
YIELD BY SIZE
> $.05
tons / acre
0.000000
0.000000
0.173751
0.059914
0.059914
40 0.293579
tons / acre
0.000000
0.239657
0.311554
0.161768
0.209700
0.922678
GRAND TOTAL YIELD (GTY)
ENVIRONMENTAL INDEX (El)
RADISH #
> $.05 > $.10
0 0
0 3
0 27
5 11
9 11
TOTAL
NO.
0
3
27
16
20
RADISH WT
> $.05 > $.10
grams
0
0
.grams
0
7
0 79
TOTAL
WT.
grams
0
7
79
FIFIELD
AREA OF
HARVEST
ft*ft
8
8
8
42 86
46 111
YIELD BY SIZE
> $.05
tons / acre
0.000000
0.000000.
0.000000
0.263622
0.389442
40 0.653064
GRAND TOTAL YIELD (GTY)
ENVIRONMENTAL INDEX (El) 0.339114
TOTAL
YIELD
tons / acre
0.000000
0.239657
0.485305
0.221682
0.269614
1.216258
0.243252
FIFIELD
TRTMNT
Kg/ha N
0
100
200
300
400
TOTALS
# TRTs
> $.10
tons / acre
0.000000
0.041940
0.473322
0.251640
0.275605
1.042507
TOTAL
YIELD
tons / acre
0.000000
0.041940
0.473322
0.515262
0.665047
1.695571
PECAN GROVE: TEAM II (LOCAT 6)
100
200
300
NITROGEN APPLICATION RATE (Kg/ac)
.00- .%-
0.754- .---- -- ---------------------------------
C"
0
LU
>-
-mr-
large
total
0.5-
0.25-
_I
n
.Ini
400
--------------------------------------------------------- ---------------------------------------------------------
---------------------------- ---------------------------------------------------- -------------- ..........
0
FIFIELD: TEAM II (LOCAT 13)
1.25
-I.-
large
1- total
co 0.75 --------------------------------------------------------------------------------------------------------------------
0.5- -------------1-----------------------------------------------------------
0 .25 ----------------------------- ------------ ------------------------------------------ ---- -----------------------
0.7
S0.5-
-j
0.5-
0.2$-----i'-------------if--------------
0 100 200 300 400
NITROGEN APPLICATION RATE (Kg/ac)
FSR ON-FARM TRIALS: RADISH RESPONSE TO FERTILIZATION
TEAM 3 TEAM 3
PECAN GROVE
PECAN GROVE
RADISH # TOTAL RADISH WT
TRTMNT > $.05 > $.10 NO. > $.05 > $.10
grams
25
139
108
Kg/ha N
0
100
200
300
400
TOTALS
# TRTs
55 102
301
TOTAL AREA OF
WT. HARVEST
grams grams
5 30
55 194
36 144
24 42
31 42
' 151
452
YIELD BY SIZE
> $.05
tons / acre
0.149785
0.832807
0.647073
0.107846
0.065906
> $.10
tons / acre
0.029957
0.329528
0.215691
0.143794
0.185734
40 1.803417 0.904704
GRAND TOTAL YIELD (GTY)
ENVIRONMENTAL INDEX (El)
FIFIELD
RADISH #
> $.05
TOTAL
NO.
RADISH WT
> $.05 >
grams
0
0
0
0
0
TOTAL AREA OF
$.10 WT. HARVEST
grams
0
6
2
5
0
grams
0
6
2
5
0
0 13
YIELD BY SIZE
> $.05
tons / acre
0.000000
0.000000
0.000000
0.000000
0.000000
40 0.000000
GRAND TOTAL YIELD (GTY)
ENVIRONMENTAL INDEX (El) 0.015578
TOTAL
YIELD
tons / acre
0.179743
1.162335
0.862764
0.251640
0.251640
2.708121
0.541624
FIFIELD
TRTMNT
Kg/ha N
0
100
200
300
400
TOTALS
# TRTs
> $.10
tons / acre
0.000000
0.035949
0.011983
0.029957
0.000000
0.077888
TOTAL
YIELD
tons / acre
0.000000
0.035949
0.011983
0.029957
0.000000
0.077888
PECAN GROVE: TEAM III (LOCAT 5)
NITROGEN APPLICATION RATE (Kg/ac)
c:
0
Q
-j
0
w
(I)
large
total
FIFIELD: TEAM III (LOCAT 12)
1.25T
0.75 ----------...-----------------------......----------------------------------------------------.........................
0.54 --- .------------------------------------------- --------------......----
0.25- --.........................................-------------------..........
I-IE
100
200
300
AMMONIUM NITRATE [45-0-0] RATE (Kg/ac)
large
total
,1.
400
Cl)
C
0
4-a
O
I
co
CE
I R K
14 ------------------------------------ --------------- --------------
-----)---- -r-
TEAM 4
PECAN GROVE
FSR ON-FARM TRIALS: RADISH RESPONSE TO FERTILIZATION
TEAM 4
PECAN GROVE
RADISH #
TRTMNT > $.05
Kg/ha N
0 0
100 3
200 4
300 4
400 4
TOTAL
NO.
RADISH WT
grams
0
15
21
17
23
grams
0
S22
57
34
35
TOTAL
WT.
grams
0
37
78
51
58
76 148
TOTALS
# TRTs
AREA OF
HARVEST
ft*ft
8
3.75
8
5.5
8
33.25
YIELD BY SIZE
> $.05
tons / acre
0.000000 '
0.191725
0.125820
0.148151
0.137803
0.603499
>$.10
tons / acre
0.000000
'0.281197
0.341511
0.296303
0.209700
1.128710
GRAND TOTAL YIELD (GTY)
ENVIRONMENTAL INDEX (El)
FIFIELD
RADISH #
> $.05 > $.10
0 0
0 0
2 9
0 3
0 2
2 14
TOTAL
NO.
0
0
11
3
2
16
RADISH WT
> $.05 >$.10
grams grams
0 0
22.5
TOTAL
WT.
grams
0
0
30
7
5
34.5
AREA OF
HARVEST
ft*ft
8
8
8
8
8
40
YIELD BY SIZE
> $.05
tons / acre
0.000000
0.000000.
0.044936
0.000000
0.000000
0.044936
GRAND TOTAL YIELD (GTY)
0.251640
ENVIRONMENTAL INDEX (El) 0.050328
TOTAL
YIELD
tons / acre
0.000000
0.472923
0.467331
0.444454
0.347502
1.732210
0.346442
FIFIELD
TRTMNT
Kg/ha N
0
100
200
300
400
TOTALS
# TRTs
TOTAL
YIELD
tons / acre
0.000000
0.000000
0.179743
0.041940
0.029957
> $.10
tons / acre
0.000000
0.000000
0.134807
0.041940
0.029957
0.206704
PECAN GROVE: TEAM IV (LOCAT 4)
0.75-------------------------------- ------------------------------------------------------------------------------
100 200 300
NITROGEN APPLICATION RATE (Kg/ac)
1.25-
large
total
)
400
14 --------------------------------------------------------- --------------------
-- - - - - -- -- - - - - - -------.----------.~-_-__ ~_,
0.5
0.25-
0:
FIFIELD: TEAM IV (LOCAT 11)
1.251
0.754 -----------...----------------------------------------------------------------- --------------
0.5 ---------------------------------------- -- --------
0.254 --------------------------------- ------------ ------- ----------------------------
200
300
r
large
total
total
400
NITROGEN APPLICATION RATE (Kg/ac)
c:
0
c)
I
w
U)
0
iLp~
..................
0
100
FSR ON-FARM TRIALS: RADISH RESPONSE TO FERTILIZATION
TEAM 5 TEAM 5
PECAN GROVE
PECAN GROVE
RADISH #
TRTMNT > $.05
Kg/ha N
0 1
100 13
200 6
300 10
400 8
TOTAL
NO.
RADISH WT
> $.05
grams
9
128
52
48
34
grams
7
24
27
TOTAL
WT.
grams
16
152
79
AREA OF
HARVEST
ft*ft
8
8
8
8
8
82 353
TOTALS
# TRTs
YIELD BY SIZE
> $.05
tons/ acre
0.053923
0.766901
'0.311554
0.287588
0.203708
1.623674
tons / acre
0.041940
0.143794
0.161768
0.059914
0.083880
0.491296
GRAND TOTAL YIELD (GTY)
ENVIRONMENTAL INDEX (El)
FIFIELD
FIFIELD
RADISH # TOTAL RADISH WT
TRTMNT > $.05 > $.10 NO. > $.05 > $.10
Kg/ha N
0
100
200
300
400
TOTALS
# TRTs
grams
0
0
0
0
0
TOTAL AREA OF
WT. HARVEST
grams grams
--0 0
7 7
6 6
4 4
4 4
ft*ft
8
8
8
8
8
YIELD BY SIZE
> $.05
tons / acre
0.000000
0.000000
0.000000
0.000000
0.000000
40 0.000000
GRAND TOTAL YIELD (GTY)
ENVIRONMENTAL INDEX (El) 0.025164
TOTAL
YIELD
tons / acre
0.095863
0.910695
0.473322
0.347502
0.287588
2.114970
0.422994
> $.10
tons / acre
0.000000
0.041940
0.035949
0.023966
0.023966
0.125820
TOTAL
YIELD
tons / acre
0.000000
0.041940
0.035949
0.023966
0.023966
0.125820
PECAN GROVE: TEAM V (LOCAT 3)
NITROGEN APPLICATION RATE (Kg/ac)
large
total
FIFIELD: TEAM V (LOCAT 10)
1.25 1
1-
0.75-
0.5-
0.25-
lit ----~ -
200
300
300
NITROGEN APPLICATION RATE (Kg/ac)
r-
0
w
LU
>-
I
0)
Q
large
total
total
400
-------------------------------------------------------------------------------------------------------------------
100
FSR ON-FARM TRIALS: RADISH RESPONSE TO FERTILIZATION
TEAM 6 TEAM 6
PECAN GROVE
PECAN GROVE
RADISH #
TRTMNT > $.05
Kg/ha N
0 1
100 9
200 20
300 13
400 11
> $.10
6
15
18
15
18
TOTAL
NO.
7
24
38
28
29
RADISH WT
> $.05
grams
5.77
56.71
160.46
70.95
61.96
grams
11.14
38.06
51.45
32.64
43.67
TOTAL AREA OF
WT. HARVEST
grams ft*ft
16.91 8
94.77 8
211.91 8
103.59 8
105.63 8
72 126 355.85 176.96 532.81
TOTALS
# TRTs
YIELD BY SIZE
> $.05
tons / acre
0.034570
0.339773
0.961383
0.425091
0.371228
40 2.132046
> $.10
tons / acre
0.066744
0.228033
0.308258
0.195560
0.261645
1.060241
GRAND TOTAL YIELD (GTY)
ENVIRONMENTAL INDEX (El)
FIFIELD
RADISH WT
> $.05
grams
0
0
25.2
9.08
16.15
41 50.43
grams
0
6.8
22.46
TOTAL AREA OF
WT. HARVEST
grams ft*ft
0 8
6.8 8
47.66 8
26.33 35.41
11.94 28.09
67.53 117.96
YIELD BY SIZE
> $.05
tons / acre
0.000000
0.000000
0.150984
0.054402
0.096761
40 0.302147
GRAND TOTAL YIELD (GTY)
0.706748
ENVIRONMENTAL INDEX (El) 0.141350
TOTAL
YIELD
tons / acre
0.101315
0.567807
1.269641
0.620651
0.632873
3.192287
0.638457
FIFIELD
TRTMNT
Kg/ha N
0
100
200
300
400
TOTALS
TOTAL
NO.
RADISH #
> $.05 > $.10
0 (
0
5
2 11
3
# TRTs
TOTAL
YIELD
tons / acre
0.000000
0.040742
0.285551
0.212156
0.168299
tons / acre
0.000000
0.040742
0.134567
0.157754
0.071538
0.404600
PECAN GROVE: TEAM VI (LOCAT 2)
NITROGEN APPLICATION RATE (Kg/ac)
large
total
FIFIELD: TEAM VI (LOCAT 9)
1.25-i
0.75 ------------------------------------------------------------------ --- --- ------ --
0.5-
0.25-
4
iiipp~
0
100
200
300
0
NITROGEN APPLICATION RATE (Kg/ac)
-I~--
large
total
total
400
CD
0
I
r-
CO3
11 t--------------------------------------------------------- --------------------
----------------------------------------------------- --------------- -----------------------------------------
FSR ON-FARM TRIALS: RADISH RESPONSE TO FERTILIZATION
TEAM 7 TEAM 7
PECAN GROVE
PECAN GROVE
RADISH #
> $.05
1
7
10
3
2
TOTAL RADISH WT
> $.10 NO. > $.05
grams
6
38
46
25
9
>$.10
grams
10
23
37
24
26
124 o 120
TOTAL
WT.
grams
16
61
83
49
35
244
AREA OF
HARVEST
ft*ft
8
8
8
8
8
YIELD BY SIZE
> $.05
tons / acre
0.035949
0.227674
0.275605
0.149785
0.053923
40 0.742936
>$.10 .
tons / acre
0.059914
0.137803
0.221682
0.143794
0.155777
S0.718970
TRTMNT
Kg/ha N
0
100
200
300
400
TOTALS
# TRTs
FIFIELD
RADISH #
> $.05
TOTAL
NO.
RADISH W1
> $.05
grams
0
52
0
0
0
T TOTAL AREA OF
> $.10 WT. HARVEST
grams grams ft*ft
0 0 8
19 71 8
3 3 8
0 0 8
4 4 8
YIELD BY SIZE
> $.05
tons/ acre
0.000000
0.311554
0.000000
0.000000
0.000000
> $.10
tons / acre
S0.000000
0.113837
0.017974
0.000000
0.023966
40 0.311554 0.155777
GRAND TOTAL YIELD (GTY)
ENVIRONMENTAL INDEX (El) 0.093466
GRAND TOTAL YIELD (GTY)
ENVIRONMENTAL INDEX (El)
TOTAL
YIELD
tons / acre
0.095863
0.365476
0.497288
0.293579
0.209700
1.461906
0.292381
FIFIELD
TRTMNT
Kg/ha N
0
100
200
300
400
TOTALS
STRTs
TOTAL
YIELD
tons / acre
0.000000
0.425391
0.017974
0.000000
0.023966
0.467331
PECAN GROVE: TEAM VII (LOCAT 1)
1.25
large
total
5 0.75- ---------------------------------- --------------------------------------------------------------------------
1------------------------------------------------------------------------------------------------ to a
LU)
0.5----- --------------------------- -------
0
0.25t-------,--..-pl~"
0 100 200 300 400
NITROGEN APPLICATION RATE (Kg/ac)
FIFIELD: TEAM VII (LOCAT 8)
400
NITROGEN APPLICATION RATE (Kg/ac)
large
total
APPENDIX II
Scatter Plots for all environments (EIs)
Scatter Plots for low El environments
NO FERTILIZATION: ALL E.I.s
Radishes; Pecan Grove & Fifield; 1992
0.4 -------------------------------------------------------------------------------
0.24 ---------------------------------------------------- ------------------ ------------------
0.1 .------- ---- ----------------- ---------------------------------- ---------------------.-------------
1.5
2.0
2.5
ENVIRONMENTAL INDEX
LIN
1.0
3.0
3.5
4.0
0.5-
0.3 --------------------------------------------------------------------------------- ------------------------------------ ---
0.0 IA0.5-A-
0.0 0.5
100 Kg FERTILIZATION: ALL E.I.s
Radishes; Pecan Grove & Fifield; 1992
3.0 --------------------------------------------------------------------------------------------------------------------------
2.0 ------------------------ : .--- ..-- ----.. .---.... .......--- -- .---.----- .......-- ... -
1.0 .-----------;L -----..-------------- ..---.--
AL
1.5
2.0
2.5
3.0
3.5
LIN
4.0
ENVIRONMENTAL INDEX
A
4.0 ---------------------------------------------------------------------------------------------------------------------------
0.5
1.0
0.0
0.0
3n
200 Kg FERTILIZATION: ALL E.I.s
Radishes; Pecan Grove & Fifield; 1992
5.0
4.0 ---------------------------------------------------------------------------- ---
3.0 ---------------------- ....-------------------------------- ---..- .--......................
2.0 ---------------------------- ---------------------------------------------------------------------------------------------
1.0 -------------- ----------------------------------------------------------------------------------------------------------
I I I
1.5 2.0 2.5
ENVIRONMENTAL INDEX
LU
C/)
Z
C/)
O
0
w
LU
LU
LIN
A
nn
0.0
0.5
3.0
3.5
4.0
r
300 Kg FERTILIZATION: ALL E.I.s
Radishes; Pecan Grove & Fifield; 1992
5.0-
W 4.0- ----------------------------------------------------------------------------------------------------------------- -
0 3.0- ------------ -- .--------------------------------------
Cr
0
C/ 3.0 ------------------------------------------------------------------------------------------------------------------
w
I- 2.0 .......-----------------------------------------------------------------------
.<
cf) 10--------------------------------- --------------------"----"--------- -- ---------
I-
ENVIRONMENTAL INDEX
AJA'e
400 Kg FERTILIZATION: ALL E.I.s
Radishes; Pecan Grove & Fifield; 1992
I I Inn
1.5 2.0 2.5
ENVIRONMENTAL INDEX
6.0
4.0 ------------------------------------------------------------------------------
3.0 ------------------------------------------------------------------------------
2.0 ------------------------------------------------------------------------- ----------------------------------
nn
U.Ua
0.0
0.5
1.0
3.0
3.5
4.0
1~1
r
>
5.0o4 --------------------------------------------------------- ---------------------------------------------------------- ................A--.
1.04 -------------------------------------------------------------------------------------------------------------------------
,a4t~aa
ll
NO FERTILIZATION: LOW E.I.s
Radishes; Pecan Grove & Fifield; 1992
0.5-1
0.44- --------------------------------------------------------------------------------------
0.34-----..-....... -----.............................................................................................
0.24..................----------------------------------- .........-----------------------------
An
I A A
0.0
A A
0.1
A A .
0.2 0.3 0.4 0.5
ENVIRONMENTAL INDEX
(D
Cd
0
_I
-U
Li
LI
0.6
0.7
0.1 -------.-..-..-.................... ....- .................... .....................-------------------------------.............-.........
100 Kg FERTILIZATION: LOW E.l.s
Radishes; Pecan Grove & Fifield; 1992
0.6-
...................................................................................................................................................................................................................
0.3 ..................--------------------------------------------------.. ..---.. ------
IA*
0.2 0.3 0.4 0.5
ENVIRONMENTAL INDEX
0
C)
a
LU
t
0n
A
0.0
0.1
0.1
0.6
0.7
1.2 ------.--------------------------------------------------------------------------------------- .- A ------------.-.-----------
0.9------------------------------------ -------------------------------------------------------------.................................----..-.......--
I
200 Kg FERTILIZATION: LOW E.I.s
Radishes; Pecan Grove & Fifield; 1992
1.5-
0.6j -------------- -- ------------------- -------- ------------------ --
A A A
.... .............--------------------------.... ..--------------------.-.--...------------ -------
0.1
1 f 8 I
0.2 0.3 0.4 0.5
ENVIRONMENTAL INDEX
LI
CD
C
0
_J
>-
0.3-
0.0
0.6
0.7
1.2 ------------------------------------- --- -- --- --- ---- -- -- --- ------- -- ------ --I
0.9 --------------------... ----- ...------------...-- --... ..--- -..-................... .. ...... ..............
nt
300 Kg FERTILIZATION: LOW E.I.s
Radishes;Pecan Grove & Fifield; 1992
5I
0.9---------------------------------------- --- -----------------
A
A
--- --- -- --- --- -- --- ----------------------..- & ------------ I ----------------------------- : ---------------------------
A
0.1
0.2
0.3
0.4
0.5
0.6
0
1.
ENVIRONMENTAL INDEX
(I)
C-
-1
W
t
0.0
.7
1.2 ------------------------------------------------------------------------------------------------------------------------
.6 ----------------------------------------------------------------------------------- --- -- --------- .A ......
0.3-
400 Kg FERTILIZATION: LOW E.I.s
Radishes; Pecan Grove & Fifield; 1992
r-
0.94------------------------------- -- ------------------
AA
A
........................................ .: ...-.... ...... .....- .- ..- .- ... .. .. .. .. .. .. .. .. .. .. .
.A "
I I I I ....
0.2 0.3 0.4 0.5
ENVIRONMENTAL INDEX
CD
L.
O
C,)
r"
0
4J
w
0.6-
0.3-
"nI
U.U 1A
0.0
.A A A
0.1
0.6
0.7
1.2 ------------------------------------
LIN
APPENDIX m
Confidence Intervals for better environments (Pecan Grove)
Confidence Intervals for poorer environments (Fifield).
I.
DISTRIBUTION OF CONFIDENCE INTERVALS
Radish; 100 Kg N/Ac; PECAN GROVE; 1992
20
15------------------------------------------------------------------------
I
c/)
1 5 .--- -- -- -- ----. -. -.-. -. .-.-.----- --- --- --- -.. ---.-.. .- ..-.-.-- . -
50 60 70 80 90 100
PERCENT WITHIN INTERVAL
DISTRIBUTION OF CONFIDENCE INTERVALS
Radish; 200 Kg N/Ac; PECAN GROVE; 1992
10
O
O
I
80-----------------------------------------------
|/ 6 -------------------------------- --------- ----------------------- -- -- --------- -- --- --- ---
S0 60 70 80 90 100
0
I --------*-------0
50 60 70 80 90 100
PERCENT WITHIN INTERVAL r
DISTRIBUTION OF CONFIDENCE INTERVALS
Radish; 300 Kg N/Ac; PECAN GROVE; 1992
<0. _------------- ---
g4 5
100
PERCENT WITHIN INTERVAL
2.5-
i-
c)
0
0
cO/)
I<
0
0)
1.5-
1-
0.5-
_n
------------------------------------------------------------------------------------------------------------------------ -
-------------------------------------------------------------------------------------------------------------- ----------
--------------------------------------------------------------------------------------------------------------------------
--------------------------------------------- --- ---------------- -----------------------
-----------------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------- m ------------------------- ------------------------
---------------------------- ) -----
----------- -- -----------------------
....... -------------
.................. --------------------
24
50O
DISTRIBUTION OF CONFIDENCE INTERVALS
Radish; 400 Kg N/Ac; PECAN GROVE; 1992,
2.5
0
I
0- .5----------------------------------|
0 --------------------- I- --------.- --.-.- .-.-- ---- -- ---------- .----- ------------------------------------- -- /
--- -------------
0 .5 ,,,,,,
50 60 70 80 90 100
PERCENT.WITHIN INTERVAL
DISTRIBUTION OF CONFIDENCE INTERVALS
Radishes; 100 Kg N/Ac; FIFIELD; 1992
8
6 ---------------------------------------------------------------------------------------------------------------------
S 4- ------------------------------------------------- --------------------------------------- ---------------------------- ------
I
0C
2 - -- - ----------------------- ---------- -- --------------- ------ ------------------- -------
C) 40- ------------------------------------------------- -------- ----- ---------------- ----------- -------------- -------
< -------- ---
-46................------....-.. .-- ). -----.-..----
6 II I I 0 I I I
P50 60 70 80 90 100
PERCENT WITHIN INTERVAL
DISTRIBUTION OF CONFIDENCE INTERVALS
Radishes; 200 Kg N/Ac; FIFIELD; 1992
3
F-
< ------------------------------------------------------------------------------------------ ----------------------
0
50 60 70 80 90 100
PERCENT WITHIN INTERVAL
DISTRIBUTION OF CONFIDENCE INTERVALS
Radishes; 300 Kg N/Ac; FIFIELD; 1992
1.5
O
1.5--------------------------------------------------------------------------------------- -- ---------
O 0
-0)
0 ----------------------------------- ---- ---- ------------------------- -- ------------ -------
-0.5- ------------------------------------------------------------------------------------------------ ------
-.05-------------------------- ---------
-1 I I I I I11 1
50 60 70 80 90 100
PERCENT WITHIN INTERVAL
i
DISTRIBUTION OF CONFIDENCE INTERVALS
Radishes; 400 Kg N/Ac; FIFIELD; 1992
0.81 1
0.6-
I
)- 0.4-
O
S0.2-
r
')
S-0.2-
-0.4-
-0.6-
5
0
100
PERCENT WITHIN INTERVAL
------------------------------------------------------------------------------------------------------------------------- -
------------------------------------------------------------------------------------------------------------------ -------
------------- 7 -------------------------------------------------- ft-ft ---------
---------------------------------- -------------- ft ------------------------- -------------------- ft -------- ft --------
------------------------------------------------ w -------------------------- -------- ----------------- --
---------------------------------------------------------------------------------------- ft -------------------------------
|
RSS
HIDE
