SL345

Fertilizer Experimentation, Data Analyses, and

Interpretation for Developing Fertilization
Recommendations—Examples with Vegetable Crop
Research1
George Hochmuth, Ed Hanlon, and Allen Overman2

Introduction The target audience for this article includes Extension state
specialists, county Extension faculty members, and profes-
Fertilizer recommendations contain several important
sionals conducting or working with research in nutrients,
factors, including fertilizer form, source, application timing,
agrochemicals, and crop production. The authors assume
placement, and irrigation management. Another important
that the reader has an understanding of basic probability
part of a fertilizer recommendation is the amount of a
and statistics. Statistical information presented in this
particular nutrient to apply. The optimum fertilizer amount
publication is intended to demonstrate the process involved
is determined from extensive field experimentation con-
in fertilizer experimentation. Explanation of the statistics
ducted for several years, at multiple locations, with several
and their calculations is beyond the scope of this document.
varieties, etc. Although rate is important, rate should be
considered as a part of the overall fertilization manage-
ment program. The important components of a fertilizer Experimentation
recommendation are discussed in Hochmuth and Hanlon The goal of research on fertilizer rate is to determine the
(2010a) Principles of Sound Fertilizer Recommendations for amount of fertilizer needed to achieve a commercial crop
Vegetables, available online at http://edis.ifas.ufl.edu/ss527. yield with sufficient quality that is economically acceptable
This EDIS publication focuses on the research principles for the grower. In Florida, these types of studies take a
behind determining the optimum rate of fertilizer, includ- slightly different approach depending on whether soil
ing experimentation and interpreting research results for testing for the nutrient in question is involved. For example,
optimum crop production and quality in conjunction with rate studies with nitrogen (N) on sandy soils would not
minimal environmental consequences. We use examples involve soil testing, but rate studies with phosphorus (P) or
from research with vegetable crops in Florida. How we potassium (K) would. In the case of N on sandy soils, the
interpret the results is as important as how we conducted researcher assumes there is minimal N supplied from the
the research. soil that would satisfy the crop nutrient requirement. In the
case of P or K, a properly calibrated soil test will reveal if

1. This document is SL345, one of a series of the Department of Soil and Water Sciences, UF/IFAS Extension. Original publication date March 2011.
Reviewed December 2017. Visit the EDIS website at http://edis.ifas.ufl.edu.

2. George Hochmuth, professor, Department of Soil and Water Sciences; Ed Hanlon, professor, UF/IFAS Southwest Florida Research and Education
Center, Department of Soil and Water Sciences; and Allen Overman, professor, Agricultural and Biological Engineering Department; UF/IFAS Extension,
Gainesville, FL 32611.

The Institute of Food and Agricultural Sciences (IFAS) is an Equal Opportunity Institution authorized to provide research, educational information and other services only to
individuals and institutions that function with non-discrimination with respect to race, creed, color, religion, age, disability, sex, sexual orientation, marital status, national
origin, political opinions or affiliations. For more information on obtaining other UF/IFAS Extension publications, contact your county’s UF/IFAS Extension office.

U.S. Department of Agriculture, UF/IFAS Extension Service, University of Florida, IFAS, Florida A & M University Cooperative Extension Program, and Boards of County
Commissioners Cooperating. Nick T. Place, dean for UF/IFAS Extension.
a response (yield and fruit quality) to the nutrient is likely (Figure 1) to gain insight into the type and magnitude of
or not. Rate studies are best conducted on soils low in the response. Plotting the raw data allows the researcher to
particular nutrient so that maximum crop response is likely inspect for apparent atypical data points that may illustrate
and that response can be modeled. errors somewhere in the data entry process.

Proper experimental design and statistical data analyses are

critical to interpretation of the results. Research begins with
a hypothesis or a set of hypotheses. One possible hypothesis
may be that there will be no effect on yield associated with
N fertilization. This hypothesis, called the null hypothesis,
is evaluated with an experiment to test crop yield response
against a range of N rates in a field likely to produce a large
response to the addition of N fertilizer.

The researcher applies a range of fertilizer rates thought to

capture the likely extent of possible crop yield responses. A
zero-fertilizer treatment is always included. Crop response
without an actual fertilizer application demonstrates and
Figure 1. Theoretical (not actually measured data) crop response to
measures the soil-supplied effects, if any. In some cases, nitrogen fertilization. There were 5 replicates of each N rate (some
sufficient nutrients, or at least a low portion of the crop data points are hidden behind others). Note the rapid increase in
nutrient requirement, may come from the soil, while in yield with the first few increments of N fertilization, then a leveling
other cases, nutrients may come from the irrigation water. off, and possibly an indication of yield reduction with excessive
fertilization. Also, note there is some yield with zero fertilization, in this
case approximately 20% of the maximum yield. We will use the term
The researcher may decide to divide the total seasonal “percent relative yield” to express this percentage of the maximum
amount of fertilizer into split-applications, following what yield. There appears to be no further increase in yield after 150–200
would likely be a recommended practice for the crop being lbs/acre N. This example is used for illustration purposes; typically in
studied. Multiple applications avoid potential large losses of field experimentation, there is more variation among the replications,
especially at the lower rates of fertilization.
fertilizer because of rainfall events, especially for nutrients
that are mobile in the soil. Typically, all treatment rates are Once the data have been collected and inspected, they are
handled similarly for timing and placement of the fertilizer analyzed statistically with analysis of variance (ANOVA).
to minimize any confounding effects with rate. Did fertilization have a significant effect on yield? ANOVA
is particularly useful in cases where a researcher might
During the growing season, the researcher may sample be evaluating the effect of fertilizer rate across several
the plant for nutrient concentrations, using whole dried varieties of crop. Here, the researcher is interested in
leaves and/or fresh petiole sap. These samples will help the whether varieties differed in response to fertilizer, which
researcher prove the response in yield was related to the will be exposed through a significant interaction term in
plant’s nutrient status. Typically, soil samples are not used the ANOVA source table. If the fertilizer treatment effect
because there is a chance of including a fertilizer particle in was significant, then the researcher will want to graphically
the sample, or there may be questions of where to sample present the results with a mathematical equation sometimes
if the fertilizer is applied by banding or through a drip called a “model.”
tape. Photographs taken during the season are useful for
documenting both growth and potential plant deficiency In fertilizer rate experiments, the rate of fertilizer is referred
symptoms. to as a continuous variable because there are many possible
rates in addition to the ones the researcher selected to use
The crop response of interest, typically marketable yield, is in the experiment. Using ANOVA, especially if the experi-
measured at the appropriate harvest time(s). For vegetables, ment had the treatments arranged in a factorial arrange-
the fruits are evaluated according to USDA grade standards ment, is a good approach to test for treatment effects and
to detect any effects of fertilization on fruit quality (size, interactions. Fertilizer rate main effects can be subjected
color, sugar content, etc.). Yields are expressed in the to polynomial contrasts, a statistical method to determine
prevailing commercial units per area of production (e.g., if there are linear or quadratic components in the overall
28-lb boxes/acre, 42-lb crates/acre, bushels/acre, tons/acre, response. Then regression methods can be applied to the
etc.). The raw data should be plotted in a scatter diagram

Fertilizer Experimentation, Data Analyses, and Interpretation for Developing Fertilization ... 2
continuous variable to develop an equation that explains this difference is a real effect. We call these differences
the significant trend in response (see the section below “significant” differences. If ANOVA detects significant
about models). differences among treatment means, then we reject our null
hypothesis.
The ANOVA statistics for a randomized complete-block N
experimental design (data in Figure 1) with five replications In the “real world,” finding no significant differences has
and nine N rates indicate that one or more N rate treat- two major implications. First, it means that farmers should
ments were statistically different from the others (Table 1). not be interested in spending extra money each year (for
In this case, our null hypothesis would have been rejected. “insurance” applications) just to gain the rare possibility
Since ANOVA tables contain estimates of several variance of a real crop response. These unjustified expenses would
components, these tables should be included in research reduce profitability. The second implication is the potential
manuscripts but are seldom included. For example, other negative impacts on the environment when a rate of
researchers may be able to use this information when sum- fertilization is applied to the crop when not needed.
marizing numerous, similar studies. While simply reporting
means and treatment effects is good for a simple research One common misinterpretation about treatment differenc-
report or presentation, this method does not contain es needs clarification. For example, assume an experiment
measures of variance, and the ANOVA table does. was conducted to test the effect of N rate on tomato yield
and the ANOVA found no significant difference between
the grower rate and the recommended (lower) rate at the
Treatment Significance 5% probability level. This finding means that there is such
Researchers cannot study every possible experimental a rare chance of a real treatment difference occurring that
treatment (rate) or combinations of treatments. In addition, we can be confident the grower can reduce the commercial
there is natural variation in the field where the research fertilizer rate. The actual means may be 2,950 and 2,920
will be conducted. The field may have variations in organic boxes/acre for the grower and recommended rates, respec-
matter, soil pH, or moisture, all of which may lead to tively. An argument could be made to someone without
variations in yield response having nothing to do with the knowledge in statistics that the 30 boxes/acre “difference”
N treatment(s). Therefore, the notion of probability comes is “worth” $600 (30 boxes at $20/box) and that amount will
into play. What are the chances that the observed differ- more than pay for the added fertilizer with the grower rate.
ences in yield are because of natural variation from plot to This conclusion is erroneous because the ANOVA indicated
plot? This inherent variability is where statistical analysis of no significant difference between the two treatment means.
the data helps to sort out the differences most likely caused Therefore, the appropriate representation of the response
by treatment (N fertilization) from the so-called “noise” or to fertilizer is the average of the two means (i. e., 2,935
random error in the production system. If we repeat the boxes per acre). Said another way, other factors on the farm
application of treatments, called replication, we can esti- impact yield more than fertilizer rate.
mate the relative amount of natural variation. Experiments
should always include replication as part of a properly A more complex experiment may be to test the response
designed experiment, one that would pass a peer-review of two cultivars to N rate. Here, ANOVA is used to test the
process. Analysis of variance is the mathematical tool we significance of the main effect of N rate, the main effect of
use for this analysis, and with this statistical tool we can test cultivar, and the interaction in the response of cultivar to
the relative proportion of the variation due to treatment N rate. There are two outcomes depending on whether or
effects against the variation due to chance. not there is an interaction of N rate and cultivar (i. e., that
the cultivars differed in their response to N rate). If there
The generally accepted probability level of 0.05 (5%) is was no interaction, then the response to N can be averaged
used in agricultural research as the probability that there using both cultivar means. If an interaction is observed,
could be a real difference when ANOVA indicates no then each cultivar response must be evaluated separately.
such difference. This probability level is the level of error
that scientists are willing to accept. In other words, a real
difference is so rare that it is of minimal practical concern. Mathematical Descriptions of the
If the experiment were repeated 20 times, there would be a Response (Models)
1 in 20 chance that our hypothesis would not be rejected. In statistical terms, fertilizer rate research employs various
Said another way, if the ANOVA indicates a difference levels of a quantitative variable, the amount of fertilizer. If
between one or more treatments, we are 95% certain that the ANOVA indicates a significant N treatment effect, as in

Fertilizer Experimentation, Data Analyses, and Interpretation for Developing Fertilization ... 3
Table 1, then the researcher will wish to further evaluate the but agronomically acceptable, recommended fertilizer rate,
response with the development of the mathematical model. saving fertilizer expense and reducing the risk of excessive
Responses to a quantitative variable can be statistically fertilizer applications that might endanger the environment.
inspected along the full range of the levels of the variable, These models include the logistic and the linear-plateau
and the responses to rates in between those actually applied models. Using the data in Figure 1, these three models are
in the field can be calculated. In most fertilizer experiments, illustrated in Figure 2, Figure 3, and Figure 4.
a set of 4 to 5 levels of fertilizer plus a zero-fertilizer control
is sufficient for most models. The results can be presented
graphically by an equation or model. The model can be
used to predict results if a second experiment similar to the
first were conducted. Models are typically developed with
regression analyses.

Various models can be fit to a set of data to explain the

responses. A linear model might explain a response that
continues upward or downward in a straight line within
the range of tested fertilizer rates. A linear response may
Figure 2. The quadratic model. Yield (tons/acre) = 6.86 + 0.14N -
mean the chosen range of treatments was insufficient to 0.00026N2. Yield max = 25.7 tons/acre at 270 lbs/acre N.
determine the maximum (or minimum) yield. A quadratic
response is typical of crop yield in which the response in-
creases with fertilizer rate to a point where yield approaches
a maximum but then might decrease at higher rates. In
other words, there is a point at which increased fertilizer
does not result in a significant increase in yield. Quadratic
models also typically have a linear component, meaning
that as fertilizer rates increase from low to medium rates
the yield also increases. At a certain point, the rate of yield
increase starts to stabilize or decline.

Linear and quadratic models are the simplest equations to

use for explaining crop responses to fertilizer, and they have
served scientists well as long as the main interest in the
research was maximizing yield. However, today there are Figure 3. The linear-plateau model. Yield (tons/acre) = 5.5 + 0.151N for
other goals in fertilizer research, including economics and N<129 lbs/acre (shoulder point), and Yield = 25 tons/acre (plateau) for
N>129 lbs/acre.
environmental issues. Several researchers have explored
different models for explaining crop responses to fertilizer
(see the articles in the list of references at the end of this
publication). Studies have found that the quadratic model
leads to overestimation of fertilizer recommendations
derived from responses to fertilizer (Cerrato and Blackmer
1987; Hochmuth et al. 1993a; 1993b; 1996; Willcutts et al.
1998). If the goal of the research was to select a fertilizer
rate to be used as a recommended practice, then the qua-
dratic model will usually predict a greater fertilizer need if
the maximum point from the model is taken as the putative
recommendation. The maximum yield mean is not always
significantly different from one or more means resulting Figure 4. The Logistic model. Yield (tons/acre) = 25/((1 + exp (1.12 –
from lesser fertilizer rates. If we inspect the plot of data in 0.0242N)). 90% maximum (25 tons/acre) yield (22.5 tons/acre) occurs
with 137 lbs/acre N. 95% maximum yield (23.8 tons/acre) occurs at
Figure 1, we might predict that there is little difference in 168 lbs/acre N.
yields among the fertilizer rates from 150 lb/acre or greater.
Other models have been identified that result in a lower,

Fertilizer Experimentation, Data Analyses, and Interpretation for Developing Fertilization ... 4
Researchers use statistics and mathematical models as slope of this model is not unusually steep. Third, the func-
tools to help explain crop response to fertilizer. We should tion does not pass through the origin; therefore, no negative
keep in mind that models are tools, and we should exercise yields would be predicted, nor are zero yields predicted
care in their use. The three models depicted here have with zero fertilizer added. Thus, this model accounts for
been fit to the same data set first presented in Figure 1. We native soil fertility. These attributes make the logistic model
know from the ANOVA that crop responded to fertilizer particularly useful for making fertilizer recommendations
in a significant way, but ANOVA does not identify which that avoid under- or over-fertilization.
fertilizer rate was superior. However, each model tells a
different story about the response, if we focus only on a In typical agronomic or horticultural crop yield response
model’s parameters. The most commonly used model in data, rarely are yields between 90% and 100% of maximum
agronomic and horticultural crop response research is the declared significantly (probability = 5%) different. Selecting
quadratic model (Figure 2). The quadratic model is easy to 95% of maximum yield to derive the putative recom-
derive by computer statistical packages, and most research- mended fertilizer rate would be a conservative approach
ers are familiar with it from their graduate training. Also, to ensure a most suitable fertilizer rate that would result in
the quadratic model is easily differentiated to show a peak profitable yields with due diligence in considering the risk
yield and its associated fertilizer rate. to the environment.

The problem with relying solely on the quadratic model Using the data set above, the considerations for a fertilizer
occurs on inspection of the mean yields versus fertilizer recommendation would include the following:
rate. It could be argued and can be shown by orthogonal
contrasts that there is a leveling-off of yield. Further, this • Quadratic model: The predicted peak crop response is
leveling-off occurs at a fertilizer rate less than the peak yield 25.6 tons/acre with 270 lbs/acre N.
derived from the quadratic model. In an environmentally • Linear-plateau model: The plateau yield is 25 tons/acre
aware society, perhaps researchers should not simply and the shoulder point fertilizer rate is 129 lbs/acre N.
interpret the quadratic model maximum as the putative
• Logistic model: 95% maximum yield (25 tons/acre)
fertilizer recommendation for rate.
occurs at 168 lbs/acre N, and 97% maximum occurs with
An optional model being used by scientists more frequently 190 lbs/acre.
is the linear-plateau model (Figure 3). This model also The list above shows that, depending on the level of con-
yields critical model parameters, the plateau and the servatism applied, the putative fertilizer recommendation
shoulder point. The plateau illustrates the notion that there could range from 129 to 270 lbs/acre N, a 100% difference.
is a leveling-off of crop yield response to fertilizer. However, Selecting the midpoint between the shoulder point of the
the linear-plateau model shoulder point could be argued to linear-plateau and the peak of the quadratic model or
be too conservative as a putative fertilizer recommendation. taking a conservative 97% maximum yield with the logistic
model yields similar results. This analysis yields a putative
Several recent research studies with vegetables in Florida fertilizer recommendation of approximately 200 lbs/acre N.
have illustrated the challenges with the quadratic and Choosing 200 lbs/acre instead of 270 lbs/acre as the recom-
linear-plateau models if used alone (Hochmuth et al. 1993a; mendation results in no sacrifice in yield but saves 70 lbs/
1993b). These researchers proposed using the midpoint acre of fertilizer. This is both an economic savings as well as
between the shoulder point in the linear-plateau model and a real removal of nutrient load from the environment.
the peak in the quadratic model as a putative recommended
rate. For our data, this midpoint would be 200 lbs/acre of N
fertilizer.
An Example from Actual Research
in Florida
A third model (Figure 4), the logistic model, has been The figures above are helpful to illustrate the principles of
proposed by Overman and colleagues in studies with research and data presentation. What about actual data
agronomic and vegetable crops (Overman et al. 1990; 1992; from Florida? There have been several research studies
1993; Willcutts et al. 1998). The logistic model is a reason- conducted with vegetables in Florida evaluating yield and
able compromise between the quadratic and linear-plateau fruit quality responses to fertilization with various models.
models. First, this model illustrates the law of diminishing One such study was conducted with watermelon (Figure 5).
returns. As the rate of nutrient is increased, the yield
increases until an area of diminishing returns. Second, the

Fertilizer Experimentation, Data Analyses, and Interpretation for Developing Fertilization ... 5
 reductions in fruit and vegetable quality parameters by
excessive fertilization (Hochmuth et al. 1996; 1999).

Figure 5. Graph of the data for the quadratic and linear-plateau

models for describing the response of watermelon to phosphorus
fertilizer from the same on-farm study in northeastern Florida. Note
the variation among replicates, especially with the zero-P treatment. Figure 6. The logistic equation for describing the response of
watermelon to phosphorus fertilizer from on-farm studies in
In the watermelon study, the shoulder point for the linear- northeastern Florida.
plateau occurred at 26.4 kg ha-1 P or approximately 53 lbs/
acre P2O5. The quadratic model maximum yield occurs Some Comments about Percent
with 75 kg ha- 1 P or 150 lbs/acre P2O5. Statistical analysis
(ANOVA and contrasts) of the data showed no significant
Relative Yield (RY)
difference in yield from 50 to 200 lbs/acre P2O5. The Crop responses are an integration of many different aspects
shoulder value is on the verge of steep yield reduction with of the entire production system to which the crop is
less than 53 lbs/acre P2O5, but the quadratic maximum yield exposed. Research completed during one season is affected
occurred with excessive fertilization. The authors of this by the crop integration process during that entire season,
research paper proposed using the midpoint between the as well as some antecedent contributors, such as nitrogen
linear-plateau shoulder point and the quadratic maximum mineralization from crop residue or soil organic matter.
point as a reasonable compromise fertilization recommen- The problem with crop responses associated with different
dation. In this case, the recommendation could be about experiments conducted by separate research groups, and
100 lbs/acre P2O5. This recommendation would result in often for different purposes, is that the observed crop yields
considerable savings in P fertilizer compared to the current in each of the individual experiments will display varia-
recommendation of 160 lbs/acre P2O5 for soils with low or tion. Plotting all the data from many experiments in the
very low Mehlich-1 P concentration. original units yields a scatter-graph that renders a general
interpretation very difficult. One method that can be used
Using the logistic model (Figure 6) yields a conclusion to get a sense of the crop response to fertilization across
similar to using the midpoint between the quadratic numerous studies is the percent relative yield. The highest
maximum and the shoulder point of the linear-plateau yield obtained in that particular experiment in that particu-
model. Using 97% of the maximum yield would result in a lar season is assigned as 100% relative yield. All other yields
fertilizer recommendation of approximately 55 kg/ha P or are calculated by dividing the observed yield by the highest
115 lbs/acre P2O5. actual yield and are expressed as a percentage.

There are additional reasons (beyond environmental) for Transforming the original data in this manner adds to
making recommendations closer to the conservative side the flexibility of looking at the relative yields, which have
of the response curve. There are numerous research reports been brought to a common scale. The value of this type of
about excessive fertilization, especially N, having a negative transformation is that researchers get a sense of how that
impact on yield and fruit quality. The slight depression particular crop responded to fertilizer additions throughout
in yield at excessive fertilizer rates, coupled with the cost many seasons, locations, and production practices. Relative
of the extra fertilizer, may lead to significant reductions yield should be used with caution to avoid putting too
in farm profits. Furthermore, research results have been much emphasis on this data transformation and resulting
published in the peer-reviewed literature documenting graph alone. For example, using all the RY values from
several experiments for subsequent regression can be

Fertilizer Experimentation, Data Analyses, and Interpretation for Developing Fertilization ... 6
quite misleading, especially for calculating actual yields. together. This problem is evident with the quadratic and
However, noticing that the variability among all responses linear-plateau models. Incorporating both models in the
decreases after fertilizer rate exceeds a certain range data response interpretation and calculating the midpoint
becomes quite obvious. as we have demonstrated above will consider both goals.
The logistic model appears to be the best single model at
There are a number of assumptions built into this trans- considering both economics and environmental goals.
formation process. The primary assumption is that most There is increasing accumulation of research documenting
or all of the response that we note in an RY graph is due the impacts of over-fertilization on yield and quality,
to fertilizer. There have been extensive arguments both thus reducing profits. Added to these reasons is the need
for and against making this assumption. In summarizing to protect the environment from nutrient pollution
this debate, Black (1992) indicates that the assumption related to farming activities. It becomes evident that how
can be considered valid when using the RY plot to explore research is conducted and how the data are analyzed and
variation across the years, seasons, and other production interpreted are critical to developing an informed fertilizer
practices. Black cautions the reader to avoid additional recommendation.
statistical evaluations of the RY plot due in part to its
statistical characteristics (not normally distributed) and the
true shape of the yield response to added fertilizer is site-
specific. The RY plot generalizes the site-specific variations
in nature of soil, fertilizer, climate, and plant interactions.
Problems with this generalization are avoided if the RY plot
is not used for subsequent regression analysis involving
actual yields and further interpretation.

For those who are interested in statistics, this type of

transformation also has a weighting factor based upon
the selection of the maximum yield. Again, this weighting
factor makes the assumptions above and is reduced to
insignificance by using the RY plot on a visual basis only
and not trying to further statistically analyze the regression Figure 7. Percent relative yield for drip-irrigated watermelon response
by other means. Black (1992) states that while these objec- to increasing N rates across several studies in Florida. The number in
parentheses after the indicated year of the report is the year in which
tions are worthy of note, the RY plot can be a useful tool in that experiment was conducted (Hochmuth and Hanlon 2010b).
fertilizer research.
References
To further illustrate the usefulness of the percent relative
Black, C. A. 1992. Soil Fertility Evaluation and Control. Boca
yield approach, watermelon yield is plotted in Figure 7.
Raton, FL: Lewis Publishers.
Note that the yields increase in all experiments and then
tend to level off somewhere between 100 and 200 lbs/acre Bullock, D. G., and D. S. Bullock. 1994. “Quadratic and
N. The current UF/IFAS N recommendation is 150 lbs/ Quadratic-plus-plateau Models for Predicting Optimal
acre N. While this graph was not used to set the UF/IFAS Nitrogen Rate of Corn: A Comparison.” Agron. J. 86:191-5.
recommendation, the graph indicates that the recommen-
dation is reasonable and supported by research. Cerrato, M. E., and A. M. Blackmer. 1987. “Comparison of
Models for Describing Corn Yield Response to Nitrogen
Summary Fertilizer.” Agron. J. 82:138-43.
Crop fertilizer response research should be carefully
conducted to account for the economics to the grower and Dahnke, W. C., and R. A. Olson. 1990. “Soil Test Correla-
protection of the environment from nutrient losses due tion, Calibration, and Recommendation.” In Soil Testing
to excessive fertilization. There are several mathematical and Plant Analysis, 3rd edition, edited by R. L. Westerman,
models to describe crop yield response to fertilizer, and 45-71. Madison, WI: Soil Sci. Soc. Amer.
these models should be employed with caution. Using a
Hochmuth, G. J., E. E. Albregts, C. C. Chandler, J. Cornell,
single model to explain crop response may not account
and J. Harrison. 1996. “Nitrogen Fertigation Requirements
for economics and potential environmental impact

Fertilizer Experimentation, Data Analyses, and Interpretation for Developing Fertilization ... 7
of Drip-irrigated Strawberries.” J. Amer. Soc. Hort. Sci.
121:660-5.

Hochmuth, G. J., J. K. Brecht, and M. J. Bassett. 1999. “N

Fertilization to Maximize Carrot Yield and Quality on a
Sandy Soil.” HortScience 34(4): 641-5.

Hochmuth, G. J., J. Brecht, and M. J. Bassett. 2006.

“Fresh-Market Carrot Yield and Quality Responses to K
Fertilization of a Sandy Soil Validated by Mehlich-1 Soil
Test.” HortTechnology 16:270-6.

Hochmuth, G. J., and E. A. Hanlon. 2010a. Principles of

Sound Fertilizer Recommendations. SL315. Gainesville:
University of Florida Institute of Food and Agricultural
Sciences. http://edis.ifas.ufl.edu/ss527.

Hochmuth, G. J., and E. A. Hanlon. 2010b. Summary of

N, P, and K Research with Watermelon in Florida. SL325.
Gainesville: University of Florida Institute of Food and
Agricultural Sciences. http://edis.ifas.ufl.edu/cv232.

Hochmuth, G. J., E. A. Hanlon, and J. Cornell. 1993a.

“Watermelon Phosphorus Requirements in Soils with Low
Mehlich-1 Extractable Phosphorus.” HortScience 28:630-2.

Hochmuth, G. J., R. C. Hochmuth, M. E. Donley, and E. A.

Hanlon. 1993b. “Eggplant Yield in Response to Potassium
Fertilization on Sandy Soil.” HortScience 28:1002-5.

Nelson, L. A., and R. L. Anderson. 1977. “Partitioning of

Soil-test Response Probability.” In Soil Testing: Correlation
and Interpreting the Analytical Results, spec. publ. 29, edited
by T.R. Peck, J.T. Cope, and D.A. Whitney, 19-38. Madison,
WI: Am. Soc. Agron.

Overman, A. R., F. G. Martin, and S. R. Wilkinson. 1990.

“A Logistic Equation for Yield Response of Forage Grass to
Nitrogen.” Commun. Soil. Sci. Plant Anal. 21:595-609.

Overman, A. R., M. A. Sanderson, and R. M. Jones. 1993.

“Logistic Response of Bermudagrass and Bunchgrass
Cultivars to Applied Nitrogen.” Agron. J. 85:541-5.

Overman, A. R., and S. R. Wilkinson. 1992. “Model Evalu-

ation for Perennial Grasses in the Southern United States.”
Agron. J. 84:523-9.

Willcutts, J. F., A. R. Overman, G. J. Hochmuth, D. J.

Cantliffe, and P. Soundy. 1998. “A Comparison of Three
Mathematical Models of Response to Applied Nitrogen: A
Case Study Using Lettuce.” HortScience 33:833-6.

Fertilizer Experimentation, Data Analyses, and Interpretation for Developing Fertilization ... 8
Table 1. Analysis of variance for the data in Figure 1, testing crop response to rate of N fertilizer. In this case, the experimental
design was a randomized, complete-block design with 5 replications.
Source of variation Degrees of freedom Sums of squares Mean squares F value
N rate 8 1655.4 206.9 163 (P<.0001)
Replication 4 1.5 0.4 0.3 (P=0.87)
Error 32 40.4 1.3
Total 44 1697.4

Fertilizer Experimentation, Data Analyses, and Interpretation for Developing Fertilization ... 9