Statstutor community project
www.statstutor.ac.uk
stcp-dataset-tomato_des
Tomato Rooting Experimental Data:
An experimenter is interested in the genetics of the rooting properties of tomato plants. He has
been working with a population of near isogenic lines (i.e. almost genetically identical), which have
been derived from a cross between an inbred commercial cultivar, M82, and a wild relative. The
lines are genetically identical to M82 except for one region (different for each line) where the lines
have the genes of the wild relative. He has identified two lines (called And B here) which appear to
differ from M82 in some rooting properties. He performs two experiments to examine in more
detail the rooting properties of these two lines relative to M82, and the relationship between root
and canopy development.
a) In the first experiment he studies the ability of roots to penetrate a barrier. He grows
tomato plants in individual plastic tubes of compost 30cm tall, with a membrane across the
bottom of the tube which resists root growth. When the plants are a few weeks old, for
each plant he counts the roots which have penetrated the membrane, records the length by
which the longest root has penetrated the membrane, and then cuts the roots off
immediately below the membrane and weighs them. He also cuts off the above ground
canopy of the plant at the top of the tube (soil level) and weights that. He oven dries both
the excised toots and the canopy and later weights the dried material
b) In the second experiment he grows the plants hydroponically (not in soil) in individual tall
(2m long) tubes. When the plants come into flower he removes them from their tubes and
measures the length of the longest root. He then cuts the roots into four sections, each
having one quarter the length of the longest root, so that section one is the quarter of the
roots closest to the surface and section four is the quarter furthest from the surface. These
sections are dried and weighed. The above-ground canopy is also weighed both before and
after drying.
The data are:


Experiment 1
o Line (variety)
o Number of penetrating roots
o Penetrating root fresh weight (g)
o Penetrating root dry weight (g)
o Canopy fresh weight (g)
o Canopy dry weight (g)
o Root length below the membrane (mm)
Experiment 2
o Line (variety)
o Final root length (mm)
o Root dry weight (section 1) (g)
o Root dry weight (section 2) (g)
o Root dry weight (section 3) (g)
o Root dry weight (section 4) (g)
o Canopy fresh weight (g)
o Canopy dry weight (g)
Categorical
Discrete
Continuous
Continuous
Continuous
Continuous
Continuous
Categorical
Continuous
Continuous
Continuous
Continuous
Continuous
Continuous
Continuous
Data contributed by Andrew Mead, University of Warwick
Statstutor community project
o
Total root dry weight (g)
www.statstutor.ac.uk
stcp-dataset-tomato_des
Continuous
There are a number of possible response variables, though main focus is the effect of line on rooting
properties, and the effect of rooting properties on canopy development.
Can be used for:
Method/Test
Descriptive
statistics
Box-and-whisker
plots
Confidence
intervals
Normal
probability
calculations
Correlation
coefficients
Scatter plots
One-sample ttest
Two-sample ttest
F-test
One-way ANOVA
Simple linear
regression
Multiple linear
regression
Simple linear
regression with
groups
Multiple linear
regression with
groups
Questions
What is the mean root length for each line?
What is the variability in penetrating root dry weight for each line?
Compare the distributions of penetrating root dry weights between the lines
Construct an interval within which the true (population) mean total root dry
weight for M82 would fall (with 95% confidence)
Assuming that canopy dry weight is Normally distributed, calculate the
proportion of plants in the population with a canopy dry weight greater than x g,
or is between x and y g
Is there a linear association between canopy dry weight and root dry weight?
Is there a linear association between root fresh weight and root dry weight?
Is there a linear association between penetrating root dry weight and number of
penetrating roots?
Is there an association between canopy dry weight and root dry weight?
Is there an association between root fresh weight and root dry weight?
Is there an association between penetrating root dry weight and number of
penetrating roots?
Are the roots of N82 grown hydroponically (second experiment) of a similar
length to what we would expect for plants grown in soil?
Are there differences in the mean root lengths between lines A and B? (could
include both the equal variance case and the unequal variance (Welch test) case
Are there differences in the variances of root lengths between lines And B? (also
directly associated with the assumption of equal variance for the two-sample ttest)
Are there differences between the mean root lengths for the three lines (M82,
A, B)?
Is the size of the canopy (canopy dry weight) linearly related to the amount of
roots (total root dry weight)?
Is the penetrating root dry weight linearly related to the number of penetrating
roots?
Is the root fresh weight linearly related to the root dry weight?
Can the prediction of the size of the canopy be improved by considering the
weights of roots at different depths below the soil surface?
Does the relationship between canopy dry weight and root dry weight vary
between lines?
Does the relationship between root fresh weight and root dry weight vary with
line?
Does the relationship between penetrating root dry weight and number of
penetrating roots vary with line?
Does the relationship between canopy dry weight and the weights of roots at
different depths below the soli surface vary with line?
Data contributed by Andrew Mead, University of Warwick