*A simple ecological study:
Proposed studies. ***Read the opening chapters in your lab manual
(read all the sections on experimental design and hypothesis testing) and the
material below before coming to class, and come prepared to go out in the field.
** If you need to review mean, standard error, and t-tests, do it. We will
meet in the lab for a brief discussion of the scientific method and a planning session for
our field work. Then, we will go to the nature reserve and you will begin making observations for your simple ecological study. Some of you may begin collecting data
this week, but you will also have next week to continue. At the end of lab you will
__ Hand in__: Fill in the worksheet

**General
Ecology Lab #4 ** **A Simple Ecological Study **

In
this lab you will conduct a simple ecological study that your group will initiate, design,
and carry out over the next two labs. The
goals of this lab are for you to understand basic ecological studies, design a sampling
regime, conduct statistical analysis of your results, and prepare a concise report of what
you did.

__Designing
an Ecological Study__: An ecological study usually begins with an
observation of a pattern in nature. For
example, you might observe that red squirrels are common in one area, while gray squirrels
dominate another. At this point your initial
observation is just subjective; you need to conduct a formal descriptive study to document
the pattern you observed. For this
descriptive phase you first need to formally state your question or *descriptive
hypothesis*: Are red squirrels dominate in
Area A? (or, stated as a hypothesis: Red
squirrels are dominate in area A, while gray squirrel are dominate in area B.) Now you would need to design a sampling protocol
that would accurately assess the squirrel populations (see section on Ecological Sampling)
followed by statistical analysis of the data (see section on statistical analysis).

*functional hypothesis* that can account for the pattern you documented. For example, you might suspect that the tree
species present determined the distribution of the squirrels, or perhaps the presence of a
predator or disease influenced the squirrels in different ways. You will need to state a functional hypothesis
that is phrased to include a cause and effect relationship. For example,
you might state your hypothesis as "if oaks are the dominate tree species in an
area, then gray squirrels will be the more abundant than in areas dominated by
pines". These "if, then" statements are a common way to express a
functional hypothesis, but in ecology it is also common to state your hypothesis
by posing a question to be addressed. For example, "are gray squirrels
more common in forest areas dominated by oaks than in forest areas dominated by
pines"? In ecology the "null hypothesis" unstated even though it is always
implied. After forming a functional hypothesis you will need an experiment
to test it. Perhaps you could introduce gray
squirrels to area A; do they survive?

__Making
initial observations:__ Many comparisons can be made, but in this lab you
should focus on *comparing one species in two
different situations*. For example,
compare the individuals of one species in different habitats. Some features you could compare:

*abundance*--the density of individuals in a habitat
might depend on the level of some resource that varies in distribution

*morphology*--the size, shape, or color of
individuals may be influenced by the environment (or you could have genetic differences in
different habitats).

*reproductive
output*--the
number of offspring produced is a critical ecological feature, and it could be influenced
by different habitats or the presence or absence of predators, herbivores, or parasites

__Formulate
a hypothesis__: For the descriptive phase of your study, your
hypothesis will be simple: e.g., The number
of individuals of Japanese honeysuckle is greater in the shade than in the sun. A *functional hypothesis* will include a mechanistic
explanation: e.g., Seed germination in sunny habitats is limited by rapid desiccation
compared to shade.

__Ecological
sampling__: Ecologists often gather information about a
population (i.e., number of seeds produced), community (i.e., number of species), or a
habitat (i.e., organic matter in the soil). We
can't measure the seed production of every individual plant in the field, so we need to
sample--we take a *statistical sample. *

You must know
what you are sampling; e.g., if you measure the seeds on an individual in August, have you
missed seeds that were produced back in June and have since fallen off? have some seeds
been eaten by mice? will more seeds be
produced in September? If you want to sample
seed production by an individual plant, you must be sure to actually sample what you think
you are sampling.

You must not
bias your samples. Which individuals will you
sample and which will be ignored? Random
sampling involves use of a random number table to select random plots to sample. The table will give you a number--say 6-- and you
sample plot 6. The problem with random
numbers is that everything you want to sample must have a number (i.e., you need to grid
off the study area so all points are numbered). Haphazard
sampling involves, for example, throwing a stake out and sampling the closest plant.

Your samples
must be replicated.

__Statistical
analysis of data__:

**Two-sample comparisons:
**This procedure allows you to draw conclusions about the similarity or
difference between the means of two sampled populations.
For example, you might compare the seed output of smooth sumac growing in
the shade and in the sun.** **Calculate the mean
(x), standard deviation (s), and standard error (sx
or SE) for each sample. Use the standard
deviation to express the population variability and the standard error to provide an
estimate of the interval where the true population mean is to be found (the interval: x __+__ 1.96 * standard error, will contain 95%
of the sample means for large samples.

A student's *t*-test will allow you to statistically contrast the
two samples--in other words objectively answer the question of whether the samples are the
same or different. Small *t* values indicate a high probability that the two
population means are the same; by contrast, large *t*
values imply lower probability. The
calculated *t* value and probability (*p*)are given in SYSTAT. Ecologists consider a difference between
population means to be significant when the probability that they are the same is less
than 5% (*p* < .05)

**Regression
analysis: **You
can use a regression analysis to constrast two different variables measured on one
species. For example, you might observe that
a taller plant produced more fruit than shorter plants.
You could test this descriptive hypothesis by measuring the height of the
plant and the number of seeds it produced. Your
lab manual will tell you how to do a regression analysis, and it is easy to do in SYSTAT.