Chi Squared Test

Skill:
Testing for association between two species using the chi-squared test with data
obtained by quadrat sampling

The presence of two species within a given environment will be dependent upon
potential interactions between them

If two species are typically found within the same habitat, they show
a positive association
Species that show a positive association include those that exhibit predator-
prey or symbiotic relationships

If two species tend not to occur within the same habitat, they show
a negative association
Species will typically show a negative association if there is competition for the same
resources
One species may utilise the resources more efficiently, precluding survival of the
other species (competitive exclusion)
Both species may alter their use of the environment to avoid direct competition
(resource partitioning)

If two species do not interact, there will be no association between them and their
distribution will be independent of one another

Quadrat Sampling
The presence of two species within a given environment can be determined using
quadrat sampling
A quadrat is a rectangular frame of known dimensions that can be used to establish
population densities
Quadrats are placed inside a defined area in either a random arrangement or
according to a design (e.g. belted transect)
The number of individuals of a given species is either counted or estimated via
percentage coverage
The sampling process is repeated many times in order to gain a representative data
set

Quadrat sampling is not an effective method for counting motile organisms it is

used for counting plants and sessile animals
In each quadrat, the presence or absence of each species is identified
This allows for the number of quadrats where both species were present to be
compared against the total number of quadrats
 Quadrat Sampling Method

Chi-Squared Tests
A chi-squared test can be applied to data generated from quadrat sampling to
determine if there is a statistically significant association between the distribution of
two species
A chi-squared test can be completed by following five simple steps:
Identify hypotheses (null versus alternative)
Construct a table of frequencies (observed versus expected)
Apply the chi-squared formula
Determine the degree of freedom (df)
Identify the p value (should be <0.05)

Skill:
Recognising and interpreting statistical significance

Example of Chi-Squared Test Application

The presence or absence of two species of scallop was recorded in fifty quadrats
(1m2) on a rocky sea shore
The following distribution pattern was observed:
6 quadrats = both species ; 15 quadrats = king scallop only ; 20 quadrats = queen
scallop only ; 9 quadrats = neither species
 Step 1: Identify hypotheses
A chi-squared test seeks to distinguish between two distinct possibilities and hence
requires two contrasting hypotheses:
Null hypothesis (H0): There is no significant difference between the distribution of
two species (i.e. distribution is random)
Alternative hypothesis (H1): There is a significant difference between the distribution
of species (i.e. species are associated)

Step 2: Construct a table of frequencies

A table must be constructed that identifies expected distribution frequencies for each
species (for comparison against observed)
Expected frequencies are calculated according to the following formula:
Expected frequency = (Row total Column total) Grand total
 Step 3: Apply the chi-squared formula
The formula used to calculate a statistical value for the chi-squared test is as follows:

Where: = Sum; O = Observed frequency; E = Expected frequency

These calculations can be broken down for each part of the distribution pattern to
make the final summation easier

Based on these results the statistical value calculated by the chi-squared test is as
follows:
2 = (2.20 + 2.38 + 1.59 + 1.73) = 7.90

Step 4: Determine the degree of freedom (df)

In order to determine if the chi-squared value is statistically significant a degree of
freedom must first be identified
The degree of freedom is a mathematical restriction that designates what range of
values fall within each significance level

The degree of freedom is calculated from the table of frequencies according to the
following formula:
df = (m 1) (n 1)
 Where: m = number of rows; n = number of columns
When the distribution patterns for two species are being compared, the degree of
freedom should always be 1

Step 5: Identify the p value

The final step is to apply the value generated to a chi-squared distribution table to
determine if results are statistically significant
A value is considered significant if there is less than a 5% probability (p < 0.05) the
results are attributable to chance

When df = 1, a value of greater than 3.841 is required for results to be considered

statistically significant (p < 0.05)
A value of 7.90 lies above a p value of 0.01, meaning there is less than a 1%
probability results are caused by chance
Hence, the difference between observed and expected frequencies are statistically
significant

As the results are statistically significant, the null hypothesis is rejected and the
alternate hypothesis accepted:
Alternate hypothesis (H1): There is a significant difference between observed and
expected frequencies
Because the two species do not tend to be present in the same area, we can infer
there is a negative association between them

Practice Question
Two species of fir tree are found along the coast of Southern California.
These two tree species are the Grand Fir (Abies grandis) and the Noble Fir
(Abies procera).
Their distribution patterns were establsihed via 150 quadrat samples, yielding
the following results:
25 = both present ; 30 = Noble Fir only ; 45 = Grand Fir only ; 50 neither
present
Activity: Use the chi-squared test to determine if these two plant species show association.