Lecture 5 - ghost crabs Flashcards
what is ecology?
> the study of the interaction that determine the distribution and abundance of organisms
what are interactions?
> between organisms and their abiotic
between organisms and others of the same species
between organisms and others of different species
what is the distribution of organisms like?
> scale dependent
random
uniform
clumped
what is dispersion?
> the pattern of spacing of individuals within a species
random:
- position of individuals is random with respect to each other
uniform, over-dispersed, hyper-dispersed, regular - further apart than ‘expected’
under-dispersed, clumped, aggregated - closer together than ‘expected’
what are some biological reasons behind dispersion patterns?
> random
- when positions of individuals are not affected by others (often used as a null model or expected)
uniform
- individuals avoid contact or die if they are close (intra-specific competition)
Clumped
- benefit from contact or use same habitat, resources, herd
what is a good test to do to measure level of distribution?
> if the number of individuals in ‘plots’ are randomly distributed then the frequency distribution of these counts, x, will follow a poisson distribution (with lambda = mean = variance)
what is an example of how to calculate distribution with crab holes?
> compare actual number of quadrats with x crabs holes with expected number calculated using a poisson distribution
if actual and expected distributions are similar then crab holes are randomly distributed in the sample guadrats
what is a simpler way to analyse dispersion? (not poisson)
> variance is a measure of the spread of a set of values about its mean
variance = sum of (xi - mean) squared / (n-1), where xi is a measure of the variable, n is the number of measures
for poisson distribution the variance = mean
what are Nearest neighbour distances?
> for Random, distance between a randomly chosen individual and its nearest neighbour should be the same as the distance between a random point and the nearest individual
for clumped, NN distances will be smaller than random point
what is sampling theory?
> how many are there? estimating population size, p
known area A? known density D (numbers per unit area), then P=DxA
take n small areas each of area a, count or census xi in each and estimate density D, D= sum of xi/n
how do you select quadrat locations, and what are some sampling techniques?
> systematic
- easy, but can be biased
simple random sampling
- proper random samples are hard to take
- random = haphazard
- need a map, random number table and gps
stratified random sampling
- good if you know there are high density and low density patches and you can map them
- sample separately, combine estimates later
cluster samples
- efficient and good for removing unconscious bias
- but sample size is number of clusters
how do you work out how many samples to take?
> balance of precision vs cost
sampling error = 2x standard deviation / sqrt(sample size)
but standard deviation is not known in advance
sample size of >15-20 allows large sample statistics to be applied
doubling precision requires 4 times sample size
what are the two types of sampling methods?
> line transects
point transects
only a portion of organisms are detected, Pd
the D = (n/2wl)Pd
Pd can be estimated from distance data. this is critical
