Quantifying Diversity Flashcards
measures of diversity?
Species richness
(how many different sp. in particular place)
Species evenness
How evenly distributed is abundance among species
Phylogenetic diveristy
How much of the evolutionary history (tree of life) is represented?
Measuring species richniess
How many species (or genera, orders…)
- identify specimens one by one
- is it a new species?
yes = increment count by 1 - plot no. of species against no. of specimens identified
makes an accumulation curve
shows how many individuals need to be identified before new species is found
new species finding gets less and less frequent as go through sample
can use mathematical model to estimate where line would flatten
i.e. how many individuals would you have to identify before stop finding new species
don’t really have to do that - curve approaches limit anyway (asymptote?)
pitfalls in random sampling
biases:
flying vs non flying (how easy are things to catch?)
collecting each specimen one time
biases in data entry into spreadsheet - can affect accumulation curve - so randomise specimen order in dataset? ( need number of new species identifying to diminish to get asymptote on graph)
identification takes time and skill
^lots of effort to quantify species richness
species evenness
can have higher no. of species (higher richness)
but if there are e.g 100 members of one species in sample
vs only on or two of each
not very even - very dominant species (so is it very diverse?)
^hyperbolic numbers in example
can have less richness in other sample - byt very similar number of individuals from each species - much more even sample
How even is the abundance of different species in the sample
shannon diversity
(formula on notes?)
take proportion of species in sample (no. of that species/total specimens)
raise proportion to power of itself
take product of all of these
D=1/product
Maximum D is the no of different species
closest to maximum the higher the species richness is
so higher richness - higher D
also Higher evenness - higher D
simpson’s diversity?
proportions are raised by 2 instead of by themselves
gives probability that 2 specimens selected at random will be the same species
inverse simpson’s:
maximum 1/lambda is number of specimens (N) when proportions are all equal (same as Shannon D)
approaches 1 as specimens are clustered into a single species (same as shannon D)
less affected by dominant species than shannon D
Phylogenetic diversity
how much independent evolution does the sample capture? (independent branch lengths) diagram illustrates well
more diverse when more phylogenetic history captured
more recent splits between sample specimens = less phylogenetic diversity captured
capturing “deep nodes” in the tree
Lineage through time plot?
plot time on the x axis(Myr ago?)
against no. of different lineages in phylogeny at that point
phylogenetic diversity and DNA sequencing?
DNA barcoding (simpler taxonomic identification [see earlier]), objective and replicable
small random samples captures deep diversity with high probability
avoids need to define taxonomic groups (species, orders families…) as COX1 sequence similarity is basis?
can still be biased by non-random sampling biases (e.g. winged creatures harder to capture - so underrepresented in sample - does not reflect reality)
bias in tree of life
large proportion of it are vertebrate and invertebrate animals - due to bias in sample collection
diversity sampling also biased by what can be seen
microevolution
evolution over matter of generations
considers processes:
natural selection, drift, mutation, and dispersal between populations (in same species??)
can be studied in lab
macroevolution
evolution at or above species level
takes place over geological time scales
major processes in this field:
speciation
extinction
range dynamics
trait evolution
studied through inference of process through patterns left behind
(what patterns does evolution leave behind, what can be inferred about extinction…)
paleontological data
same type of fossil found in one rock layer (stratum)
same thing found in later stratum
can infer that it:
-originated some time prior to first fossil
-went extinct some time after last one
-existed in time period between them
its STRATIGRAPHIC RANGE
paleontologists bin the time periods (time bins)
how many genera/families existed in this bin?
can look across periods to see the number of taxa through time
(taxon richness in different time periods)
pitfalls of paleontological data
lots of variation exists within families (e.g. ducks and geese in same family)
many things used to distinguish these species within families is not represented (not preserved well) in fossil record
-plumage
-behaviour (song, courtship…)
so more specific diversity information can’t be conferres/is lost so can’t get full picture of diversity
sequence data and inferring past diversification patterns
huge increase in sequence data available (growing technology)
form molecular phylogeny
could contain signature of past diversification
phylogenetic patterns that inform about past diversification
tippy tree:
short branches early on
many longer branches later
rapid diversification early on
slows down
stemmy tree:
fewer, longer, branches early on
slower diversification early on
diversification speeds up later on
many shorter branches at ends
balanced tree:
even rate of among branches
imbalanced tree:
much more branching in some branches than others in the tree
LTT and phylogenetic pattern:
LTT on log scale:
linear
on log so this means slower diversification earlier and faster later
TIPPY TREE
learning about extinction?
fossil diversity curves
number of taxa found in different time periods
sharp downturns in marine families correlate to extinctions
however some people saw these downturns and said they were sampling artifacts/biases
mass extinctions - exceptional/artifacts?
researchers:
took time bins
how many taxa went extinct in each bin?
did a regression - see how extinction rates were changing over time
got measure of how extreme the extinction rate would have to be to be an “exceptional” mass extinction - separate from just the background extinction rate
found 4 of 5 mass extinctions were above
somthing different going on for them
though also showed background E rate decreasing over time - which is not true
not reason why we see more diversity today
mass extinction unpredictability
mass extinctions are unpredictable as in which groups will succeed and which won’t
there are things we can’t understand about macroevolution from just scaling microevolution up (theory popular in 80s, 90s, not so much now, ok??? and???)
marine bivalve groups extinction
species rich groups survive longer
the more widespread the distribution of species in the group - the longer it survives
would predict that widespread speceis rich groups would survive mass extinction better
however there was not really a difference between groups for proportion surviving and proportion not
only general correlate of mass extinction survival is range size of clade (not species range size) as if clade found everywhere - more likely to survive
ecological restructuring after mass extinctions
if look at motile vs non motile groups separated - no pattern
just diversification increasing towards present
however if look at proportion of motile:non-motile groups
=proportion of motile increased towards present with big step changes at mass extinctions
mass extinctions restructuring ecology towards more motile species after MEs
same pattern in proportions of predator vs prey species with prey taxa step increasing at MEs
background extinction - Phylogenetically clustered?
looking at marine bivalves
background extinction rates cluster in phylogenetic groups
more extreme for some groups than others in different time periods
-clustering of extinction
perhaps shared attributes within families that make them more susceptible to extinction at that time than other families
ONLY REFERS TO BE not MEs
MEs still unpredictable
Red Queen Hypothesis
plot survival time
against no. of genera that have said survival time
approximately linear (negative slope)
consistent where if all taxa have constant probability of extinction
due to competing in zero-sum world
with all taxa competing at same time
this co-adaptation gives constant deterioration of environment for a taxon
due to e.g. predators adapting to be better, competing groups adapting to be better at competing
need to constantly adapt in order to stay at same level in environment and not go extinct
red queen - running in place
molecular phylogenies to study extinction
Pure birth model
Birth-death model
pure birth model:
simplest process phylogeny
every lineage has same probability of speciating at any moment
+ no extinction
-branches fairly constant size throughout
-balanced
-exponential growth - shows up on log plot LTT as positive gradient straight line
stork phylogeny from earlier - doesnt exactly show this - shows asymptotic instead
Birth - death model (birth minus death)
each lineage has constant rate of speciation AND extinction
on LTT plot:
Myr ago vs no. of taxa
on log plot:
linear increasing
BUT shows UPTICK before present
because recently speciating branches have had little time to go extinct
[PULL OF THE PRESENT]
steepness of this upturn gives a good estimation of speciation rate
lower less steep part of plot can give estimate of birth minus death rate
and so can use this model to estimate speciation AND extinction rates with mol. phylogeny
HOWEVER this shape of LTT is not consisten with what we see with the storks
so can rarely estimate extinction rates from molecular phylogenies and must use fossil record
other processes could also produce above LTT plot
biases in fossil record raw diversity curves
-some taxa fossilise better than others (e.g. bird skeletons bad, least soft tissues and most hard skeletons best)
-fossils not laid down at constant rate - some time periods have better conditions for fossilisation (bias toward them in record)
-preservation of already laid fossils is much better in some periods than others
-more recently laid down fossils have had less time to degrade - more likely to have survived to present
“pull of the recent”
some upturns in some phylogenies’ LTT patterns coud be due to increased diversification OR just pull of the recent
AMBIGUITY
-sampling bias - some periods have more attention than others
so have different sample sizes for different time periods
correcting for inconsistent sample sizes?
RAREFACTION
randomly take a specified number of specimens from each sample
(e.g. 3 from each time period sample)
count no. of genera in each
repeat many times at this specified number
now repeat these steps for different specified sample sizes
can now plot average number of genera found in each specified sample size
can allow for correcting for sample size differences
= fair comparison between time periods
after rarefaction:
in predicted number of genera over time plot
-no longer see overall increase in diversity towards present (pull of present corrected for)
-mass extinctions look less pronounced (except end permian)
in curve where global temps and invertebrate diversity where correlated
no correction = more in colder
after correction - flipped to more in warmer
rarefaction - throwing baby out with bathwater
amount of diversity correlates with amount of sedimantary rock:
because:
- more sediment = increased chance of preservation and therefore overrepresentation
-> this correlation between sediment and diversity reflects a causal relationship. this is a bias to certain periods and needs to be corrected for - large scale geological processes exist that affect both amount of sediment and macroevolution
->this is an indirect correlation (not causal) between sediment and diversity caused by a third variable
this correlation should not be corrected for as it throws out important factors of diversity
diversity feedback on diversification rates
as diversification happens
niches get filled
opportunity to diversity into those niches is reduced
=> reduced diversification rates
niches get filled
then competition prevents species from extending range
reduces opportunity for speciation via range splitting (allopatric speciation) and increasing extinction
=> reduced rates
DIVERSITY DEPENDENCE
shows on phylogeny as much shorter branch lengths early on in tree that slow down later
(more tippy tree?)
exponential diversification
if there was no feedback on diversification rates
might be moments on plot where rate looks exponential
but not much - doesn’t explain diversity over time
the logistic equation
model that allows for negative feedback of diversity on diversification rates
has variable K - carrying capacity (no of species/individuals resources can support in that area)
asymptote on graph^
when N=K
diversification rate=0
fossil record evidence for diversity dependence?
look at origination rates of new taxa and extinction rates of old ones
new genera formation peaks just after mass extinctions
backs up fact that mass extinctions allow more rapid diversification
ME brings N much lower than K
allowing for much faster diversification
(niches emptied, ranges can extend too)
higher diversity - lower diversification
vice versa
evidence for Diversity dependence - Dendroica(passerine) phylogeny
root to tip of phylogenetic tree:
slowing down pattern (longer branch lengths later on)
diversity dependent model gives better results over just pure birth-death model
same shown with stork phylogeny
evidence for Diversity dependence -
Anolis lizard adaptive radiation
radiated into lots of niches which fill different parts of trees
across different islands in Caribbean the same niches on tree were filled by different Anolis species by convergent evolution
can reconstruct when different lizards reached different islands
slow down at different rates on different islands
samller islands slow much more rapidly
because thay have lower K
area big factor on carrying capacity
other factors causing branching rate to slow down?
-incomplete sampling
-failure to recognise youngest species (which would be later in tree) have not speciated far enough to be recognised as different - slow down illusion
-Peripatric speciation
speciation via dispersal
dispersal to different isalnds
speciate awat from each other
gives linear plot on LTT, looks like slow down when plotted on log
-gene tree vs species tree - using enetic data makes species branching events look earlier
pools later speciation events earlier - slow down illusion
why are some species winners?
Hoatzin - 1 lineage
Passerines - 5000 species in similar time frame
what makes one group diversify more than the other?
winners v losers?
does more variation among clades in species richness exist than expected by random?
YES
there were transitions both to faster div rates and to slower div rates
why do diversification rates vary among taxa?
key innovations
e.g. phytophagy
many insects feed on plants - but characteristic is restricted to a few orders
-> phytophagy appears hard to evolve
hypothesis: once phytophagy evolves there is a huge diversity of plants available to feed on
hard for other groups to access
ADAPTIVE ZONE - gives opportunity for adaptive radiation (and so higher div rates)
species richness and diversification rates vs latitude (latitudinal diversity gradient)
higher species richness in tropics
BUT
higher diversification rates in poles (calculated on mol. phylogeny)
potentially because more niches are available in poles becasue species are farther from their carrying capacity there (because polar environment newer than tropics)
fossil record - evidence for LDG?
scenario 1.
in greenhouse environment- lower extinction at higher latitudes
during global warming event - less decline in polar due to said lower extinction rate
scenario 2.
extinction rates increase in tropics during G warming event
due to too high temps
and that is why peak divesity shifts polewards during greenhouse
due to peak in tropics only in icehouse conditions