Evolutionary Ecology Flashcards
What did this study find?
Compared EQ to mean group size across primate phylogenies
Cassien et al. 2017
Mean group size was not a good predictor of brain size
Diet seemed a stronger predictor
Some diets may need greater cognitive ability like frugivores
Higher energy turnover & higher quality diets may lead to larger brains.
It is an example of the comparative approach, and showed that ecology could be more of a predictor of brain size than sociality
Define the Comparative Approach
Studying patterns and processes across different species or populations to understand the evolution of traits, behaviors, and ecological interactions. By examining these variations, researchers can infer evolutionary relationships, adaptations, and the selective pressures that have shaped ecological strategies.
What did this study find out? What is it an example of?
Looked across the Americas at nest predation in different species
Predicted that where physical conditions are easy, interrelationships between species would become paramount adaptive problem. Thus tropics would have more predation
Freeman et al. 2020
An example of the comparative approach again
Found very little latitudinal trend in predation
HOWEVER found that predation rate was negatively correlated with nesting period - longer nesting, less daily predation rate
BUT for same nesting period length, tropical species would experience higher daily nest predation rates than temperate species - Concluded that adaptation is dampening geographic patterns in interaction rates
Define adaptive radiation theory
Predicts that niches are partitioned early in the clade history in the presence of ecological opportunity
Cetacean study found – evidence for declining diversification rates in cetaceans overtime, and instead found the rate to be quite s—-.
Study supported evidence for an early burst in d—— but NOT d——-
Slater et al. 2010
No evidence, cetaceans, steady.
Disparity within clades was lower than expected with a constant-rate model
disparity, diversity
What is correlated evolution?
Tries to answer if changes in one trait are correlated with changes in another
Use statistical analysis to see if the differences observed are down to chance or not
Regression or correlation analysis is the most common form of comparative analysis
Why may we begin to use comparative methods instead of intraspecific studies alone? Give a classic example of this
- Including more species can create more variation in life history and ecology
to better test a theory - A typical example of this would be Tinbergen’s gulls in 1962, although this was just 2 species.
- Black-headed gulls were ground nesting, with cryptic nests & chicks, and removed egg shells. They had higher chick mortality
- Kittiwakes were cliff nesting, didn’t remove shells, and had more obvious nests and chicks
Give an example of the comparative approach applied across mammal taxa
Harvey et al. 1991
Correlated BMR and size across many mammal taxa
Found positive correlation across family, genus and species levels
Give some general pros of the comparative approach? 4
- Very general
- Incorporates large amount of variation
- Incorporates large amounts of data, which can be found in online databases
- Can guide experimental / observational work
Give the 2 main problems of using phylogeny in the comparative approach
- Pattern vs process - black and polar bears are different colours, but we still don’t know what process drove that pattern (if we don’t know what the ancestor was like)
- Statistical Non-Independence - Could result from taking data from same place, same animal and same time. Species may show similar traits just due to relatedness
Give an example of pattern vs process in a real study
Pagel & Meade 2006,
Mating systems in old world primates.
Tried to see if female oestrous displays correlated with multiple partners
Looked at phylogeny to see what came first and come to a conclusion
What could cause statistical Non-independence in the comparative approach? How can you control for this?
Could result from taking data from same place, same animal and same time. Species may show similar traits just due to relatedness. This may show different correlations across different taxa even if individual taxa show different ones
ie. Passerine and non-passerine birds have significantly different levels of correlation between BMS and body size. - this is an example of the comparative approach MASKING patterns
Can control for this for accounting for phylogenetic relatedness in comparative analysis
Give 3 reasons for phylogenetic dependence
- Phylogenetic niche conservatism - the main one, species (such as Darwin’s finches) inheriting niches from their ancestors giving them similar ecologies
- Evolutionary lags such as vesitigal organs and evolutionary inertia. A situation in which a species or population fails to keep pace with changes in its environment through the process of evolution.
- Different adaptive responses: Macaw and Kea having similar beaks but for different purposes
Give 3 consequences of phylogenetic dependence
- Important patterns masked
- Spurious correlations generated
- Statistical tests compromised
Fill in the gaps: To account for phylogenetic dependence we rely on m———— models of trait evolution
This is a m———- or c———- representation of how species traits evolve through time.
This predicts how similar or dissimilar species are as a function of the evolutionary d——- between them
mathematical
mathematical, computational
distance
Give 3 pros of the Brownian model
Makes minimal assumptions
Mathematically easy to work with
Makes predictions which look a lot like real data
Give 4 assumptions of the Brownian model
Traits change all the time
Evolution is neutral - traits can increase or decrease in equal probability
Rate of evolution is constant - the accumulation of differences from ancestors always the same
Species independent of one another
Define the Brownian model
A theoretical framework used in evolutionary biology to describe the random, continuous change of quantitative traits over time. Can allow us to see if there may be a correlation between two traits - correlated evolution.
REVIEW LECTURE 3 when revising. This lecture does not especially lend itself to flashcards
What does this describe?
1. Use data to infer ancestral states of these traits
2. Calculate contrasts between different lineages by subtracting them from one another
3. Perform correlation analysis between the two traits having accounted for part of trait due to phylogenetic dependence
The Felenstein 1985 method for Brownian evolution and calculating phylogenetic contrast. Thus accounts for phylogenetic dependencies when looking for evidence of correlated evolution. Thus allows statistical control for non-independence
Give the four main limitations of the Brownian model
- Variance can increase forever
- No ecological basis
- Species are independent of one another once they have split
- Rate is constant - no punctuate equilibrium and such
Non-brownian evolution adds on some extra t— evolution that is i——– of the phylogeny to try and account for a species specific l—– adaptations
trait, independent, local.
Brownian + Non-Brownian evolution = net amount of evolution
Look at lecture 4 to review actual maths and that
How do you apply non-brownian methods to real-world data?
Try out different values and see which ones best fit patterns seen in real-world data
Uses a statistical model called ‘maximum likelihood’ to do this
You can then see if real data is alining best to Brownian or Non-Brownian patterns
What does lambda do in the Brownian model? What does it mean when lambda is closer to 1?
Tests for phylogenetic dependence
Looks for evidence of niche conservatism
Allows for differing levels of phylogenetic dependence in statistical models.
A lambda closer to 1 indicates higher levels of phylogenetic dependence
What does the adaptive constraints / the Ornstein-Uhlenbeck model do? What letter is used for this constraint?
Addresses the unrealistic Brownian assumption that trait variation can increase indefinitely by showing evolution selecting against traits when they become too big / too small.
Thus you don’t see so much splitting / diverging in traits between two species over time.
Alpha is used in reference to this constraint
We can measure the rates of evolution separately between two or more sets of branches and perform a statistical test -
Tested on pr——- / semipr——– and al——birds
T—- used in reference to this
Finds evidence that indeed p—– birds have higher level of evolution as less c——-
Thomas, Szeckeley & Freckleton 2006
Precocial, semiprococial, altricial - precocial are born quite independent, altricial more hopeless, featherless and blind
Theta
precocial, constrained
To address another Brownian assumption, how do you test for the influence of Geography upon a phylogeny? What letter measures the relative contribution of spatial distance to a phylogeny?
- Species living together may be expected to be more similar in trait values
-*Compute phylogenetic contrasts
*Compare with phylogenetic & spatial distance
*Compute index phi – measures relative contribution of spatial distance
What did Freckleton & Jetz, 2008 test the comparative influence of geography vs phylogeny on?
Applied to bodymass, temperature & range size in mammals
Asks whether these show phylogenetic, spatial signals, or both
Found strong relationship between spatial distance and trait variation NOT phylogeny for temperature
Found neither significantly influenced for range size
Found phylogeny stronger factor for bodymass
Does the comparative approach count as an observational, experimental or theoretical approach? Why?
Is observational as uses real-world data from nature. Fits models to data which are then used to make predictions
Jetz and Freckleton fitted an evolutionary model to body size in mammals and found that…
This could make very accurate predictions (about 65-95% accuracy) about unknown mass. Did this by removing known values and seeing what model predicted, then seeing how close to the true value this was.
phylogeny & s—– data can be used to predict species traits
spatial - based on their geographical closeness to one another
Jetz & Freckleton
Looked at mammals
Looked at range size, body mass, human encroachment, human impacts and interactions between range size and body mass
Used phylogenetic & spatial information to make predictions and fill in gaps such as body mass
What data were they using?
What did they find out?
Jetz & Freckleton, 2012.
Note: Study of the same citation also done on predicting unknown sizes in mammals.
To discern the threat level DD species were probably experiencing
The relationship between observed and predicted values was very good
Data deficient species predicted to have a HIGHER threat probability across many mammalian orders
NON-RANDOM probability of data-deficient species being more threatened - probably as these species have a low population size, live in remote areas
Compare the difference in findings about how widespread phylogenetic dependence was across ecological and behavioural data.
Why may there have been a difference?
So freckelton aimed to see how widespread phylogenetic dependence was across ecological data, and found it to be present in 88% of data sets
Bloomberg did the same thing with behavioural data and found a much weaker signal of dependence, potentially as behavioural adaptations are more labile, and can be taken on relatively quickly
Phylogenetic comparative methods can go beyond just measuring correlations and can make p———- too.
They can link e—– outcomes to e—— models
predictions,
ecological, evolutionary
What does an R-squared analysis tell you about the correlation between 2 (or more!) traits?
When R-Squared is higher, there is a stronger correlation between the 2 traits. Just like any other regression analysis. The slope value in this result in r can also tell you the gradient of this relationship
Define the incubation limitation hypothesis
*In bird species where males incubate but are smaller than females, egg size may be constrained by male body size, and hence ability to incubate the eggs.
What did Buckey & Jetz discover about the correlation between Anolis lizard density on islands and presence of competitors/predators?
Lizard density found to be higher on islands with fewer competitors and predators. This shows density compensation as a result of ecological release.
What would happen to intraspecies interactions as population density increases?
Intraspecies interactions could become more frequent and stronger
Increased intraspecific competition would favour niche expansion, and increased phenotypic diversity of the population
What did Svanback & Bolnick do to those sticklebacks? What did they find out about ecological release?
Svanback & Bolnick, 2007
They manipulated the population density of three-spine sticklebacks.
Used a paired experimental design
Control measures were taken outside of the enclosures
They found that:
Prey density declined at high biomass
Thus increased population density reduced prey availability
Individuals added alternative prey types to their diets
Diet variation among individuals INCREASED relative to low-density control enclosures
Niche breadth increased across the whole population
EVIDENCE FOR ECOLOGICAL RELEASE