Phils stuff Flashcards
Caughly 1994- what did he do
Came up with 2 paradigms because he noticed a lack of integration between practitioners of conservation and scientists
- Small pop paradigm
- Declining pop paradigm
Still major challenge to unite these paradigms
Small population paradigm
Theoretical work dealing with the problems of small populations. Dominated conservation biology up to this point.
Environmental stochasticity- randomness in the environment eg inter annual variation in food available. Cant predict but have big impact on small pop
MVP- minimum viable pop
PVA- Population Viability Analysis. How we might manage a population to get them viable. The conditions under which they are viable.
Metapopulation management- how we manage sub populations. Can we move individuals between them to alleviate interbreeding etc?
Captive breeding- take away from wild threat and put them back when threat removed
Protected area- one big one or multiple small?
Declining pop paradigm
Empirical work identifying and attempting to prevent processes causing the decline of populations. Practitioners use this.
Identify cause of decline and treat it. More practical value than various other approaches- essential to determine causal threats
- correlation isnt causation
- management must be run as an experiment or nothing learned
- consequences of actions must be evaluated according to standard experimentation rules
Issue is often situations are so specific it can be difficult to generalise the results - so not of interest to high profile publications
Example of declining pop paradigm working
Lord Howe Woodhen
- series of ecological disturbances. Pigs introduced, cats, rats. Human settlement in 1834
- Woodhen is a flightless rail endemic to the island that suffered a steady decline.
- By 1853 restricted to mountainous regions,
- by 1920 only on 2 summits of least accessible mountains.
- By 1969 conservation started. Numbers stabilised at 8-10 pairs.
- Intensive field work to assess diet and habitat ruled out food shortages and habitat loss as a possible cause of decline
- Surveyed rat population = rat most abundant where woodhen were= not that
- Mapped pig distribution- fine in 2D but in 3D realised pig couldn’t access the remote mountain peaks and so …
- Pigs eradicated between 1971-1989.
- 1980-83 CB programme
- ‘87 population stable at 180 birds, habitat saturated.
- No inbreeding depression —> shouldn’t just write off populations because they are small
- $200,000
Single threat identified in closed island system
Example of declining pop paradigm failing
Californian condor
- north ameiricas largest bird. Rapid decline
- 60 individuals in 50s, 20 in 83
- 1987 last 7 birds taken into captivity
- 1960s thought food shortages may have caused the decline. So put feeding stations up in 71-73.
- Also learnt about organochlorides causing shell thinning. They were banned in 1972. Circumstantial link to OC accepted as cause of decline.
- Then thought lead was the cause and identified as cause of death in 3/5 birds. Lead likely from deer carcasses because hunts use lead bullets. Condors have very strong digestion so can break down and assimilate lead. Tried to get hunted to use copper bullets but they were reluctant
- 91 reintroductions started
- 06- 44 individuals in the wild >6 years old (maturity age)
$35,000,000!!
Multiple interacting threats and condor are large and slowly multiplying and maturing birds. Mortality from electrical power lines and egg collectors too –> little follow up, nothing rigorously tested
Tiger causes of decline
Poorly studied
Poaching, fragmentation (loss of genetic variability), habitat loss to agriculture, loss of food sources,
Karanth and Smith 99 believed it was because of loss of food sources. This was a new cause at this point, after 20/30 years of trying to stop the tiger decline. Used a simple population model for this
Chapman et al 2008- used data on life history parameters. Tiger have slower life history so vulnerability to poaching i higher than for other solitary big cats. Even good juvenile survival must be accompanied by high adult survival for population stability. So they conclude poaching bigger threat than food shortages.
Other cats mature quicker, have bigger litters, short birth intervals so can withstand larger demographic insult than tigers
Cougar adult survival can be 0.5 and juvenile 0.7 and still be stable
Tigers adult can be 0.75 if all juvenile survive (1)- only need low levels of poaching to cause adult survival to go down resulting in a population decline
Pic on phone
Both studies model based- not that reliable. Need stats
Why is population modelling important?
Data we observe is very noisy with many interacting factors so it can be difficult to find out what is causing a particular response.
Scale of problems we are facing are huge. Make models to replicate environment and study there.
Manipulation of the environment could be too risky. If endangered species often have to make rapid decisions cant wait for generations to see potential impact of what we’ve done
What can population modelling be used for?
Predict how a population can change with interventions, what has the best impact
Can assess risk eg temperature changing, how this will influence the population. Can figure out the scale and urgency of the problem from this without having to collect many years of data
Understand trophic interactions. Manipulation of 1 may influence other. Can see which has the most influence and target.
Understand link between where a species occur and what climate is like there. Can then predict changes. This is species distribution modelling. Can understand how large the threats of CC are.
An expected demographic response can be modelled for each potential cause and decline and compared with independently collected data –> more exploratory than explanatory because doesn’t incorporate explicit info on the cause of decline eg marbled murrelet Peery et al 2014
Write out a simple population model for
- single number
- pop size at time intervals
- continuous time pop growth model
pic on phone
continuous time pop growth model = how much time elapsed up to the point you want to get next estimate
Draw a discrete time model
phone
Draw a continuous time model
phone
How can we model pops with greater realism
Add carrying capacities, density dependence, info about pop structure too eg spatial structure, meta populations, age stages and matrices, individuals
Types of density dependence
Negative (competition)- can be direct or indirect eg scramble competition depletes resources available but not in contact with each other. Contest- competition over indivisible resource- only 1 gets it
Positive (allee effects)
Example of positive density dependence
Wild dogs- need big groups to hunt and care for pups and protect kill from kleptoparasites
Small pops may struggle to find mates. Plus reasonable chance there is a skew in the sex ratio- retard pop growth.
When in company individual more likely to escape predator- dilution effect eg nesting birds
Co-operative vigilance eg meerkats more likely to see predator
Why is it difficult to collect data about allee effects
Populations often decline so rapidly hard to collect data, also the focus should be on helping the pop not modelling their decline
Would also need lots of data
Must distinguish between component and demographic effects
Instead of measuring populations as they declined they looked at eg individual growth rate,
Component- allee effects that act on 1 component of fitness. said the grow more rapidly when other indivs around so must be allee effect. Also did this with clutch size, surviving etc. They measured some component of fitness and showed it was positively related to pop size or density. This isn’t the same as demographic allee effect because there could be some other component you could be measuring?
Eg individuals may survive better when there are other individuals around- eg stop predators- but at the same time they are competing more strongly bc lots of indivs. So their fecundity might be reduced. May have increased survival but decr fecundity. As long as the decr outweighs the incr in survival still see negative density dependence. component
Demographic—overall relationship between pop growth rate and pop size or density. Act on whole population
stochasticity and models
Simple models are deterministic- given conditions always lead to the same outcome and unrealistic eg allow partial growth like 1.3 individuals
Read populations are subject to stochasticity
- demographic- births and deaths happen as integers, sex ratios not always 50:50. So individual input to a population eg you either have 1 or 2 kids (p=0.5). Affects all indivs the same and not limited to small pops
- environmental-impact lots of individuals in the same way eg drought or storm. Conditions change yearly. So in a good year have 2 kids, bad have 1 (p=0.5)
Both examples suggest 1.5 kid a year
demo= all indivs differ
environmental= offspring is a function of the year
Demographic models don’t always escape the issues with small pops but if they reach carrying capacity fluctuations are limited because large numbers of individuals and CLT tells us itll be an average so 1.5 so extinction is less common in this model
Environmental has much more variation and induces catastrophes
Central limit theorem
Demographic stochasticity affects individuals independently - less important as population grows. If you’re looking at random processes the larger your sample size the closer you’ll be to what you predict the mean to be
Spatial modelling
Models can be spatially implicit = space implied but nothing about the model records locations where individuals are - sites are modelled as separate
- rules govern movement
- specific geographic relationship unimportant
Or explicit= models a specific landscape
- usually tied to GIS
- often incorporates much finer scale details
- if specific location then that locations environment and resources specifically
Panmixis
Everyone has the opportunity to mate with everyone. Most don’t exist like this and have sub populations. Not all 1 big population
Spatially implicit
Often grid based, movement rules, edge effects optional.
If interested in edges then can have hard edge so less imigration or emigration opportunities at them
If edges aren’t of interest then taurus shape- wraps around itself
Landscape is infinite
Spatially explicit
Fine scale, physical attributes
Can keep track of multiple subpopulations that may have different dynamics that influence (eg 3 areas closely joined and one further away so can have imi/emigration from ne cell)
Eg P4 have most time empty because limited connectedness so limited prospect of rescue effect. So lowest average pop size
Linked populations
Metapopulation models can model these dynamics of extinction, immi, emigrations
They can be modelled;
explicitly- as sum of dynamics of each patch
implicitly- by assigning p of extinction or recolonisation of each patch
Should patches be connected or no
Yes (IUCN is for)
- permit metapopulation processes
- permit gene flow
- permit large-scale processes eg can track climate
- linked pops can overcome MVP constraints
“Bigger, better, more joined up” John lawton
Can have large scale connections too eg to promote migration
Models can inform us of the importance of connectivity at national and international scales
No
- eggs in one basket- eg disease
- facilitate predation/expansion of invasive
- requirements hard to characterise
- providing something like a bridge or culvert, only some species use and some predators loiter by them and wait for animals so incr bred
