1/110
Looks like no tags are added yet.
population dynamics
how populations change through time
basic conservation questions about populations
stable, declining, or increasing?
how many individuals are there?
life cycle and reproduction
sex ratios
the baseline historical information
food resources
habitat fragmentation
endemic or not
prioritization- is there support?
how this population affects others
general threats
Models
exponential, logistic, and matrix models assess threats and find solutions
population viability models predict extinction
this information from models is important to determine recovery plans and ESA doesn’t have this information for 50 percent
exponential growth
continuous
exponential growth equation
dN/dt=rN
dN
change in number of individuals
dt
change in time
N
number of individuals
r
intrinsic rate of increase with birth rates - death rates (individuals/ind*time), always constant regardless of N
combine effects of birth and death rates
why r is different for different species
larger r means faster growth, the shorter the doubling time
doubling time
how long it takes for a population to double
continuous growth (r)
growth happens continuously over time. No time lags behind when individuals enter the population
discrete growth (lambda)
growth happens at distinct time intervals
no differences in predictions, just time difference
discrete growth examples
sea turtles lay eggs once a year
frogs
tulips
assumptions of exponential growth
continuous growth
density-independent growth
deterministic growth
homogenous individuals
no immigration/emigration
no spatial structure
density-independent growth
constant rate of growth (r or lambda don’t change) with a changing population size (N)
deterministic growth
constant values of r or lambda and all parameters influencing growth rates over time
no stochastic or random variation (environmental)
homogenous individuals
all individuals are the same with no genetic, age or size differences
r (b and d) represent averages across all individuals
homogenous example
age structure in humans ignore that not everyone reproduces or are the same age
spatial structure, immigration, emigration can be a problem because
emigration can look like death
immigration can look like birth
we can ignore it if low, but metapopulations need to consider it
how does exponential growth actually occur?
is unlikely because resources are limited in nature and real populations cannot increase without bound
however we know that resources may be temporarily unlimited in nature and all populations have the potential to grow exponentially
examples of populations growing exponentially
invasive species have small numbers with access to resources
when predators die, prey can grow exponentially in a short amount of time
domestic animals
viruses
humans
remove human impacts on animals with conservation concern
examples of populations that grew exponentially after conservation efforts
sea otters
whooping crane when we stopped hunting and helped habitat
sea turtles
Kirtland’s warbler
gray whales
model used for conservation
exponential growth
logistic growth model
continuous (r)
dN/rN= rN(1-N/K)
As N increases, (1-N/k) reduces rate of rN or exponential growth
if N>K then dN/dt>0
if N<K then dN/dt<0
k
carrying capacity which is the limit of how many individuals a habitat can support
1-N/K
unused portion of K
logistic growth assumptions
continuous growth
deterministic
density-dependent growth meaning growth rate changes with N
homogeneous individuals
no immigration/emigration/spatial structure
logistic growth examples
gray wolf was listed in 1974 and exponentially grew that finally reached logistic growth because the populations leveled out
for humans most populations have reached a stable k
instantaneous growth rates
exponential and logistic
are difficult to get data because not all species grow this way and scientists collect data at discrete time intervals
in practice, different models are used for…
discrete growth is used for real data
continuous models are used for theoretical exploration
most conservation modeling uses discrete and matrix models
Discrete growth equation
Nt+1= lambda (Nt)
Nt+1
number of individuals at time t + 1, one year after t
Nt
number of individuals at time t
lambda
finite rate of increase
t
time which is often in years but could be others
discrete growth assumptions
discrete growth
density-independent
deterministic
exponential
homogeneous individuals
no immigration/emigration/spatial structure
general equation to predict N after t time
Nt=lambdat (No)
lambda greater than 1
population grows
lambda equal to 1
population is stable
lambda less than 1
population shrinks
how to convert from lambda to r
er=lambda
r=ln(lambda)
life cycle
individuals vary on survival, growth, reproduction which can depend on age, stage, or size of individuals
life cycle diagram
age-based, stage-based, or size-based
arrows are for order and probabilities of moving to the next thing, staying, or reproducing
age-based
separate in age intervals
stage based
separate in different stages
egg, larvae, pupa, adult
size-based
intervals are in different sizes, a certain size is needed in order to produce seeds
example: trees
life cycle to matrix
make a table with “From” on top and “To” on side of each stage/age and fill in the numbers on arrows
demographic model only looks at
females for the matrix
top row of matrix
average number of female offspring produced by a female with age x
diagonal line starting at 1 to 2
probability that a female of age x will transition to next age
PA,1
probability of transition from age 1 to adult
M
a more detailed version of lambda
Nt+1=MNt
matrix multiplication
always do the 1st column with 1st row then 2nd column with 2nd row
benefits of matrix model
you can calculate lambda by measuring demographic parameters (instead of population sizes through time)
incorporates size, stage, and age structure which allows for non-homogenous individuals
populations w/ constant lambda have stable age or stage distribution
can identify specific actions that will result in a lambda greater than 1
determines parts of life contributing most growth (sensitivity and elasticity)
stable age or stage distribution
lambda is constant when matrix values remain constant
high young and less old (is not stable)
stable is low young and high old
measuring demographic parameters
lambda can be manually calculated by projecting population forward until it stabilizes, lambda=N1/No
lambda is eigenvalue of the matrix
sensitivity
absolute change in lambda per unit change in a particular parameter
example: an increase of a parameter by 0.01 will increase lambda by 0.2
small changes in parameters can have a big impact on lambda
elasticity
proportional change in lambda per unit change in particular parameter
ex: an increase in a parameter by 10% increase in lambda by 25%
Threat analysis and applying matrix models
why did it decline and what’s preventing its recovery?c
chinook salmon life history
fish eggs are laid in inland streams and lakes
adults are in the ocean and migrate upstream to inland lakes to reproduce and then die
chinook salmon abundance
lots of species are in crisis or not keeping pace
main factors affecting salmon
habitat degradation
climate change
fish passage barriers (dams)
predation
harvest
hydropower
hatcheries
method in ideal world
use control vs treatments
replication
performed controlled and manipulative experiments
actual method
observational studies and matrices
how do we find out if the hydroelectric dams are the primary factor affecting the salmon in the Snake River?
tag individuals and track how they survive
did an observational study on salmon before/after dams
salmon observational study
created latters which helps the salmon hop to get over dams
had control, bad dams, and good dams with passageways
before dams salmon
could move more freely
salmon results
building dams reduces upstream spawners
dams reduce downstream migration of recruits
better dams allow salmon to move downstream
salmon conclusions
dams without passageways are hurting recovery of salmon populations
dams with passageways are not limiting the recovery of salmon populations
dams aren’t solely the reason why populations are declining
salmon matrix model
when survival rates both up and downstream were increased to 100%, salmon still declined
other reasons why salmon are declining
habitat quality
stream habitat is not “pristine” wilderness and has 20 fish invasive species including the brook trout that are direct competitors that we raise in hatcheries in order to fish for them
how do brook trout affect salmon?
PIT-tag wild chinook throughout Snake River Basin
detect a boner Granite Dam
calculate survival from proportion and probability of detection (mark recapture)
generally salmon had better survival rates when brook trout were absent
habitat quality for salmon
% no forested
temperatures
water diversions
fraction rangeland
annual rainfall
fraction steep hillslope
no brook trout, good habitat, resulted in best salmon survival
Do hatchery releases of millions of steelhead affect native salmon survival?
more hatchery release = decline in native salmon survival
using a matrix model, these strategies were predicted to increase salmon populations and get lambda above 1
improve stream habitat to reduce fry mortality by 9% (increase p2)
reduce mortality by estuary by 4% (increase p3)
reduce mortality in streams (p2) (3%) and estuaries (p3) (1%)
Loggerhead matrix model
widespread, global range
lives less than 50 years
mature at 22 years
lays eggs on beaches and spend the rest of life at sea
been declining in Atlantic 3-5% for a long time
loggerhead IUCN status
vulnerable, subpopulations are critically endangered to least concern
loggerhead ESA status
recognizes 9 distinct populations with 4 threatened and 5 endangered
primary threats for turtles
overharvesting eggs for food
destruction of nest, beaches (vehicles, lights, humans)
killing of immature and adult turtles for shells
accidental drowning in fishing gear
initial thought for major problems for loggerheads
beaches due to vehicles crushing and tracks
egg predation
shore lights
loggerhead initial response
20-30 years of beach-based protection
beach closures, captive rearing and release
didn’t work
loggerhead change of direction
reducing mortality of juveniles and adults by looking at shrimp trawls where turtles get caught through bycatch
loggerhead solution
Turtle Excluder Devices (TEDs) recommended to potentially decrease mortality by as much as 35%
manipulative experiments to identify most important threats are often not possible because
large scale is hard to replicate like climate change
ethical issues and harming at risk populations
so we build models to do observational studies
Strategies to protect loggerheads
boost reproduction by supplementing eggs and hatchings
boost reproduction by protecting beaches and nests
reduce mortality of juveniles and adults with TEDs
Population viability analysis (PVA)
quantitative assessment of probability of population extinction
PVA components
demographic parameters (birth, death rates, age structure
Environmental factors (habitat quality, climate change)
Genetic considerations (inbreeding, genetic diversity)
PVA begins with a growth model
estimates lambda or M
estimates current N (population size)
projects the current population into the future
when we use PVA
evaluating the risk of extinction for endangered species (% chance of extinction)
assessing the impact of habitat loss or fragmentation
informing management strategies for species
predicting responses to environmental changes (like climate change)
examples of PVA
loggerhead
California condor
California condor
only had 22 individuals in 1982
california condor threats
habitat destruction
lead poisoning from lead ammo (used by hunters)
micro trash ingestion (glass, metal, anti-freeze)
california condor intensive management
captive breeding program produce chicks that are released into the wild
GPS monitors
food provisioning
vet checks, vaccinations against west nile
lead chicks and chelation therapy
removal of trash around nests
california condor results
considered a success but we wondered if it was self-sustaining
argued only recovered and is only surviving on captive breeding and lead is still out there
we need to ban lead ammo
PVA includes stochasticity which means
considers randomness
chance of extinction depends on..
current population size
trend (slope)
variability over time (more variation = higher chance of extinction)
stochastic growth models
show randomness
how large will the population be in 20 years stochastic
uncertain
mean + 95% confidence interval
what is the probability of extinction? stochastic
proportion of iterations that fall below some threshold