Transcript 1091-Lec15(PVA2P)

Population Viability Analysis Lecture15
Puerto Rican parrot
Black footed ferret
What sort of PVA model is appropriate?
DISCUSS
What is gained from using PVAs?
USES OF PVA
1. Assessing the extinction risk of population
Questions
What type of model to use?
Simple - 1 population, all individuals equal
- 1 population, with st(age) structure
- multisite- patch occupancy models
Complex - multisite, structure+ spatially
realistic
Issues
Data quality
Parameter values may
MORE USES OF PVA
1. Assessing the extinction risk of
population
2. Comparing the relative risk in 2+
populations
3. Identify key life history stages to
protect
4. Determining minimum reserve size
5. Determining numbers to release
6. Setting harvest guidelines
2. Compare relative risks of populations
Pacific salmon populations Allendorf et al. 1997
Aim: Assess threat, rank priority of acting
2. Compare relative risks of populations
Allendorf et al. 1997
Approach: use PVA to assess threat
Used Stage-structured models
Designated three levels of risk
Very high: 50% within 5 yrs
High:
20% within 20 yrs
Moderate:
5% within 100 yrs
Q. What additional factors may impinge
on ranking management priorities?
2. Compare relative risks of populations
Allendorf et al. 1997
Combined estimated extinction risk with
1
2
3
To get overall priority for management
3. Identifying key stages for management
E.G. RED COCKADED WOODPECKER
Endemic to southeastern US
Nest in cavities in old pines
Endangered
Live in groups
Males - breeder, helper, floater
Females - breeder, floater
STRUCTURED MODEL
STRUCTURED MODEL
Modelling approach
Identify stages (age/types)
Data needed - transition probabilities
- stage specific survival/fecundity
Track numbers of each stage
ASK - What parameter influences viability
Sensitivity analyses
Sensitivity
change in  due to small absolute change
in stage specific vital rates
Elasticity
change in  due to small % change
in stage specific vital rates
Parameter (M=male, F=female, Fldg =Fledgling)
Q What factors had the biggest impact on lambda
STRUCTURED MODEL
RESULTS
Model most sensitive to
1. Changes in female mortality
2. Mortality of female dispersers
3. Fledglings/brood
RECOVERY PLAN
1. Enhance dispersal to new sites by
making nest cavities
2. Remove competing flying squirrels
4. Estimating the reserve size necessary to protect
grizzlies in BC
Aim: establish 1 benchmark population/biogeoclimatic zone
benchmark = unhunted, naturally regulated population
How many bears and how big an area?
Approach- use PVA to get N with low risk of quasi-extinction
(ie risk of falling below 100) in 20 years
Data from 6 different grizzly bear popn’s
----> mean±SD
Age-specific survival
Age at first birth
Fecundity (female cubs/adult female/year)
Maximum age
cubs
yearlings
Sub-adult F
Stochasticity - environmental +demographic
- no catastrophes
Density dependence - none below K
Adult F
RESULTS
Initial N
100
150
200
250
Pextinction
1
0.7
0.05
<0.01
Flathead Valley, seBC
South central BC
Textinction (years)
1
2
19
>20
Density
6.4/100km2
2.3/100km2
Area need
3900 km2
10,700 km2
But bears can/do move outside any fixed area
--->
Recommendation??
Wielgus 2002 Biol Conservation
Final recommendation for 6 reserves
MVP = 250 animals
= 11-15% BC pop’n
Total area
51,000 km2
5% of land area of BC
Uses of PVA
5. How many individuals can be released
Bearded vultures
Disappeared from Alps in 19th century
Restricted to Pyrenees, southern Balkans, Corsica and Crete
Uses of PVA
5. How many individuals can be released
1970’s reintroduction attempt - fails
---> captive breeding program
1986
2nd reintroduction program
1993
100 captive, 53 released into Alps
Demand for captive birds was increasing
How many birds can be released per year?
When is the wild population viable?
Bustamente 1996 Cons Biol 10 822-831
Approach
Do PVA on captive population with varying numbers of
fledglings used for reintroductions
PLUG DATA INTO VORTEX
Age specific mortality
Fecundity - breed 7-31yrs; not density dependent
No Inbreeding depression
Catastrophe 1% chance
Initial pop’n 50m 39f > 1 yr
K = 200 set by $$$ cost
Kids released
3m3f
4m4f
5m5f
6m6f
7m7f
extinction rate
(% in 200 yrs)
1.6
88
8.6
70
22.8
71
52.6
64
74.4
62
time
actual
Q. How would you inform captive breeders what
would best increase numbers to be released?
How viable is the reintroduced population?
Approach
PLUG MORE DATA INTO VORTEX
Data available
Fecundity - Pyreneean population
Mortality - from released birds
Inbreeding - likely
Variance ???
OUTPUT - IF mortality as low as minimum
estimates current population < 5% chance
extinction in 200 yrs with no additional
releases
Uses of PVA
6. Setting harvest limits for American ginseng
Panax quinquefolium
(COSEWIC)
1988 “threatened”
1989 export discontinued
Harvest continues - but what is sustainable?
Approach used Nantel et al. 1996 Cons Biol 10
Stages in the life of a long lived perennial plant
SeedSeedling1-leaf2-leaf3-leaf4-leaf
Assess mortality transitions and seed production
4 populations southern Ontario /Quebec
Incorporate variance
use between yr/between population variation
Death
Seeds
Seedling
1 leaf
2 leaf
3 leaf
4+ leaves
Produce seeds
Seedling
1 leaf
2 leaf
3 leaf
4+ leaves
4+ leaves
Harvest
Estimate growth rates of 4 populations
under different harvest regimes
Estimate minimum viable population
Ie lowest N0 where extinction < 0.05 over 100 yrs
MVP with no harvest estimated as 170 plants
Quebec 3 popn’s >500 most <50
Ontario populations average 10-20 plants
Canada
12 known popns > 170
Implications
Most pop’ns too small to harvest
Annual harvest >10% will lead to extinction
Rotation 5+ yrs would be prudent
The Major Challenges for PVA models
Data requirements
Developing robust variance estimators
Incorporating change to the environment
What makes a good PVA?
Understanding of
the species ecology
environmental disturbances/threats
Species response to threats
Assessment of
Current state of population
Future risks
Evaluation of habitat as well as population
Relevance to managers
But NOT necessarily amount of data
Why?
Burgmann and Possingham 2000
PVAs - the bottom line
Stochastic population models
Tool for evaluating risk to populations
Numerous uses
-Accuracy - debated - but allow assessment of the relative value
of management options
Major value may be that they
organise information, engage stakeholders