ESPM 114 Lecture Notes - Lecture 12: Metapopulation, Metamodeling
Lecture 12: METAPOPULATIONS
●Missing from single-pop models:
○R of change dep only on b and d
○I/E only sometimes included
○Fluctuations in other pops not included
●Limitations:
1. Single pops can’t recover from extinction ( colonization )
2. Single pops can’t account for spatial heterogeneity (habitat patches, connectivity)’s efect
on pop dynamics
3. Single pops can’t explain unequal b and d rates in a stable pop
a. Sink pop : b<d, lambda<1, i>e
b. Source pop : b>d, lambda>1, i<e, exceeds K if no dispersal
●Solution: simultaneously model multiple pops and their interactions
●Metapopulation : grp of regularly-interacting (thru i/e) local pops of same sp.
○Patch : single pop, discrete
○Unsuitable matrix: surrounding patches
○Connected by limited dispersal
○Dynamics of pops are asynchronous
○Rescue efect : declining pop saved by immigration
●Types of metapopulation models:
1. Patch occupancy models:
a. Occupancy (P) = prop of occupied patches/the prob of any given patch being
occupied (# suitable)
b. Simple but data-intensive
c. Levins Patch Occupancy Model: first meta model, useful for general theory
i. Assumes: infinite patches, all patches equal in area and isolation
ii. Requires c>e for metapop persistence
iii. Var of log model where local K is P
iv. Upside-down U
v. e =prob of extinction, c= prob of colonization
d. Multiple patches “spreads the risk” of extinction
e. Local vs regional extinction
f. Persistence : prob metapop persists over given t
2. Spatially realistic patch occupancy models:
a. Incorporates spatial structure of finite patch network
b. Models prop of occupied patches (P)
c. Moderately data-intensive → GIS
d. Stochastic patch occupancy model (SPOM) simulations
e. Hanski’s Incidence Function Model:
i. Prob of colonization and extinction varies by patch
ii. Afected by patch area and isolation
iii. Area increases, extinction dec, colonization output increases