Island biogeography: lecture topics

1
Island biogeography lecture topics

• Introduction to island communities Empirical
  patterns of species diversity on islands
• MacArthur Wilson Dynamic Theory of Island
  Biogeography
• Theory
• Tests
• Criticisms/problems
• Environmental application of island biographical
  theory Conservation Biology

2
community patterns of isolated oceanic islands
like Cocos

• Depauperate fauna, flora
• Disharmonious fauna, flora
• Defined as non-representative types of organisms,
  including absence of typical mainland types
• Cocos Island, for example, has no native land
  mammals, no amphibians, no bees, only one
  butterfly species, two lizards, four resident
  landbirds and a disproportionate abundance of
  melastome plants (gt12 species)
• Proportionately many endemic species
• Closest relatives found nearby--Galapagos,
  mainland Central South America

3
Cocos Island, Costa Rica Remote (isolated)
island in E. Pacific Ocean (6.25º N. Latitude)
4
Aerial view of Cocos Island (rainforest covered)
(Photo T.K. Sherry)
5
Only four Resident landbird species, of which
Cocos Finch shown here (used as example of
ecological release niche expansion in prior
lecture) (Photos T.W. Sherry T.K. Werner)
6
Dispersal is important aspect of which species
(and how many) occupy oceanic islands

• Colonization process, movement to newly created
  (or scoured) volcanic islands well documented
• Krakatoa in Java Straight
• Surtsey near Iceland
• Mt. Saint Helens colonization well documented
• Waif dispersal arrival by chance, based on
  wind, ocean currents, flying off course, floating
  on debris, etc.
• Thus, the kinds of species on islands are
  non-random subsets of mainland pool of potential
  dispersers (waif dispersers are those species
  best adapted to disperse, survive during
  transport--e.g., r-selected!)

7
Wind, sea-dispersed organisms disproportionately
abundant initially on Krakatau Island all
represent waif dispersal
8
Next pattern Species-area curve illustrates
dependence of species richness on area. Example
shown here for herpetofauna (reptiles
amphibians) of Greater Antilles and nearby
smaller Caribbean islands
9
Mathematics of species-area curve

• Number of species on islands increases with area
  of island (same relationship holds for areas of
  mainlands!)
• Most prevalent mathematical form of relationship
  S CAZ
• S number of species on island, A island area,
  C, z are fitted constants
• I.e., log(S) log(C) Zlog(A)
• This is equation for a line on logarithmic axes
• log(S) y axis
• log(A) x-axis,
• log(C) y-intercept constant
• Z-values slope, determined empirically
  average 0.3 range 0.15-0.35

10
Generalizations about z-vales (slope) of
species-area curves

• Mainland (habitat island) z-values (0.15-0.25)
  tend to be lower than real (oceanic)
  islands--why? (mainland areas at all scales tend
  to have transient individuals, because
  dispersal barriers reduced on mainland islands)
• Spp. that disperse well tend to have lower
  z-values
• E.g., on Nevada mountain tops, birds z 0.165,
  mammals z 0.326
• Deserts below not such a barrier for birds, which
  disperse more easily, show less effects of
  isolation (i.e., lower z next slide)

11

Z-values (slope in species-area curve) tend to be
lower in groups that disperse well (such as
birds, graph b) than in groups that show more
effect of isolation (such as mammals, graph c)
12
Whats biological basis for species-area curve?

• Smaller areas tend to have less habitat (e.g.,
  Lesser Antilles islands tend to be homogeneous
  dry forest, whereas greater antilles have desert,
  rainforest, high mts.)
• Area effect per se? Some studies have shown
  statistically that area has an effect
• I.e., controlling for effect of habitat
  heterogeneity by considering at any one time only
  islands with similar heterogeneity
• E.g., Barbuda (160 km2) has 20 resident landbird
  species vs. Anguilla (90 km2) has 11 spp. (both
  are 300 m high islands, with similar habitats)
• Smaller populations on smaller area tend to go
  extinct? (Very recent empirical support, but
  still weakly supported)
• Smaller islands less of a target for colonizing
  species?

13
Species-area relationship extends to herbivores
feeding on host tree species as islands
14
Final pattern Isolation effect, shown by fewer
species on isolated islands, in species-area
curve for birds of warm ocean regions red
triangles represent isolated islands (gt300 km
from next largest land mass) (from Paul Slud)
Cocos Island, Costa Rica
15
MacArthur Wilson (1967) Dynamic theory of
island biogeography

• Attempt to synthesize into a model empirical
  knowledge from foregoing discussion
• Inductive theory that integrates immigration and
  extinction processes
• Its an equilibrium theory that predicts number
  of species at point that immigration (gain) rate
  equals extinction (loss) rate of species on
  islands
• Its one of most successful paradigms/models in
  ecology
• Graphical model plot of I immigration rate, E
  extinction rate, both as a function of number
  of species on island

16
Graphical model of one variation of
MacArthur-Wilson Dynamic Theory of Island
Biogeography
17
(No Transcript)
18
Things to note about model?

• First, remembering axes is key to understanding
  model
• s-hat is equilibrium number of species...is this
  a dynamic equilibrium?
• What is maximum I value, and why? Why does I
  curve decrease?
• Why does E curve increase?
• Why are curves probably curvilinear, and not
  linear?
• What effect in the model of island size,
  immigration constant?
• What effect in the model of isolation (distance
  effect), island size controlled?

19
Effect of island area, distance held constant
20
Effect of island distance (isolation), area held
constant
21
Both isolation (distance) effects on immigration,
and island size (area) effects on extinction,
combined into one model--showing different
predicted equilibrium species richness values
22
Dynamic theory of island biogeography makes
variety of deductive predictions (some new to
this theory)

• If decrease area, should decrease number of
  species
• If increase distance, should reduce number of
  spp.
• Constant no. of species, if island is in fact at
  equilibrium
• Existence of species turnover (new ones arrive,
  others go extinct)
• Turnover rate should be greater in near, small
  islands than far, large ones

I
Rate (immigration, extinction, turnover T)
E
T
No. Species on Island
23
Dan Simberloff tested MacArthur-Wilson
predictions (turnover, area effect, equilibrium)
with defaunated mangrove islets, Bay of Florida
24
Simberloffs results, from defaunated mangrove
islets
25
Simberloffs results...

• There exists an equilibrium species diversity on
  the islands (return, post-defaunation, to
  pre-treatment species diversity, based on island
  area)
• Experimental reduction of mangrove areas caused
  decrease in species (OVERHEAD Fig. 19.12,
  Stiling text)
• Turnover of species composition, due to continual
  immigration and extinction on each island (data
  not shown)

26
Experimentally reduced areas for rainforest
birds, Manaus, Brazil (Tom Lovejoys Minimum
Critical Size of Ecosystems study)
Result? Many animal species disappeared from
forest fragments, especially rainforest interior
species big edge effects on trees other species
27
Criticisms of MacArthur-Wilson theory

• Few tests conducted to date...and they dont
  usually identify underlying processes
  (immigration, extinction)
• Underemphasizes evolution on islands (Mark
  Williamson continental ids. dominated by
  ecological processes, oceanic ids. by evolution)
• E.g., Cocos Id., Costa Rica, is distant (gt300 mi.
  from mainland in Pacific Ocean), dominated by
  endemism adaptation---gtlow turnover rates of
  landbirds
• Speciation important in archipelagos such as
  Gapalagos, Caribbean Ids.--e.g., birds, reptiles,
  amphibians, plants
• Ecological barriers to colonization (David Lacks
  idea)
• Disharmoniousness makes environment hostile to
  colonization by some organisms
• Thus colonization is more than just dispersal!

28
Cocos Island Vagrant, starving great blue heron
(due to disharmonious-ness of fauna, harsh
climate) would be unlikely to colonize
successfully. Thus successful dispersal does not
guarantee successful colonization.(Photo T.W.
Sherry T.K. Werner)
29
Additional criticisms/additions to
MacArthur-Wilson theory

• Rescue effect (Brown Kodric-Brown) distance
  affects not just immigration, but also extinction
  rates...near islands have immigration that
  reduces chance of extinction
• Target effect (Whitehead and Jones) area affects
  not just extinction, but also colonization...large
  r islands provide larger targets to colonists
• Importance of history
• E.g., landbridge islands such as Trinidad have
  not yet relaxed (extinctions still ongoing) to
  expected lower equilibrium species richness of
  island its size
• Rickleffs work on molecular phylogenies History
  more impact on Caribbean bird distributions than
  equilibrium

30
Relaxation effect Number of species on
continental islands decreases back towards
MacArthur-Wilson equilibrium over time, from
super-saturated condition when island formed by
post-Pleistocene sea-level rise
31
Applications of island biogeography theory?

• Design of parks natural preserves so as to
  maintain native diversity
• Parks islands in sea of inhospitable (e.g.,
  agricultural) habitats
• Best designs to maximize area and minimize
  dispersal distances between units of park Next
  slide
• Related design feature Parks with top predators
  better than those without (bcause top predator
  helps maintain diversity of organisms at lower
  trophic levels)
• Important implications for field of Conservation
  Biology!

32
Application of island biogeography ideas for
design of nature preserves
Better design
Worse design
a.
Area
Less subdivision
b.
c.
Proximity
d.
Proximity
e.
Corridor
Edge effect (perimeter)
f.
33
Conclusions

• Many patterns of species richness on islands well
  known for a long time--e.g., spp-area
  relationship
• MacArthur Wilson synthesized these ideas into a
  predictive, mechanistic model, dynamic theory of
  island biogeography
• Many deductive predictions made from this model,
  and many of these are supported by both
  observational and experimental tests
• Some problems with model are well documented, and
  these add to our understanding of when model is
  applicable, rather than invalidating model
• This is well supported equilibrium model,
  attesting to balance of extinction with
  origination processes influencing community
  structure

34
Acknowledgements Some illustrations for this
lecture from R.E. Ricklefs. 2001. The Economy
of Nature, 5th Edition. W.H. Freeman and
Company, New York.