Hurricane Katrina Aug. 29, 2005

Hurricane KatrinaAug. 29, 2005

Disturbance & Ecological Succession

Image from http://earthobservatory.nasa.gov/Newsroom/NewImages/Images/katrina_goe_2005241_lrg.jpg


Succession – directional change in community composition at a site (as opposed to simple fluctuations), initiated by natural oranthropogenic disturbance, or the creation of a new site

Some biologists restrict the definition to directionalreplacement of species after disturbance

Disturbance – a discrete event that damages or kills residents on a site; either catastrophic or non-catastrophic

(Platt & Connell 2003)

Photo of W. J. Platt at Camp Whispering Pines, LA from K. Harms; photo of J. H. Connell from UCSB


Catastrophic disturbance – a disturbance that kills all residents of all species on a site; i.e., creates a “blank slate” (Platt & Connell 2003)

Photo of Mt. St. Helens from Wikipedia

Mt. St. Helens, Washington, U.S.A.May 18, 1980

Luquillo Experimental Forest, Puerto Ricojust after 1989 Hurricane Hugo


Non-catastrophic disturbance – a disturbance that falls short of wiping out all organisms from a site; i.e., leaves “residual organisms”

or “biological legacies” (Platt & Connell 2003)

Photo of Yellowstone in 1988 from Wikipedia; Photo of Luquillo Forest, Puerto Rico in 1989 from http://pr.water.usgs.gov/public/webb/hurricane_hugo.html

Yellowstone Nat’l. Park, U.S.A.just after 1988 fires


Primary Succession – succession that occurs after the creation of a “blank slate,” either through catastrophic disturbance or

de novo creation of a new site

Photo of Mt. St. Helens in 1980 from Wikipedia; Photo of Anak Krakatau from http://amazingindonesia.net/2008/06/mount-krakatoa-the-wrath-of-earth

Anak Krakatau, Indonesiaappeared above water ~ 1930



Secondary Succession – succession that occurs after non-catastrophic disturbance (including “old fields”)

Photo of Yellowstone in 1988 from Wikipedia; Photo of Luquillo Forest, Puero Rico in 1989 from http://pr.water.usgs.gov/public/webb/hurricane_hugo.html

Luquillo Experimental Forest, Puerto Ricojust after 1989 Hurricane Hugo

Yellowstone Nat’l. Park, U.S.A.just after 1988 fires


Henry David Thoreau (1859) is often credited with coining “succession” as applied to directional changes in plant communities

Photo of Thoreau from Wikipedia

Thoreau made many remarkable observations at a time when many still believed in such phenomena as spontaneous generation

“Though I do not believe that a plant will spring up where no seed has been, I have great faith in a seed.

Convince me that you have a seed there, and I am prepared to expect wonders.”

Examined species composition of Lake Michigan sand dunes & concluded that the dunes were older further inland, i.e., formed a “chronosequence” from which temporal change could be inferred (space-for-time substitution)

Believed that succession tended toward a

stable equilibrium that was never (or at least rarely) reached


Photo of Cowles from http://oz.plymouth.edu/~lts/ecology/ecohistory/cowles.html; photo of Lake Michigan sand dune from http://ebeltz.net/folio/cfol-5.html

A brief history of observations and ideas…

H. Cowles (1899) – stressed the dynamic nature of “plant societies” (“phytosociology”)


H. Gleason (1926, 1939) – “individualistic view of succession” in which “every species is a law unto itself”

Our modern population-biology view derives primarily from Gleason’s conceptual model, even though Clementsian ideas of deterministic progression through seral to climax stages dominated ecological theory well into the 20th century (see Connell & Slatyer 1977)


F. Clements (1916, 1928) – radical, “superorganism” view of communities; species interact to promote a directed pattern of community development through “seral” stages, ending in a “climax” community

Photos from http://oz.plymouth.edu/~lts/ecology/ecohistory/history.html



F. Egler (1954) – made distinctions between primary succession (“relay floristics,” in which initially there is no vegetation) vs. secondary succession (following non-catastrophic disturbance of existing vegetation)

Egler thought secondary successional patterns were driven by propagules present when the disturbance occurs (“initial floristic composition hypothesis”)

In addition, he thought that changes in species abundances reflected differences in longevity of species



Four classic papers demonstrate the maturation of thought concerning the nature of trade-offs & colonization history within Gleason’s “individualistic” framework

Horn & MacArthur (1972) – mathematical models of competition among fugitive species in a harlequin environment

Drury & Nisbet (1973) – verbal models of succession driven by differences in dispersal & competitive ability, growth & survival

Platt (1975) – empirical demonstration of mechanisms of coexistence of fugitive species on badger-mound disturbances

Bormann & Likens (1979) – introduced the “shifting-mosaic steady-state” concept; within-patch non-equilibrial dynamics average to an equilibrium pattern at the scale of many such patches taken together


Three models of succession:

1. Facilitation – Early species enhance the establishment of later species (if it occurs, it is perhaps most likely in primary succession)

2. Tolerance – Early species have no effect on later species

3. Inhibition – Early species actively inhibit later species


Connell & Slatyer (1977) – Reacted against an emphasis on life-history strategies & competition alone; recognized a variety of species interactions that could impact succession


Photo of Glacier Bay National Park, Alaska from Wikipedia

Primary succession along the Glacier Bay chronosequence

One of the world’s most rapid and extensive glacial retreats in modern times (so far); eliminated ~2500 km2 of ice in ~200 yr, exposing large

expanses of nutrient-poor boulder till to biotic colonization



Classical view of Glacier Bay succession based on 50 yr of research, which employed the simple chronosequence assumption:

- Mosses - Mountain Avens (Dryas); shallow-rooted herbs - Willows (Salix); first prostrate, then shrubby species - Alder (Alnus crispus); after 50 yr forms thickets to 10 m - Sitka Spruce (Picea sitchensis); invade alder thickets - Hemlock (Tsuga heterophylla); establish last

Succession is driven by N-fixation (Dryas & Alnus)

Alnus acidifies the soil, allowing Picea invasion

Accumulation of soil carbon through succession improves soil texture and water retention, ultimately allowing invasion by Tsuga



Fastie (1995) – Reconstructed patterns of stand development at several sites within the chronosequence; intensively analyzed tree-rings

Figure from Fastie (1995)



Fastie (1995) – Identified 3 alternative pathways of compositional change (not a single chronosequence of events):

1. Sites deglaciated prior to 1840 were colonized early by Picea & Tsuga

2. Sites deglaciated since 1840 were the only sites colonized early by N-fixing Alnus

3. Sites deglaciated since 1900 were the only sites dominated relatively early by black cottonwood (Populus trichocarpa)



Oldest sites: Dryas Picea & TsugaIntermediate sites: Dryas Alnus PiceaYoungest sites: Dryas Alnus Populus Picea

What accounts for these among-site differences in composition?

Differences are unrelated to soil parent material

Strong effect of seed source: Refugial Picea stands are concentrated at the mouth of the bay; distance from the nearest seed source explains 58% of among-site variance in early Picea recruitment

Younger sites received more of their seed rain from new communities colonizing exposed surfaces than from refugial populations



What about facilitation?

Succession of Alnus to Picea was considered a textbook example of facilitation in the mid- to-late 20th century

The real pattern is more complex!

Alnus was absent on older sites, so Picea does not require it for establishment

Alnus may either inhibit or facilitate seedling establishment of Picea

Chapin et al. (1994) – Found net positive effects of Alnus on Picea on glacial moraines, but net negative effects on

floodplains


Facilitation along cobble beaches of New England

Bruno (2000) – Determined mechanisms by which Spartina alterniflora is a facilitator of relatively large impact on the community (i.e., a “foundation species” - Drayton [1972]; “keystone modifier” - Bond [1993]; “ecosystem or keystone engineer” - Jones et al. [1994])

Observations: Spartina occurs along the shore; cobble-beach plants occur behind Spartina

Cobble-beach community is absent along breaks in the Spartina phalanx

Photo by J. Bruno



Bruno (2000)

Question: At which life stage(s) is colonization cobble-beach plants limited to sites behind Spartina?

Results: Only seedling emergence & establishment were adversely affected by the absence of Spartina

Experiment: Addition experiments to determine limiting life stages (seed supply, seed germination, seedling emergence, seedling establishment & adult survival) for cobble-beach plants



Bruno (2000)

Question: By what mechanism(s) does Spartina facilitate seedling emergence & establishment of cobble-beach plants?

Results: Substrate stability increased seedling emergence & establishment, whereas manipulations of the other factors had limited influence

Experiment: Conducted manipulations of water velocity, substrate stability, herbivory & soil quality in sites lacking Spartina



Bruno (2000)

Conclusions: Spartina alterniflora acts as a foundation species, keystone modifier & ecosystem engineer) by stabilizing the substrate, enabling seedlings of cobble-beach plants to emerge & survive

Photo by J. Bruno

Anak Krakatau


Primary succession on Krakatau & Anak Krakatau

Explosion of Krakatau (1883) The loudest explosion ever heard by humans Created tsunamis that killed 30,000 people on larger islands & mainland

The island was effectively “sterilized”

Anak Krakatau (“Child of Krakatau”) appeared out of the ocean in ~1930 & has been growing ever since

First analyses of colonizing vegetation were by Doctors van Leeuwen (~1930s); more recent expeditions by Robert J. Whittaker

Photo of Anak Krakatau from http://amazingindonesia.net/2008/06/mount-krakatoa-the-wrath-of-earth



Whittaker (1994) – Examined dispersal characteristics of plant arrivals

Nearest mainland site is Sumatra (~ 50 km away); Nearest island is ~ 21 km away

First arrivals (within 4 yr of eruption) were either wind or water dispersed

Early zoochorous plants were dominated by figs; 17 of 24 fig species on the island arrived in the first 30 yr and are now canopy dominants, which suggests that bats have been very important dispersal vectors or mobile links (Old World bats have gut-retention times up to 12 hr)



Whittaker (1994) – There are now 124 zoochorous species on Anak Krakatau

Doves and pigeons (> 4 hr gut retention time) have been important dispersers subsequent to colonization of the island by figs (an indirect mechanism of facilitation by bats operating through figs?)

Many large-seeded species are absent relative to Sumatra & the mainland flora

Anak KrakatauImage taken June 11, 2005 from Ikonos satellite



Image from http://earthobservatory.nasa.gov/Newsroom/NewImages/Images/krakatau.IKO2005_06_11_lrg.jpg

Primary succession on the flanks of Mount St. Helens

May 18, 1980 – the north face of the previously symmetrical mountain collapsed in a rock-debris avalanche that essentially wiped clean 60 km2 of forest



Photo of Mt. St. Helens from Wikipedia

Fagan & Bishop (2000) – Examinedthe influence of herbivores on the rate of spread of lupines (Lupinus lepidus), the site’s main “colonizing” species



Lupines are efficient N-fixers & trap detritus; they are often facilitators in ecological succession

Lupines colonized from remnant populations elsewhere on the volcano to form patches

Spread rapidly initially and then slowed

Figure from Fagan & Bishop (2000)

Why?



Fagan & Bishop (2000) – Ruled out various alternative explanations for slowed population growth rates & focused on the effect of insect herbivores, whose colonization lagged behind the lupines by 10 yr

Experimental test:Established plots at the center of lupine patches (core) and at the edge of expanding patches (edge)

Sprayed half of the plots with pyrethroid insecticide

Much higher incidence of damaging insects at

patch edges

Higher leaf damage at patch edges




Why was there more herbivore activity at the edge?

Densities of predators (e.g., spiders) & parasitoids (e.g., a tachinid fly) were 4x higher at the core vs. edge

Predators may be more abundant in the core where plant density & productivity are higher

EdgeSite

CoreSite




Lower seed productionat patch edges




Fagan and Bishop (2000) – Diffusion model showed how reduced seed production at the edge affects rates of lupine spread (assuming no long-distance, jump-dispersal events)


Modeling secondary succession

Developed simple Markov models of successional replacementof temperate-zone tree species

Forest consists of cells, each occupied by a single tree

– Horn (1975)

Probability of replacing an individual tree with a new individual of a given species is calculated from a transition matrix

Example of transition matrix for four species (GB=grey birch; BG=black gum; RM=red maple; BE=beech)

GB BG RM BEGB 0.05 0.36 0.50 0.09BG 0.01 0.57 0.25 0.17RM 0 0.14 0.55 0.31

BE 0 0.01 0.03 0.96

Initial composition vector: (100, 0, 0, 0)

After 1 time step: (5, 36, 50, 9)

Iterate this process & plot the changes in relative abundance…

BE

RMBG

GB


Figure from Horn (1975)

Modeling secondary succession – Horn (1975)

One approach for estimating transition probabilities: proportional to the fraction of each species as saplings beneath adults, e.g., if 5% of

saplings beneath GB are GB, then P(GB|GB)=0.05


If the same transition matrix is used throughout, then a stable composition (the dominant Eigenvector) will result (here dominated by BE)

However, the Markov approach is phenomenological, so…Why do recruitment probabilities vary, i.e., what biological traits lead to

different colonization rates & relative abundances?

Modeling secondary succession – Horn (1975)


Modeling secondary succession – Pacala et al. (1996; SORTIE)

The most recent generation of forest simulation models; precursors include FORET (Shugart & West 1977)

Spatially explicit, mechanistic simulation model developed to predict dynamics of succession for 9 species of northeastern U.S.A. hardwoods

Early occupation by Red Oak (Quercus rubra) & Black Cherry (Prunus serotina) followed by late dominance by Beech (Fagus grandifolia) & Hemlock (Tsuga canadensis), with Yellow Birch (Betula alleghaniensis) present in gaps



Basics of SORTIE:

Spatially explicit model predicting the fate of every individual tree throughout its life

Individual performance is affected by resource availability at each tree’s location (original SORTIE only included competition for light)

Species-specific functions predict each individual’s growth, mortality, fecundity & dispersal; estimated from data collected in the field

Four sub-models determine the fate of each individual throughout its life



(1) Resource (light) submodel: Calculates light available to an individual based on its neighborhood; the process is analogous to taking a fisheye photo above each plant

Calculates a projected cylindrical crown for each individual based on data relating crown diameter & depth to stem diameter

Computes whole-season photosynthetically active radiation (PAR) for each plant based on the location & identity of neighbors

Figure from Pacala et al. 1996



(2) Growth sub-model: Species-specific equations predict radial growth from diameter & light availability





(3) Mortality sub-model: Species-specific equations predict probability of death from an individual’s growth rate over the past 5 yr



(4) Recruitment sub-model: Species-specific equations predict the number & spatial locations of seedlings based on the sizes of adult trees




Community-level output: From randomly seeded initial composition Hemlock & Beech clearly dominated after 500 yr



The mechanistic approach taken in this model allows one to ask:


Which key traits define species performance?

How sensitive are model predictions to parameter values (and therefore sampling errors in parameter estimation)?

How would hypothetical species with different parameter values perform in this community? What would constitute a “superspecies” (i.e., one of J. Silvertown’s ecological / evolutionary “demons”)?

How many species could potentially coexist, e.g., > 50 spp. for > 10,000 yr?

How would changing abiotic / biotic conditions affect forest trajectories?


See Doug Deutschman’s on-line visualization of SORTIE!Link to SORTIE visualization


Baseline without disturbance

Heavy disturbance

Large, circularclear-cut

Figures from Deutschman et al. 1997

http://www.sciencemag.org/feature/data/deutschman/index.htm


Succession may involve changes beyond species composition…

Community and Ecosystem Properties:

Diversity – often increases throughout succession

Standing-crop biomass – often increases throughout succession

Elemental cycling & other biogeochemical processes – e.g., the Hubbard Brook experiments in New Hampshire, and Peter Vitousek’s work in Hawaii

Susceptibility to disturbance – may be a function of time since last disturbance, e.g., fire and the accumulation of fuel loads

Anthropogenic Disturbance & Ecological Succession

If “all species have evolved in the presence of disturbance, and thus are in a sense matched to the recurrence pattern of the perturbation”, why

are anthropogenic disturbances often so damaging? (Paine et al. 1998)

Anthropogenic disturbances often differ from the naturaldisturbance regime in timing, frequency, or intensity

Paine et al. (1998) also argued that: “more serious ecological consequences result from compounded perturbations within the

normative recovery time of the community in question”


A marine example: Corals in the Caribbean

Hughes (1994, Science)

One-two punch of overfishing (“selective disturbance”) & “natural” mass mortality of dominant urchins (Diadema) has created a “phase-shift” from coral-dominated to macroalgae-dominated reefs

Caribbean coral reefs may never recover!

Photo of macroalgae-dominated reef from http://news.mongabay.com/2008/0108-hance_coral.html


A terrestrial example: Dipterocarps in southeast Asia

Curran et al. (1999, Science)

Photo of dipterocarp forest from http://biology.ucsd.edu/news/article_012706.html

One-two punch of logging & increased frequency of El Niño events (due to anthropogenically induced climate change?) resulted in elimination of recruitment by dipterocarps in forests of Borneo!

May result in a large-scale “phase-shift” away from dipterocarp domination of the forests [dipterocarps are the principal food of giant squirrels, bearded pigs, several species of parakeet & myriad specialist insects, etc.]

Hurricane Katrina Aug. 29, 2005

Documents

harms photo of

coining succession

blank slate platt connell

puerto ricojust

catastrophicplatt connell

hurricane hugoyellowstone

novo creation

hurricane katrinaaug