State-and-transition Models:Current Status and Future DirectionCurrent Status and Future Direction

David D. BriskeDavid D. Briske Ecosystem Science & Management Texas A&M UniversityTexas A&M University

Professional Reinvention

State-and-transition Model Framework

State 2State 1

Threshold

Community pathway

Threshold

Threshold

Community phases within stable states

State 3

Stringham et al. 2003

Presentation Objectivesj

E l li k b t STM d iliExplore linkage between STMs and resilienceAssess the current effectiveness of STMsInvestigate role of empirical data in STMsComment on the future direction of STMs

Thresholds vs ResilienceThresholds vs Resilience

Thresholds Resilience

Resilience – degree of modification that an ecosystem can absorb prior to transform to an alternative stateabsorb prior to transform to an alternative state. Threshold – resilience limit of an ecosystem.

Resilience based ManagementResilience-based Management

Ecological Resilience Ecological Resilience

State I State II

Feedback switch

ThresholdTrigger

Community Phase

Community

Negative Feedbacks Negative Feedbacks

PhaseAt-risk

Community Restoration pathway

Positive Feedbacks Positive Feedbacks

Negative Feedbacks Negative Feedbacks

Modified from Briske et al. 2008

Positive and Negative Feedbacks

Grassland Woodland Threshold

State StateProgression

Positive Feedbacks• woody plant cover• coarse fuel loadsw

itch

• coarse fuel loads• propagule limitations

ack

Sw

Negative Feedbacks• grassland productivity• fine, continuous fuel loadsFe

edb

fine, continuous fuel loads• propogule limitations

Mountain Clay Ecological Site, OregonReference StateIndicators: High perennial grass cover, dispersed sagebrush cover, minimal juniper and bare soil. Feedbacks: Herbaceous cover retains water on

Idaho FescueBluebunchwheatgrass Idaho Fescue

Bluebunchheatgrass

At-risk Community Phase: Herbaceous cover reduced, sagebrush decadence, juniper visible and bare

Feedbacks: Herbaceous cover retains water on site and provides fuel to support a fire return interval of less than 50 years.

wheatgrassMountain big sagebrushJuniper

Idaho FescueBluebunchwheatgrassMountain big

b h soil patches increasing, potential fire frequency reduced.

Trigger: Drought and intensive grazing promote juniper establishment through reduced fire frequency.

sagebrush

JuniperSagebrush

Threshold: Juniper attains a height and density that reduces fine fuel load and fire-induced tree mortality. Large, inter-connected bare soil patches occur with redistribution of nutrients/soil beneath juniper canopies.Restoration Pathway: Bunchgrass

(BG) density > 1 m2 requires mechanical juniper removal only;

SagebrushIdaho FescueBluebunchwheatgrass

mechanical juniper removal only;BG density < 1 m2 requires juniper removaland grass reseeding, if soil is intact.

Alternative State

JuniperIdaho FescueSandberg bluegrass

Alternative State Indicators: Mature juniper dominant, Idaho fescue only beneath juniper canopies, large interconnected bare soil patches, sagebrush decadence.. Feedbacks: Juniper dominates resource use, water and wind redistribute soil and nutrients beneath juniper, minimal grass and sagebrush establishment .

Effectiveness of STMs

Say, aren’t here supposed to be holes in this mask!

How Effective are STMs?How Effective are STMs?

Survey 47 rangeland professionalsSurvey 47 rangeland professionals26 Agency Managers21 Research Scientists21 Research Scientists

Purposes of STMsM d l St thModel Strengths Model Weaknesses Construction and Review

STM Purposes S u poses

Guide management (87%)Guide management (87%)Managers 92%; Researchers 81%

D ib l i l d i (70%)Describe ecological dynamics (70%)Managers 65%; Researchers 76%

Id if bl h h (40%)Identify testable hypotheses (40%)Managers 12%; Researcher 76%

Communications tool (38%)Managers 35%; Researchers 43%

STM StrengthsSTM StrengthsImprove decision making (87%)Improve decision making (87%)

Managers 92%; Researchers 81%D ib t d i (70%)Describe system dynamics (70%)

Managers 65%; Researchers 76%I i i (38%)Improve communication (38%)

Managers 35%; Researchers 43%Identify relevant questions (34%)

Managers 19%; Researchers 52%

STM WeaknessesSTM WeaknessesInsufficient information (43%)Insufficient information (43%)

Managers 30%; Researchers 57%M d l l l (26%)Models overly complex (26%)

Managers 38%; Researchers 10%L k f i d (21%)Lack of time and resources (21%)

Managers 27%; Researchers 14%Potential misrepresentation (17%)

Managers 8%; Researchers 29%

Construction & ReviewConstruction & ReviewExpert knowledge critical (43%)Expert knowledge critical (43%)

Managers 47%; Researchers 37%Mi i l i i l k l d (43%)Minimal empirical knowledge (43%)

Managers 34%; Researchers 61%M d l i i (26%)Model inconsistency (26%)

Managers 34%; Researchers 13%Mechanisms for validation (87%)

Managers 87%; Researchers 88%

Areas of STM RefinementAreas of STM RefinementManagement vs ecological driversg gRole of expert vs empirical knowledgeCriteria to define thresholdsCriteria to define thresholdsAppropriate model complexityModel review and revisionModel review and revision

Value of Empirical Data

“Then a miracle occurs” !!occurs !!

Attributes Idaho Data SetsAttributes Idaho Data Sets

Idaho National Lab US Sheep StationIdaho National LabPlant density

2

US Sheep StationPlant density

234 m2 plotsSampled 10 times

15-26 m2 plotsSampled 23 times

1950 – 2006N = 340 samples

1930 – 1957N = 545 samples

Species = 55 MAP = 220 mm

Species = 54MAP = 300 mm

Idaho Falls ID Dubois ID

Data AnalysisData AnalysisDr Sumanta BagchiDr Sumanta Bagchi

Identify communities with cluster analysisVerify community membership against species dissimilarity

BIC-parsimony, ANOSIM, SIMPERRecord community transitions in timey

Categorize transition frequency and attributes

Species Composition Species Composition

80

100

e (%

)

othersSa.kaPa.smAl.de

b) INEEL

20

40

60

age

abun

danc

e

Me.alBr.teLa.ocOp.poHe co

0

20

1950

1957

1965

1975

1978

1985

1990

1995

2001

2006

Aver

a He.coEl.elEl.laCh.viAr.trd

80

100

e (%

)

othersEl.alKo.maPo.se

a) USSES

20

40

60

age

abun

danc

e

Br.tePs.spTe.caPh.loPh ho

0

20

1930

1932

1934

1936

1938

1940

1942

1946

1949

1951

1954

1956

Aver

a Ph.hoCr.acCh.viBa.saAr.trp

Transition DissimilarityTransition DissimilarityTransition DissimilarityTransition Dissimilarity

Idaho National LabUS Sheep Station

50

67

Idaho National LabUS Sheep Station

ency 30

40

quen

cy 45

Freq

ue

1020

Freq

12

3

0.0 0.2 0.4 0.6 0.8 1.0

0

S i di i il it

0.0 0.2 0.4 0.6 0.8 1.0

0

Species dissimilaritySpecies dissimilarity

Empirical STMsEmpirical STMsEmpirical STMsEmpirical STMs

) USSES) USSES

A5

7

a) USSES

A5

7

a) USSES

D E7 12

7

b) INEEL

D E7 12

7

b) INEEL

D B11

76

13D B

117

6

13

46

4 73

318

46

4 73

318

E1

3

22

E1

3

22 AC 18

36 9 5

AC 18

36 9 5

C EC EF B

9 4

611F B

9 4

611

66

Community TransitionsCommunity Transitionsyy

40

50

m) 25

30

ive

)

Precipitation (cm) Cheatgrassa) USSES

10

20

30

40P

reci

pita

tion

(cm

5

10

15

20

25

heat

gras

s re

lat

abun

danc

e (%

40 70Precipitation (cm) Cheatgrassb) INEEL

0

10

1920 1930 1940 1950 1960 1970

P

0

5 Ch

152025303540

pita

tion

(cm

)

3040506070

gras

s re

lativ

end

ance

(%)

05

1015

1940 1960 1980 2000 2020

Pre

cip

01020

Che

atg

abun

Temporal Dynamics Temporal Dynamics p yp y

1.0

a) Abrupt change 1.0

c) Gradual change5

0.0

0.5

DC

A ax

is 2

50.

00.

5

-1.0 -0.5 0.0 0.5 1.0

-1.0

-0.5D

-1.0 -0.5 0.0 0.5 1.0

-1.0

-0.5

0.5

1.0

is 2

b) Frequent reversal

0.5

1.0

d) Relative stability

0-0

.50.

0

DC

A ax

0-0

.50.

0

-1.0 -0.5 0.0 0.5 1.0

-1.0

DCA axis 1-1.0 -0.5 0.0 0.5 1.0

-1.0

DCA axis 1

Summary Idaho Data Setsy

Transitions occurred in a 10 yr windowyAssociated with increasing cheatgrass density

Transitions decreased at maximum densityyAlternative stable state formed

Cheatgrass is a ‘biotic trigger’Cheatgrass is a biotic trigger Interaction with precipitation patterns

Feedbacks rapid and unrelated to fireFeedbacks rapid and unrelated to fire Likely induced by plant-soil processes

Similar patterns occurred at both sitesSimilar patterns occurred at both sites

Value of Empirical Data?Value of Empirical Data?Empirical data can support STMs:Empirical data can support STMs:

Describe community transitionsIdentify temporal scalesIdentify temporal scalesAssess feedback mechanismsRefine resilience hypothesesRefine resilience hypotheses

Vegetation records insufficient: Adaptive management best approachAdaptive management best approachMonitor management outcomesConsider autogenic & climatic processesConsider autogenic & climatic processes

F t f STMFuture of STMs

Strong, consistent support among stakeholders

Continue resilience-based foundation

Adaptive management supported w/ monitoring

Science-management partnerships