risk from release of biotech trees James S. Clark Nicholas School - - PowerPoint PPT Presentation

Our charge:

Forest health measures for evaluating risk from release of biotech trees

James S. Clark Nicholas School of the Environment Department of Statistical Science Duke University

My charge: Climate-change emphasis

• Biotechnology risks to forest-health
• How they differ from non-biotech efforts
• Modelling to assess risks

Ele lements of f forest health–why size-species str tructure?

• ‘Human health’: concern for

the individual

• Mammals: ‘body condition’—

we mostly don’t care about individuals

• Demography: individual scale

performance impacts populations

• Forests: ’dynamic size-species

structure (SSS)’

• The basis for conservation and

management: thinning practice

• Species diversity: is it declining?

Individual species at risk

• Example services:
• Masting system—base of the food chain
• Carbon storage
• Wood/fiber
• Recreation/spiritual renewal
• Sustainable: change in size-species

structure (SSS)

• Monitored throughout N America

and most European countries

• Combined effects of demography
• Concentrations in large size

classes: recruitment failure, succession

• Concentration in small size

classes: recruitment bottleneck in crowded stands, invasion

Siz ize-species structure (S (SSS): : climate, management, disturbance

Clark et al, National Climate Assessment (2016)

Main points

• Risks
• Escape: pollination, seed

dispersal

• Super competitors that beat

the pervasive limitations

• Pathogens/insects released

from natural enemies

• Exotic invaders suggest

the possible

• The role of monitoring
• How much is enough?
• Models could do better
• but not much

The pervasive limitations

• Escape
• Pollen expensive, dispersal

inefficient, short-term viability, hybridization

• Seed expensive, dispersal

inefficient

• Ex: individual produces 106 seeds

in a population with average replacement of 1 tree

• Demography
• Fecundity: few individuals
• Recruitment:
• Understory water, light
• Important dynamics limited to
• pen environments
• Attribution: noisy data, but

compound interest

• Pathogen regulation
• Damping off common
• Poorly understood, hard to

attribute

• Complications: both host and

pathogens respond similarly

A novel invader

• Novel traits leading to:
• Super-competitors—novel traits that beat the pervasive limitations
• Release from natural enemies—vulnerable recruitment stages
• Introduction and spread:
• LDD of pollen (Quercus, Pinus) and seed (wind Populus, many by animal

vectors)

• Hybridize (e.g., Quercus)
• Recruit: not only open environments, but also understory invasion
• Summary: abundant pollen and seed, hybridize, withstand

fungal attack and understory competition

Super competitor potential

Ailanthus beats the limitations in competitive stands:

• Early maturation, high fecundity, root suckering
• Establishes in the understory
• Must withstand pathogen attack
• Allelopathy

Monitoring can be critical: Ailanthus invading closed forest

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• 2000

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2001

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2001

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2002

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2003

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2004

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2005

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2006

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2007

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2008

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2009

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2010

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2011

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2012

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2013

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2014

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

Ailanthus invading closed forest

• ●
• ●
• 2015

DF_BW

50 m

• ●
• ●
• 10

2000 2005 2010 2015 10 20 30 40 50 Year Seeds per m2 0.00 0.02 0.04 0.06 Basal area (m^2/ha) DF_BW

The value (and limits) of monitoring

• Critical insight on changing size-species structure (SSS)
• Track the full demographic process for Ailanthus invasion
• Expensive: could not devote this effort widely and for

unknown future invaders

• How much monitoring?

The pace of species loss, fecundity anticipates mortality

2000 2005 2010 2015 Year 0.0 0.5 1.0 1.5 2.0 Basal area (m^2/ha) CW_LG CW_UG 1990 1995 2000 2005 2010 2015 Year Seeds per m2 10 20 0.0 0.2 0.4 0.6 0.8 1.0 CW_218

Cornus and dogwood anthracnose Tsuga and hemlock woolly adelgid

Undetected change: Mult

ltiyear drought‐induced morbidity precedin ing tree death in in south-eastern U.S .S. . forests

Berdanier and Clark (2016) Ecol Applications

• Protracted death thwarts attribution

Demography: when is it an indicator?

• Predictable growth,

volatile fecundity

• Sensitivity ≠

vulnerability

• Small changes

magnify with compound interest

Clark et al. GCB 2011

Fecundity Growth

Monitoring may not document changing range li limits

Expanding N, retreating from S Not evident in data Contrasting trends in SST

Zhu et al. GCB (2012)

Do bio iotic in interactions control range li limits?

Katz and Ibanez, 2016, J. Ecology

S ← → N

• range edges: small effects on annual rates

unclear

• Closed forest are not the place for rapid

change

The value of monitoring SSS (a (and when is it over-rated?)

• Demography: the processes of change
• SSS: the consequences of change
• Careful design needed where biotech risks

anticipated

Modeling important

• Answers limited by data or by the models that could

exploit them?

• Demography/SSS models could be better: fit at the

scale where predictions needed

• Biotech risks are not the place to oversell predictive

capacity of better models

Contemporary climate trends

• Warm, short winters: reduced

snow cover, soil freezing, fine root death

• Energized atmosphere: runoff,

peak flow, storm surge

• Drought: moisture stress,

reduced streamflow

• Fire frequency, size, severity:

interaction with fuel loads

• Biological stressors: invasive

plants, pathogens, insects

• All affect the SSS

Current understanding of climate impacts: slow change, hard to evaluate

• High-profile diebacks in the western US
• Interactions involving drought, insects, pathogens, fire
• Fast change, aftermath slow, many decades
• Eastern US mixed
• Clear impacts on growth
• cryptic effects on survival –protracted morbidity complicates

attribution of death

• large (but volatile) fecundity response combines with masting cycles
• recruitment responses slow, poorly understood

Models could be better: Species distribution models (SDMs)

Individual Population Community

climate abundance diversity, productivity

• bserved in

boxes Inference Extrapolate

Scale mismatch: Inference on species, predict biodiversity

Models could be better: Demographic models: PDEs, MPMs, IPMs, IBMs

Individual Population Community

size structuret demography tree size, survival, fecundity climate

• bserved in

boxes Inference Extrapolate

Scale mismatch: Inference on individuals, predict population

Processes are individual

growth survival reproduction

processes: weather, competition sample size: # individuals absence: no data responses: growth, survival, fecundity prediction: fitness limitation: no connection to population

weather

Individual

Predictions are community

# communities zero size-species distribution biodiversity complex

processes: weather, competition sample size: # individuals absence: no data responses: growth, survival, fecundity prediction: fitness limitation: no connection to population

Community

weather

Individual

Size-species-space-time (3ST)

Individual Population Community

size-species distribution demography tree size, survival, fecundity climate

• bserved in

boxes Inference Prediction

inference/prediction at the same scale data synthesis at latent processes stage

Acer negundo

Predictive distributions: Range limits controlled by competition (including interaction with climate)

• bserved at red dots

With competition Without competition

Main points

• Biotech risks
• Escape: pollination, seed

dispersal, hybridization

• Super competitors that beat

the pervasive limitations:

• Fecundity—few individuals do it
• Overwhelming light/moisture

limitations in closed stands

• Release from natural enemies that

are poorly understood

• Pathogens/insects released

from their natural enemies

• The role of monitoring: when

change is happening

• Design critical
• Models could do better
• but not much
• model fitting/prediction at the

same scales

• Not at the expense of

monitoring/experiments