presented - Global Change Consulting Consortium, Inc.

advertisement
Overview of the biology of extreme events
Vincent P Gutschick
Global Change Consulting Consortium
Las Cruces, NM
vince.gutschick@gmail.com)
Hormoz BassiriRad
Biological Sciences, University of Illinois
Chicago, IL
hormoz@uic.edu
2008 Fall Meeting
American Geophysical Union
San Francisco, CA
15-19 December 2008
PowerPoints available at:
http://biology-web.nmsu.edu/vince/agu08.ppt
and ../agu08long.ppt
KH Raffa et al., BioScience 58(2008): 501-517
Extreme events are…
1) … rare, and impactful … attention-getting
2003 heat wave in Europe
- for humans
- for atmospheric fluxes
Hurricane Katrina
Conifer dieoff across Western North America
Rise of atmospheric CO2
Snowball Earth
…and many more
….appearing (but not really) categorically different from
(more) normal events  extremity is a distribution
Extreme events are…
1) … rare, and impactful … attention-getting
2) Inordinately weighted in the physiology, ecology, and
evolution of organisms
Hence, also, in biogeochemical cycles
Katrina: 0.1 Pg C reinjection expected from tree death
Conifer dieoff: 0.3 Pg C reinjection
2003 heat wave in Europe (-NPP, -Rd, -WUE, -APAR; drought)
Snowball Earth: end of CH4 atmosphere, major greenhouse
Yet, mostly anecdotal to date – no comprehensive framework
Extreme events are…
GF Gravatt, Unasylva 3(1949):3-7)
1) … rare, and impactful … attention-getting
2) Inordinately weighted in the physiology, ecology, and
evolution of organisms
3) Of diverse origins:
● meteorological – heat wave, hurricane
● biotic
- evolutionary – Snowball Earth
- anthropogenic
– land-use change; deforestation  ΔT, ΔP (incl. monsoons)
- introduction of exotics (chestnut blight,
Dutch elm, Oz rabbits…)
● orbital – ice ages
● tectonic- Toba eruption, 70 kya
Extreme events are…
1) … rare, and impactful … attention-getting
2) Inordinately weighted in the physiology, ecology, and
evolution of organisms
3) Of diverse origins: meteorological, biotic, orbital, tectonic
4) Occurring/recurrent, over wide scales
- of time
Mega-years: Snowball Earth, C4 plant evolution
Years: Darwin's finches, evolution of beak size
Hours or days: flower drop in extreme T events
- of space (same examples)
JR Ehleringer, Oecologia 95(1993): 340-346
Extreme events are…
…….
5) For biological EEs…and for much of consequent effects on
climate, hydrology, etc.:
A) Defined by organismal effects, not drivers
Chihuahuan desert: heat, drought effects  native vs. introduced
B) Physiological in effect – Leaf damage from heat, frost, cold+light
And/or ecological – Cheatgrass and fire regime; pollinator timing
And/or evolutionary
Toba eruption signature on human genome
20th-year extreme drought: Ci/Ca selected in Encelia farinosa;
Pinecone scale size selection in abundant years
Microsites  soil water availability  glycolytic enzyme in pinyon pine
Extreme events are…
…….
5) For biological EEs…and for much of consequent effects on
climate, hydrology, etc.:
A) Defined organismally
B) Physiological, ecological, evolutionary
C) Likely more important than mean values of conditions
Recent manipulative experiments
Natural conditions: frosts structuring ecosystems
Fitness effects likely concentrated in EEs
D) Likely: fitness effects concentrated in the recovery phase
Drought recovery: resource acquisition restarts in recovery, not event
Recovery can be protracted, even for native species
Extreme events are…
PM Brown & R Wu, Ecology
86(2005):
3030-3038
1) …
rare, and impactful
… attention-getting
2) Inordinately weighted in the physiology, ecology, and
evolution of organisms
3) Of diverse origins: meteorological, biotic, orbital, tectonic
4) Occurring/recurrent, over wide scales of time and space
5) For biological EEs: A) Defined organismally; B) Physiological,
ecological, evolutionary; C) More important than “normal”
conditions; D) Often more evident in recovery phase
6) Positive as well as negative: e.g., pluvial events for forest
reestablishment after fire
Extreme events are…
……
Aust. J. Plant Physiol. 25(1998): 27-37
6) Positive as well as negative: e.g., pluvial events for forest
reestablishment after fire
7) Generally defined by sequences, not points in time;
Order and rate of driver events matters
High T's in spring vs. winter;
Cold-weather leaf drop in Larrea tridentata of warm deserts
High-T flower drop in Opuntia cacti in spring (see earlier slide)
Hardening to low T in Eucalyptus pauciflora
B) Thus, they are defined as events, not trends
Trends increase/ decrease probability of pointwise extremes in
driving variables, such as T
Example: Eucalyptus pauciflora in pastures
Time scales important – rise of CO2: trend  event in evolutionary time
Extreme events are…
……
5) For biological EEs: A) Defined organismally; B) Physiological,
ecological, evolutionary; C) More important than “normal”
conditions; D) Often more evident in recovery phase
6) Positive as well as negative: e.g., pluvial events for forest
reestablishment after fire (Brown & Wu 05)
7) Generally defined by sequences, not points in time;
8) Often defined by multiple variables (e.g., drought and T,
in conifer dieback of 2000 ff.)
9) And by cascades, which may involve other organisms
Drought, high T  beetle outbreaks  conifer death
Reduced CH status
Increased CO2  Altered T regime  Frost damage in eucalypts
Altered frost sensitivity
Extreme events are…
……
7) Generally defined by sequences, not points in time;
8) Often defined by multiple variables (e.g., drought and T,
in conifer dieback of 2000 ff.)
9) And by cascades
10 Because of sequence-dependence, multiple correlated drivers,
and cascades: challenging to describe statistically
-even if only observationally, much less, predictively
For defined sequences: ARIMAs, etc.
Who knows the important sequences, as for frost damage?
Can GCMs do adequate statistics of extremes in drivers?
Challenging, even with rainfall timing
Simple observations: now finding heavy tails in floods – other
surprises
waiting?
A Dai & KE
Trenberth, J. Climate 17(2004): 930-951
Extreme events are…
……
7) Generally defined by sequences, not points in time;
8) Often defined by multiple variables (e.g., drought and T,
Science
2068-2074
in conifer dieback of
2000 289(2000):
ff.)
9) And by cascades
10) Because of sequence-dependence, multiple correlated drivers,
and cascades: hard to do statistics upon
11) Changing in spectrum
A) Drivers themselves, as shifts in PDFs – e.g., extremes in precipitation
Annual amounts, or heavy events
B) Crossing thresholds – recasting of even the form of PDFs
Non-analog climates
Thermohaline circulation shutdown in past, and perhaps in future
 New “normals” can constitute extremes for extant genotypes
Extreme events SJ
are…
Crafts-Brandner & ME Salvucci, Plant Physiol. 129(2002): 1773-1780
……
8, 9, 10) Often defined by multiple variables, cascades  challenging
statistics
11) Changing in spectrum
12) Thus, challenging to predict for effects
E.g, “March of the trees” with rising CO2 & climate change:
A) Potential effects on radiative balance and biogeochemical fluxes
B) Responses are in physiological time, of the individual organism:
May depend as much on extremes as on means
New climate regimes (non-analog climates) may constitute extreme
Physiological & ecological responses to EEs: poorly known
Acclimation is response - may reach limits in rates (Rubisco activase
in maize) or amounts  dysacclimation as a definable extreme)
Direct responses to CO2 :poorly known, and extremely
variable  potential biogeographic chaos (VP Gutschick,
Ecol. Mod. 200(2007): 433–451)
Annual mean precip.
Extreme events are…
…
12) Thus, challenging to predict for effects
A, B) radiative balance, biogeochemical fluxes; physiological time
C) Response in evolutionary time – nuanced, at the very least:
Fitness effects are likely concentrated in extreme events
Responses are constrained by population genetic structure
Who says responses are adaptive?
Neutral theory of Kimura (1983 ff.)
Adaptive responses to high CO2 – almost all lost in 20 My – N dynamics
D) How good are our predictive capacities fordrivers of EEs?
Point events – observations (return times) – good
Correlated multiple variabales – in development
GCMs’ abilities to capture spectrum (PDF) of point events – fair
GCMs’ abilities to capture sequences that define EEs – unknown
JP Iorio et al., Climate Dyn. 23(2004): 243-258
Biotic effects, such as disease prevalence- !!
Recap: lots of specifics to account for, lots of conceptual
and mathematical challenges
* Inordinately weighted in physiology / ecology / evolution / biogeochem cycles
* Diverse in origins
* Diverse in scales of time & space
* Defined by organismal response, not drivers
* Likely more important than mean values
* Can be positive, too
* Defined by sequences, not points; multiple variables; cascades
* Challenging for statistics
* Changing in spectrum
* Challenging in knowing organismal responses, even for adaptiveness
* Challenging to capture proper statistics in their (meteorological…) drivers
We need observational tools, and, more so,
a comprehensive framework, with
* biology: physiology, ecology, evolution
* meteorology and climatology
* statistics
Who's going to plan it?
OR will we default and live with greater uncertainty
than we need to for prevention and amelioration?
PowerPoints available at:
http://biology-web.nmsu.edu/vince/agu08.ppt
and ../agu08long.ppt
Download