Transcript Annual Precipitation (mm) (average over Prairie Ecozone)

Vulnerability of Prairie
Grasslands to Climate Change
Jeff Thorpe
Saskatchewan Research Council
March, 2011
Copyright © SRC 2010
smart science solutions™
Modeling future climates:
•Use 1961-90 normals as baseline (PRISM model)
•Select future climate scenarios based on Global
Climate Models:
•Future climates projected for the 2020s, 2050s, and
2080s.
Scenario
ECHAM4 A2 (cool scenario)
Manitoba
Sask.
Alberta
X
X
X
GFCM20 B1
X
CGCM3_T47_2 A1B
X
MIMR B1
X
HADCM3 B2
X
HADCM3 A2 (warm scenario)
X
CSIRO2 B2
X
X
X
Growing Degree-Days (average over Prairie Ecozone)
3500
3000
2500
AB, warm scenario
AB, cool scenario
2000
AB, baseline
SK, warm scenario
SK, cool scenario
SK, baseline
1500
MB, warm scenario
MB, cool scenario
MB, baseline
1000
500
0
1960
1980
2000
2020
2040
2060
2080
2100
Potential Evapotranspiration (mm) (average over Prairie
Ecozone)
900
800
700
AB, warm scenario
AB, cool scenario
AB, baseline
SK, warm scenario
600
SK, cool scenario
SK, baseline
MB, warm scenario
MB, cool scenario
500
MB, baseline
400
300
1960
1980
2000
2020
2040
2060
2080
2100
Annual Precipitation (mm) (average over Prairie Ecozone)
900
800
700
AB, warm scenario
AB, cool scenario
AB, baseline
SK, warm scenario
600
SK, cool scenario
SK, baseline
MB, warm scenario
MB, cool scenario
500
MB, baseline
400
300
1960
1980
2000
2020
2040
2060
2080
2100
Proportion of Precipitation in Summer (average over Prairie
Ecozone)
0.75
0.70
AB, warm scenario
AB, cool scenario
0.65
AB, baseline
SK, warm scenario
SK, cool scenario
SK, baseline
0.60
MB, warm scenario
MB, cool scenario
MB, baseline
0.55
0.50
1960
1980
2000
2020
2040
2060
2080
2100
Modeling of vegetation zones in
relation to climate
• Most existing studies deal with the forest/grassland boundary,
but not with vegetation zones within the grassland.
• SRC model based on vegetation zones from Canadian
Prairies to Colorado and Nebraska
• Use the U.S. as an analogue for a warmer future Canada.
• Three climatic variables (1961-90 normals):
• annual precipitation
• proportion of precipitation in May-Sep
• annual potential evapotranspiration
• Statistical relationships between climate and vegetation zone.
Ecoregions of Canada
Vegetation types in the U.S.
State and transition diagrams
• Standard tool used in range management
for representing vegetation dynamics.
• Boxes represent vegetation states and
community phases within each state.
• Arrows represent types of transition
between them.
Dry Mixed Grassland, SW Saskatchewan
Crested
Wheatgrass –
native
grasses
←exotic
invasion
Northern
Wheatgrass –
Needle-and-thread
↓heavy grazing
recovery ↑
Needle-and-thread
– Wheatgrass –
June Grass – Blue
Grama
↓heavy grazing
recovery ↑
Blue Grama –
Needle-and-thread
– June Grass –
Western
Wheatgrass
Dry Mixed Grassland, SW Saskatchewan
Crested
Wheatgrass –
native
grasses
←exotic
invasion
climate change
(cool scenario)
→
Blue Grama –
Needle-andthread –
Western
Wheatgrass
climate change
(warm scenario)
→
Blue Grama –
Needle-andthread –
Western
Wheatgrass
Northern
Wheatgrass –
Needle-and-thread
↓heavy grazing
recovery ↑
Needle-and-thread
– Wheatgrass –
June Grass – Blue
Grama
↓heavy grazing
recovery ↑
Blue Grama –
Needle-and-thread
– June Grass –
Western
Wheatgrass
climate
change
(warm
scenario)
→
Blue Grama Buffalogass
Aspen Parkland, North Battleford
Snowberry – native grasses
↓fire
Smooth
Brome –
native
grasses
no fire↑
←exotic
invasion
Plains Rough Fescue –
Northern Wheatgrass –
Western Porcupine Grass
Kentucky
Bluegrass –
native
grasses
←exotic
invasion
↓heavy grazing
recovery ↑
Western Porcupine Grass –
Northern Wheatgrass –
Sedge – Pasture Sage
↓heavy grazing
recovery ↑
Sedge – Pasture Sage –
Western Porcupine Grass –
Northern Wheatgrass
Aspen Parkland, North Battleford
Snowberry – native grasses
↓fire
Smooth
Brome –
native
grasses
no fire↑
←exotic
invasion
Plains Rough Fescue –
Northern Wheatgrass –
Western Porcupine Grass
Kentucky
Bluegrass –
native
grasses
←exotic
invasion
↓heavy grazing
recovery ↑
Western Porcupine Grass –
Northern Wheatgrass –
Sedge – Pasture Sage
↓heavy grazing
recovery ↑
Sedge – Pasture Sage –
Western Porcupine Grass –
Northern Wheatgrass
climate
change
(cool
scenario)
→
Western
Porcupine
Grass –
Northern
Wheatgrass
climate
change
(warm
scenario)
→
Northern
Wheatgrass –
Needle-andthread
climate
change
(warm
scenario)
→
Blue Grama
– Needleand-thread –
Western
Wheatgrass
Implications of zonation model:
• Zonation model does not give exact prediction of future
vegetation:
• shifts in species may lag behind changes in climate.
• new combinations of species could result from
differences in migration rate.
But zonation model shows probable
future trends:
•gradual reduction in tree and tall shrub cover.
•regeneration failure after disturbance in the boreal fringe
•shrinking aspen groves
•reduced woody encroachment on grasslands
•shifts in structure of grasslands: decrease of
midgrasses, increase of shortgrasses.
•decrease in cool-season grasses, increase in warmseason grasses (other literature supports this).
•gradual introduction of plant and animal species
currently found only in the U.S. (e.g. buffalograss, big
sagebrush).
Modelling of grassland production
•Same three climatic variables (1961-90 normals):
•annual precipitation
•proportion of precipitation in May-Sep
•annual potential evapotranspiration
•Measured grassland production at various locations in
Canada and the United States.
•Restrict to loamy upland soils (Loam Ecosite).
•Statistical relationship between climate and production.
Grassland Production on Loam (kg/ha)
3500
3000
Dauphin, baseline
Dauphin, cool scenario
Dauphin, warm scenario
2500
SE Manitoba, baseline
SE Manitoba, cool scenario
SE Manitoba, warm scenario
Lloydminster, baseline
2000
Lloydminster, cool scenario
Lloydminster, warm scenario
Estevan, baseline
1500
Estevan, cool scenario
Estevan, warm scenario
Cardston, baseline
Cardston, cool scenario
1000
Cardston, warm scenario
49°N/110°W, baseline
49°N/110°W, cool scenario
500
0
1960
49°N/110°W, warm scenario
1980
2000
2020
2040
2060
2080
2100
Why not larger decreases in
productivity?
• Large increase in potential evapotranspiration (PET)
suggests much lower moisture indices (“desertification”).
• However, precipitation has a bigger effect than PET.
• Campbell et al. (1997): 90% of the variance in
productivity in grasslands can be accounted for by
annual precipitation.
• Rise in temperature has a secondary negative effect,
probably because of direct evaporation from the soil.
• Decrease in proportion of precipitation in summer also
has a secondary negative effect.
• Other literature supports conclusion of modest
productivity changes.
Carbon fertilization effect
• These models do not account for possible carbon
fertilization effect:
• rate of photosynthesis increases with ambient CO2
concentration.
• stomatal conductance decreases, meaning improved water use
efficiency.
• Field experiments with CO2 enrichment chambers:
• average global increase in grassland production of 17%.
• greatest response when moisture supply is limited.
How important is carbon
fertilization?
• Is carbon fertilization effect large enough to compensate
for the effect of a drier climate?
• Other factors such as heavy grazing or nutrient
deficiency could reduce the ability of plants to take
advantage of carbon fertilization.
• Some research shows that forage quality declines under
carbon fertilization, so cattle would have to consume
more to achieve a given weight gain.
• Overall, effect of carbon fertilization in our grasslands is
uncertain, but it may help to reduce the impact of a drier
climate.
Effects of Drought
• These models represent the average climate (30 year
normals) – what about year-to-year variation?
• Droughts are a characteristic feature of grassland
climates.
• immediate response – reduced grassland productivity
• multi-year response – shift in species composition from
taller to shorter species.
• Some studies indicate that climate change will
increase variability in precipitation, possibly resulting in
more frequent and more intense droughts.
• This could be more important than the changes in
average productivity.
Year-to-year variation in measured
production at Manyberries, Alberta
1000
900
800
700
600
500
yearly
average
400
300
200
100
0
1930
1940
1950
1960
1970
1980
1990
2000
2010
Yearly Production at Manyberries, and Effect
of Climate Change on Average Production
1000
900
800
production (kg/ha)
700
600
measured
500
cool scenario
warm scenario
400
300
200
100
0
1920
1940
1960
1980
2000
2020
2040
2060
2080
2100
Modeling effect of drought on
production
• SRC Forage Calculator
• use “forage-year precipitation” – sum from
previous September to current August.
• calculate percent deviation of current forageyear precipitation from long-term average.
• predict percent deviation of current forage
yield from the long-term average.
• apply to average forage yield to estimate
current forage yield.
Climate change and effect of
drought
• based on variability of precipitation from historic
data.
• use statistical distribution to estimate “dry
threshold” – precipitation level below which the
driest 20% of years fall.
• calculate drought threshold for precipitation in
future scenarios (warm scenario shown).
1. assume same variability as in historic data
2. assume increasing variability
Drought and grassland
composition
• Changes during the drought of the 1930s were well
documented.
• General U.S. trends – taller grasses decreased,
shorter grasses increased; elimination of most forbs,
increase of cactus.
• In northern mixed prairie (including Canada), the
main change was increase of early-growing species:
Sandberg’s bluegrass, June grass, sedges.
• Impacts of drought were made worse by the heavy
grazing practiced at that time.
Drought and woodlands
• During the drought of 2001-2002, average growth of
aspen stands declined to near zero.
• Drought interacts with outbreaks of forest tent caterpillar
in reducing growth.
• Substantial mortality of aspen (“aspen dieback”) was
observed, especially in the Aspen Parkland.
• Tree mortality during droughts could be one of the major
processes in shifts in vegetation zonation.
Drought and animals
• Most grassland birds are less abundant in dry years
than wet years.
• Shifts in species: e.g. drought of 1988 in North
Dakota:
• Grasshopper Sparrow, Sprague’s Pipit, Clay-colored
Sparrow, and Baird’s Sparrow decreased.
• Horned Lark and Western Meadowlark became more
dominant.
• A wide range of insects declined during the drought of
the late 1980s.
• However, outbreaks of plant-eating insects
(grasshoppers, forest tent-caterpillar) are often
preceded by warm, dry weather.
Climate change and wetlands
• It is well known that the number of wetlands and number
of ducks depend on weather cycles, declining in dry
years.
• In the long term, models predict decreasing pond
numbers and duck populations with climate change.
• The most productive area for ducks, in southeastern
Saskatchewan, southern Manitoba, and the Dakotas, will
become a more episodic, less reliable source of
waterfowl production, similar to the drier areas further
west.
• Favourable water and cover conditions will be found
further north and east.
• Interaction with land use: drainage of wetlands
exacerbates impact of climate change.
Classification of wetlands (Stewart
and Kantrud 1971):
General impacts on biodiversity –
geographic shifts
• One way species can adjust to climate change is by
moving their ranges.
• Globally, average range shift 6.1 km northward per
decade over 20th Century (Parmesan and Yohe 2003).
• Species vary in migration rate, so there will be sorting of
species along the migrational front, led by the most
invasive and trailed by the least invasive.
• Impacts of fragmentation - habitat specialists with poor
dispersal ability will be the least able to keep pace with
climate change.
Phenological shifts:
• Another way in which species can adjust to climate
change is by phenological change.
• Globally, average shift toward earlier spring timing of 2.3
days per decade through the 20th Century.
• At Edmonton, first-flowering date of trembling aspen
advanced by 26 days.
• At Delta Marsh, 25 of 27 bird species showed earlier
arrival dates over a 63 year period.
• Phenological change can lead to mismatches in timing
between predators and prey, pollinator and plant, etc.
• Snowshoe hare – colour change is driven by daylength,
so may be mismatched with snowmelt.
Evolutionary change
• Another form of adaptation in place is evolutionary
change.
• Contemporary evolution in thermal tolerance has been
observed in frogs, insects, and plants.
• But probably less important than range shifts.
• e.g. following Pleistocene glaciation, there was northward
movement of existing species rather than evolution of new
species.
• Weedy species are likely to show fastest evolution in
response to climate change, because of large population
size and short generation time.
Advantages of invasive species
under climate change
• Faster evolution (adaptation in place).
• Efficient dispersal allowing faster range shifts.
• Large native ranges, indicating broad climatic
tolerances.
• High habitat connectivity because of use of
disturbed habitats.
Increasing susceptibility to invasion:
• Climate change could be a stress that makes
communities more susceptible to invasion.
• Existing late-successional plant species could become
increasingly ill-adapted to the climate, so more likely to
be out-competed by newly arriving species.
• However, invasion also depends on resources:
temporary surplus of water or nutrients favours invasion.
• In dry environments, invasion increases with water
availability, so increasing drought would actually reduce
the risk of invasion.
Species at Risk
• Grassland birds – impacts depend on the species.
• shift to shorter, more open grassland will reduce
habitat for Sprague’s Pipit but increase habitat for
Burrowing Owl.
• loss of tall shrubs will reduce habitat for Loggerhead
Shrike.
• Many of our species at risk are northern fringe
populations of species that are common in the U.S.
• examples: Buffalograss, Western Spiderwort, Hairy Prairieclover.
• climate change should increase the area of suitable climate for
these species.
Impacts on specific grassland types
Mixed Prairie – largest area of remaining
grassland.
• Mix of growth-forms and photosynthetic types, so shifts in
composition can occur initially by increase in species already
present.
• Northward migration of southern species is more likely to be
successful because of larger grassland area and relatively lower
level of fragmentation.
• Dry climate and incidence of drought will reduce risk of exotic
invasion.
• However, prolonged, severe droughts leading to soil erosion are
more likely in the Mixed Prairie.
• Livestock producers already practice conservative stocking and plan
for drought so may be better equipped to deal with climate change.
Northern Fescue Prairie
• Lower grassland area and higher fragmentation
compared to Mixed Prairie.
• Shrinking aspen groves and reduced woody
encroachment could benefit grasslands and livestock
grazing.
• Existing fescue grasslands will shift toward mixed prairie
composition – loss of a unique community.
• Shifting of fescue northward will be impeded by high
fragmentation of native habitats. Native habitat to north
is boreal forest – unknown successional pathways as
trees decline.
• Greater risk of exotic invasion – already a bigger
problem in Northern Fescue Prairie.
Foothills Fescue Grassland
• Zonation models indicate that this type will
persist in its current location.
• May shift towards characteristics of
Montana foothills grasslands.
Manitoba grasslands
• Part of Aspen Parkland climatically, but with shifts in
composition eastward (tallgrass species, Kentucky
bluegrass), and with extensive poorly drained soils.
• Risk of in exotic invasion with climate change will be
greatest in Manitoba because of moist climate and moist
soils. However, shift toward somewhat drier Mixed
Prairie may reduce exotic invasion and woody
encroachment.
• Tallgrass species are generally warm-season (C4)
grasses, which are expected to benefit from climate
change. However, future climate may not provide
enough moisture over much of southern Manitoba.
• Problems of excess moisture in extreme wet years are
greater concern in Manitoba.
New grasslands on former forest
land
• Decline of forests will create new grasslands – could
significantly increase the area of grazing land. But what
kind of grasslands?
• With initial canopy opening, forest understory species
will grow more vigorously.
• With complete loss of tree cover, some forest shrubs
could persist, but most forest herbs will eventually be
replaced by sun-adapted species.
• Best-case scenario is forest surrounded by native
grassland; propagules will gradually spread into the
opening forest, and increasingly out-compete the forest
herbs as light levels increase.
Former forests:
• But in many situations the propagules that
are most available and most aggressive
will be exotics such as smooth brome and
Kentucky bluegrass.
• may already be in the grassland
• may have already invaded the forest
• nearest source of sun-adapted species may
be roadsides or fields seeded to tame forage.
The resilience of native
grassland
• Prairie grasslands have evolved in a highly variable
climate – higher drought tolerance than forests or
annual crops.
• Prairie grasslands are made up of a mix of species:
• taller and shorter species
• warm-season and cool-season species
• drought-tolerant and moisture-requiring species
• The grasslands of the Great Plains show continuous
variation over a huge area (Canada to Mexico).
Resilience continued:
• The drought of the 1930s has shown that native
grasslands can adjust by shifts in the proportions of
species, while maintaining a grassland ecosystem.
• So there is a large potential for prairie grasslands to
adjust to changes in climate.
Barriers to resilience
• Species that are already here can shift in
abundance (e.g. increase in native warm-season
species).
• But habitat fragmentation may slow the movement
of southern species into Canada.
• Maintaining connections between habitats will
facilitate range shifts.
• Idea of “assisted migration” – human intervention to
help southern species to migrate northward.
Maintaining range health as an
adaptation to climate change
• Vigorous plants are better able to cope with climatic
stress.
• Litter cover helps to “drought-proof” grassland.
• Increased efforts to control exotic invasion may be
needed under climate change.
• Good range management practices will become
increasingly important.
Producers already adapt to
drought
• Predict production in the coming year based on fall and
winter precipitation.
• Heavier culling of herd than in normal year.
• Keep part of herd in yearlings – reduce number in dry
years.
• Rent alternative grazing land in moister regions.
• Keep portion of grazing land as reserve for dry years.
• Sow annual crops for grazing.
• During years of poor crop production, turn cattle out on
crops rather than harvest them.
• Buy more feed.
Producers and drought
• Producers in drier regions experience more frequent
drought, so tend to have emergency plans in place.
• making contacts for alternative grazing
• maintaining larger feed reserves
• grazing more conservatively
• With climate change, producers in moister regions may
have to shift their management in this direction.
Need for monitoring
• Rangeland monitoring networks can provide the
information needed to assess the impact of climate
change.
• shifts in species composition
• long-term trends in productivity – use to adjust
stocking rate recommendations.
• Examples:
• Alberta Rangeland Reference Area Program (183
plots)
• Manitoba Forage Benchmarking Project (63 plots)