Transcript asm_poster_final_doug

Prescribed Fire Effects on Above-Ground Net Primary Production in a
Southwest Desert Grassland
Douglas Moore, Esteban Muldavin, Scott Collins,and Karen Wetherill
Sevilleta Long Term Ecological Research Program
Biology Department, University of New Mexico, Albuquerque, NM 87131
Abstract
2004
The effects of a fire on aboveground net primary productivity (ANPP) in a southwestern desert
grassland were studied. In June, 2003, a prescribed burn was applied to a large area (5,000 ha)
of desert grassland and shrubland on the east side of the Sevilleta National Wildlife Refuge.
The area was dominated by black grama (Bouteloua eriopoda), blue grama (Bouteloua gracilis),
and creosotebush (Larrea tridentata) in pure and mixed stands. The summer growing season
following the fire was one of the driest on record (54 mm – mean is 144 mm), and, based on
visual inspection, little or no growth occurred on either burned or unburned sites. In the spring
of 2004, paired, side-by-side aboveground net primary productivity (ANPP) plots were
established in burned and adjacent unburned areas representing the major community types: 1)
black grama dominated, 2) a mixture of black and blue grama, and 3) a creosotebush stand with
an understory of black grama. The fire did not travel significantly into creosotebush shrublands
without grass nor into pure blue grama grasslands, and hence these communities were not
sampled. The plots consisted of grids of 40 permanent 1-m2 quadrats at 10-m intervals. They
were sampled in the spring and fall of 2004, and in the winter, spring and fall of 2005. Sampling
followed the standard Sevilleta ANPP double-sampling protocol in which volume by species was
estimated on each quadrat and regressions of weight-to-volume for each species were
developed by harvesting the species of various sizes from adjacent areas. While 2003 was dry,
both 2004 and 2005 had above average moisture (298 mm and 340 mm – mean is 257 mm)
that generated significant growth responses. While the fire generally removed creosotebush
canopies, the large majority of plants resprouted from root crowns. Among grasses, black
grama mortality was high on burned sites and subsequent production was primarily among
bunches that escaped fire. Blue grama was not significantly impacted by the fire and most
plants recovered. Overall, on burned plots ANPP was higher in black grama sites, but the
relative abundances of the dominants changed following the fire from B. eriopoda to forbs such
as Sphaeralcea wrightii and other grasses (Sporobolus spp.) Production in the
Creostebush/Black Grama site was still significantly depressed relative to other sites after three
years. The timing of the burn at the peak of an 12-month drought may have intensified impacts
to production. Indications are that full recovery, if any, may be a decadal process, particularly
with respect to black grama and creosotebush communities.
Results
2006
Black Grama
Black grama site ANPP
150
ANPP (g/m2)
125
100
Forbs
Other shrubs
75
Other grass
Black grama
50
25
2004
2005
Burn
Blue Grama/Black Grama
2004
• Black grama ANPP was significantly reduced
on the burned site, primarily as a function of
mortality, and there was little recovery through
2005.
• In contrast, forb production was significantly
higher on the burned site and made up as much
as 70% of the ANPP.
• As a result, burning resulted in total ANPP that
was greater than or equal to that in the
unburned plots.
0
2005
Control
Blue Grama -Black Grama Mixed Site
Mixed grass site ANPP
150
ANPP(g/m2)
125
Introduction
Forbs
100
Other shrubs
Other grass
75
Blue grama
50
Black grama
• While black grama ANPP was again
significantly lower on the burned site, blue
grama productivity was greater than on the
unburned plots.
• Burning did not result in increased forb
production as seen in the Black Grama site
25
• Fires are thought to significantly impact vegetation community
composition and structure in desert grasslands of the Southwest,
but their effect on aboveground net primary productivity (ANPP)
production has been little studied.
• Hence, total ANPP was lower on the burned
site than in the unburned site.
0
2004
2005
2004
Burn
Creosotebush/Black Grama
• Following a landscape-scale prescribed burn on the Sevilleta
National Wildlife Refuge, in June 2003, ANPP monitoring plots
were established in burned areas previously dominated by, 1)
black grama, 2) a mixture of black grama and blue grama, and 3)
creosotebush in a matrix of black grama. Unburned plots of
similar pre-burn compositions were established in adjacent sites
for comparison.
2005
Control
Creosotebush/ Black Grama Site
Creosotebush/grass site ANPP
• Creosotebush production was significantly
lower on the burned site and attributed mostly to
resprouting shrubs.
150
ANPP (g/m2)
125
• Here, we evaluate the ANPP responses following the first three
growing seasons (through 2005) in terms of the effects on
selected dominant species and life forms.
Forbs
100
Other shrubs
Creosotebush
75
Other grass
50
Black grama
25
0
2004
2005
Burn
Precipitation (mm)
Water-year precipitation (2001-2005)
400
Precipitation
350
•The burn was conducted at the peak of a severe drought that engulfed
the region the previous 12 months.
• Summer 2003 was one of the hottest and driest on record and the
monsoon rains that did occur arrived late, near the end of the growing
season.
• 2004 and 2005 were above normal for winter precipitation while summer
precipitation was close to normal.
300
250
Jun-Sep
200
Oct-May
150
100
0
2001
2002
2003
2004
2004
2005
Control
• As with the other sites, black grama ANPP was
lower on the burn.
• Forb production was only slightly increased on
burned plots.
• Consequently, total production in the burned
site was only about half of that on the unburned
site.
Discussion
50
• The effect of fire on total ANPP was different among desert grassland community types. While
black grama had significant declines without recovery across all three types in response to fire, this
was compensated for by increased forb production on the Black Grama site and increased blue
grama production on the mixed Blue Grama-Black Grama site. Yet, there appeared to be no similar
compensatory effect on the Creostebush/Black Grama site.
2005
Water year (Oct - Sep)
ANPP Sampling
• Three grassland community types were sampled with paired burned and
unburned plots (Fig. 1B):
Black Grama, burned and unburned
Blue Grama-Black Grama mix, burned and unburned
Creosotebush/Black Grama mix, burned and unburned
Methods
Site and Fire Characteristics
• A prescribed burn of approximately 5,000 ha of desert grassland was
conducted from June 19-22, 2003 on McKenzie Flats on the Sevilleta
National Wildlife Refuge in central New Mexico (Fig. 1a).
• Each plot had 40 permanent 1-m2 quadrats either on a 10-m grid or as
part of a set of circular rodent-sampling webs (Black Grama Unburned).
• The mid-June timing mimics when lightning-caused fires are most likely
to occur, i.e., the hottest and driest time of the year and when dry-lightning
associated with the beginning of the Southwest monsoon provides an
ignition source, typically without rain.
• Sampling followed standard Sevilleta ANPP double-sampling protocols in
which volume by species was estimated on each quadrat and regressions
of weight-to-volume for each species were developed by harvesting the
species of various sizes from adjacent areas. Positive change in green
biomass for each species from one season to the next on each quadrat was
considered NPP.
• Wind condition were moderate to high, and the fire generally moved
quickly across the landscape with low residence times, but due to patchy
fuels and unexpected transient increases in humidity, much of the total
area still did not burn. However, study plots were completely burned.
Black Grama Site
Figure 1A. 2003 prescribed-burn location
on Sevilleta National Wildlife Refuge, NM.
Figure 1B. Location of aboveground net
primary productivity (ANPP) plots on
McKenzie Flats.
• Plots were sampled in the spring and fall of 2004, and in the winter, spring
and fall of 2005. In 2003, sampling was not conducted because the lack of
rainfall during the growing season precluded any significant production.
• Burn timing is important. The antecedent 12-month severe drought period before the burn in 2003
and the lack of rain in the remainder of the growing season that year may have led to increased
mortality for black grama, and possibly even creosotebush, and to intensifying declines in ANPP in
two out of the three ecosystem dominants within the burn area.
• The sizeable reduction in creosotebush production in the first three growing seasons suggests that
fire may be effective for shrub control in this ecosystem, but the lack of compensatory production of
either grasses or forbs, along with the creation of more open sites for potential shrub colonization
argues for caution in fire use, particularly over the long term.
Acknowledgements
This research was made possible by NSF Long Term Ecological Research Program and its support of the Sevilleta LTER.
We especially thank those people who have helped collect and manage the data: Maya Kapoor, Tessa Edelman Julieta
Betinelli, Charity Hall, Yang Xia, Michell Tohmey, Jay McLeod.
A special thanks to the U.S Fish and Wildlife Service and especially the Sevilleta National Wildlife Refuge and the care
taken by its fire team led by Don Kearny.