N Budgeting at LTERs:
Implications for C sequestration
Herman Sievering and Kate Lajtha
This workshop was attended by 54 ASM scientists and graduate
students. Discussion moved from a review
of the ecological community’s knowledge regarding N cycling and atmospheric N
deposition (Ndep) at forested LTERs and a few ILTER sites (reviewed by Bill Munger [HFR], Jed Sparks [HBR], and Herman Sievering [NWT])
to the need for a cross-site study across a gradient in Ndep
(e.g., from 5 [or less] kgN/ha-yr. to 30 [or more] kgN/ha-yr.). The
focus, it was agreed, should be on impacts of elevated N deposition &
ecosystem N retention upon forest ecosystem function, including
especially C sequestration (as determined by NEE, NEP,
biomass increment). At the workshop
itself, five LTERs - AND, BES,
CDR, HFR, NWT - and the SIK
(Hungarian Oak Forest) ILTER expressed interest in: undertaking collaborative
research (i.e., a cross-site study) with consistent experimental procedures and
protocols (i.e., to allow for data synthesis and coherent modeling inputs) to
elucidate the impact of globally changing N deposition. At the Res. Init.
#1, Altered Water and Nutrient Cycles, breakout group discussions that Kate Lajtha and Herman Sievering
convened (dealing with Qs #3 and #5), the CWT and HBR
LTERs, and the Bialowieza
Forest Polish ILTER indicated they would like to join the earlier
workshop-identified LTER/ILTERs in an, overall, nine-site
collaboration and N&C Relations field
experimentation project.
Interest at the workshop focused on forested ecosystems, not
only because the large majority of global land biomass is found within forested
areas, but also because initial evidence for a possible strong correlation of C
sequestration (as represented by annual NEE) across a gradient in Ndep was presented at the workshop. Figure 1 shows a statistically significant
dependence (p<0.05) of NEE, in gmC/m2-yr, versus Ndep (including fog/cloud intercepted N, where
found to be important, as well as wet plus dry N deposition), in gmN/m2-yr. The non-linear dependence of NEE on Ndep indicates there may be a leveling off in
the enhancement of NEE with increasing Ndep
once a loading of ~2 gmN/m2-yr. is
surpassed. The continued decrease in NEE
with loadings >2 gmN/m2-yr. may be a first piece of ecosystem
scale, quantitative evidence that large amounts of atmospherically-deposited N
at forest canopies leads to reduced productivity. The 24-site data base behind Figure 1
includes the consideration of precipitation, elevation, latitude, and LAI. Only precipitation entered
the multiple regression of NEE ahead of Ndep
and none of the other variables were significant (p>0.2). The 24-site Ndep
data have substantial uncertainty (+- 25-30%) that makes the statistical
strength of the Fig. 1 relationship suspect.
Participants concluded that a
coordinated cross-site study of changes in N deposition, and the resulting
effects on ecosystem function and C sequestration, would have a much greater
scientific impact than would a series of independent and uncoordinated N
deposition studies. Long-term,
cross-system research is now recognized as crucial to our understanding of
ecological phenomena and environmental change.
It was further concluded that only through consistent measurements and
coordinated experiments conducted at multiple locations with an Ndep gradient can LTER/ILTER scientists test the
generality of ecological models developed at single locations. Recently, such networks of coordinated
ecological research have yielded important insights on global and regional
scale controls on ecosystem N retention (NITREX and EXMAN; Tietema et al. 1997),
the role of respiration in the carbon balance of forests (EUROFLUX
and AMERIFLUX; Valentini et
al. 2000), decomposition (LIDET; Moorhead et al.
1999), trace gas emissions (TRAGNET; Ojima et al. 1992) and stream N dynamics (LINX; Peterson et
al. 2001). LTER sites have figured
prominently in the last three networks, among others. There is no current long-term network
of coordinated experiments that combine a detailed examination of N inputs to
forested ecosystems with studies of changes in C sequestration processes, yet
such a long-term, cross-site perspective is critical for models of global
change.
Discussions solidified two tentative goals that may (upon
future discussion) be chosen as project foci:
a) The
accurate measurement of N deposition (including N species contributing to dry,
cloud, and fog deposition) across a gradient from relatively low N inputs to
relatively high. It was noted that at
most LTER sites, just as for the NADP network, only
rates of inorganic N wet deposition are measured accurately. Yet, dry deposition and organic N wet
deposition may provide significant inputs of N at many forests and may change
more rapidly than inorganic N wet deposition.
Without accurate measures of total N deposition, changes in N inputs to
sites, and the ecological effects of changes in N inputs, cannot be detected.
b) The
effects of elevated N deposition & retention on ecosystem function,
especially C sequestration, and the mechanisms of such sequestration (e.g.,
retention in biomass increment, increased stabilization of recalcitrant SOM, reduced soil heterotrophic respiration, altered NEE).
Both the workshop participants and the breakout group
considered the Integrated Forest Study (IFS - Johnson and Lindberg, 1992) to be
a model and a starting point for the LTER/ILTER cross-site study that we are
considering. In the IFS, scientists
measured and/or modeled deposition of different N species, using technology
that was available at the time, and attempted to balance the N budget at each
site. The IFS was not designed as a
long-term study, nor were measurements of N deposition integrated into studies
of soil retention, litter decomposition, and C sequestration. However, the IFS was very effective in
demonstrating that different ecosystems in different regions of the country
experience different forms of N deposition, with some inputs, such as cloud
deposition, contributing greatly to Ndep
at some forests and not at all significantly at others.
The breakout group felt that this N&C
Relations project is integral to the goals of Research Initiative #1, Altered
Water and Nutrient Cycles. Participants
in this group agreed that a central goal should be to develop a network of
sites with accurate N input measurements and coordinated N fertilization
experiments across a gradient of Ndep,
with a suite of coordinated measurements of C-sequestration.
While there was not enough time at the workshop or the Research
Initiativ breakout group to develop a list of
measures of C sequestration, clearly NEP,
decomposition, soil respiration, trace gas emissions, DOC and DON leaching,
soil enzyme activity are critical measures that will help determine both
amounts and mechanisms of altered C sequestration in ecosystems. More detailed analyses, such as 13C-NMR
measures of changes in soil organic matter chemistry, should be decided upon at
a preliminary workshop of interested participants.
At this time, scientists at eight of the nine LTER/ILTERs have offered to be site contacts for this proposed
project: BES (John Hom,
also USFS), CDR (Jean Knops), CWT (Brian Kloeppel), HFR (Alison Magill and Bill Munger),SIK ILTER (Janos Toth), Bialowieza (Maciej Zalewski), and, of course, AND (Kate Lajtha)
and NWT (Herman Sievering).
From the viewpoint of the Data, Experiments, and Models
outline of requested response at the breakout group, only very little time was
available to address this outline.
However, the proposed project would clearly have these aspects to
consider:
DATA - data gap filling is crucial in order to have a common
starting point for any planned experiments;
EXPERIMENTS - the only way to undertake the breadth and
depth required of a study that will uncover the linkages of the C cycle and N
cycle at forested ecosystems is through cross-site collaboration within an
already existing network of ecosystem science study sites; i.e., the LTER/ILTER
network of sites;
MODELS - it was felt that the synthesis of our considered experimental
data products is absolutely essential to the parameterization process that is
at the heart of model development and testing.
Models that were mentioned as, possibly, having existing useful
structures include Biome-BGC, Century, and PnET.
Fig. 1. Annual
NEE vs. annual Ndep across 24 high
(>300 gC/m2-yr) C- sequestration forested sites (13 European, 11