Biogeosciences Aeolian nutrient fluxes following wildfire in sagebrush steppe : implications for soil carbon storage

Pulses of aeolian transport following fire can profoundly affect the biogeochemical cycling of nutrients in semi-arid and arid ecosystems. Our objective was to determine horizontal nutrient fluxes occurring in the saltation zone during an episodic pulse of aeolian transport that occurred following a wildfire in a semi-arid sagebrush steppe ecosystem in southern Idaho, USA. We also examined how temporal trends in nutrient fluxes were affected by changes in particle sizes of eroded mass as well as nutrient concentrations associated with different particle size classes. In the burned area, total carbon (C) and nitrogen (N) fluxes were as high as 235 g C m−1 d−1 and 19 g N m−1 d−1 during the first few months following fire, whereas C and N fluxes were negligible in an adjacent unburned area throughout the study. Temporal variation in C and N fluxes following fire was largely attributable to the redistribution of saltation-sized particles. Total N and organic C concentrations in the soil surface were significantly lower in the burned relative to the unburned area one year after fire. Our results show how an episodic pulse of aeolian transport following fire can affect the spatial distribution of soil C and N, which, in turn, can have important implications for soil C storage. These findings demonstrate how an ecological disturbance can exacerbate a geomorphic process and highlight the need for further research to better understand the role aeolian transport plays in the biogeochemical cycling of C and N in recently burned landscapes. Correspondence to: N. J. Hasselquist (niles.hasselquist@slu.se)

One of the direct consequences of aeolian transport is the redistribution and potential loss of soil and associated soil nutrients on eroded surfaces (Trimble and Crossan, 2000;Reynolds et al., 2001;Okin et al., 2004;Li et al., 2007Li et al., , 2008;;Ravi et al., 2009).The mass of nutrients moved in aeolian sediments is a function of transport type (i.e., suspension, saltation or creep), amount and nutrient concentration of transported sediment.Wind-driven surface creep and saltating particles dominate at small scales (i.e., several meters) and are largely responsible for local redistribution of sediment within a landscape (Stout and Zobeck, 1996).Suspension-sized particles, on the other hand, can be transported over long distances (i.e., kilometers) and can be moved at regional and even global scales (Goudie and Middleton, 2006).Nutrient concentrations tend to vary among different sizes of sediment (Zobeck et al., 1989;Li et al., 2009) and particle size distribution of aeolian sediment varies with time in areas undergoing wind erosion (Zobeck et al., 1989;Gillette and Chen, 2001).The disturbance that triggers erosion can itself affect the biogeochemical composition of moved sediment, especially in the case of fire and the direct effects of combustion (Ravi et al., 2006(Ravi et al., , 2009;;Rau et al., 2008).Thus, a mechanistic understanding of horizontal nutrient fluxes associated with aeolian transport on recently burned landscapes require an assessment of how particle size distribution, as well as nutrient concentration, of the eroded sediment varies in time after fire.
This study was part of a larger study examining aeolian sediment transport following wildfires in a semi-arid sagebrush steppe ecosystem on the eastern Snake River Plain (ESRP), Idaho (Sankey et al., 2009a(Sankey et al., , b, 2010)).During the first year following fire, Sankey et al. (2009a) reported a 28 to 67-fold increase in aeolian transport in a burned relative to an adjacent unburned area.Temporal and spatial variability in aeolian transport on the ESRP was largely attributable to climate and soil moisture until vegetation recovery in the first post-fire growing season (Sankey et al., 2009a, b) as well as differences in surface roughness (Sankey et al., 2010).Although Sankey et al. (2009aSankey et al. ( , b, 2010) ) examined the occurrence and causes of aeolian transport following fire on the ESRP, the direct consequences of aeolian transport on the redistribution of soil nutrients are still not well understood in this ecosystem.
The main objective of this study was to determine the horizontal fluxes of carbon (C) and nitrogen (N) occurring in the saltation zone (i.e., 1-2 m above the soil surface) as a result of aeolian transport following a wildfire in a semi-arid sagebrush steppe ecosystem.In a burned and an adjacent unburned area, we compared (1) the amount of aeolian transport, (2) the percent of C and N in the transported sediment, and (3) how relative changes in the particle size distribution and nutrient content of the aeolian sediment contributed to differences in the horizontal fluxes of C and N. We hypothesized greater C and N fluxes would occur in the burned relative to the unburned area as a result of increased aeolian transport, despite the potential reduction in C and N concentrations in the soil surface caused by the wildfire.We also discuss the direct and indirect affects that increase horizontal nutrient fluxes have on soil C storage in recently burned sagebrush steppe.

Study site
This study was conducted within a 16 km 2 area (Lat.43 • 30 N, Lon.112 • 38 W, 1650 m elevation) near the southern border of the Department of Energys' Idaho National Laboratory (INL) on the eastern Snake River Plain (ESRP; Fig. 1).The area has a mean precipitation of approximately 216 mm annually.Mean annual temperature ranges between 7 • C-13 • C, based on climate data from 1950-2005(NOAA INL Weather Center, 2008).The area we sampled is relatively flat with prevailing winds from the southwest.

Experimental design and sediment sampling
In July 2007, the Twin Buttes fire, a naturally caused wildfire, burned nearly 3800 ha of semi-arid sagebrush steppe on the INL.The Twin Buttes fire produced a severely burned landscape with no vegetation cover until the following spring.Maximum basal cover in the year following the fire was about 5 % and was comprised mainly of grasses and forbs (Sankey et al., 2010).The burned and unburned sites were topographically and geomorphically similar (Sankey et al., 2009a), and therefore differences in aeolian transport between these two areas were largely attributable to the fire.We selected five locations in the burned area and five locations in an adjacent unburned area (Fig. 1).At each location, we installed a tower with five omnidirectional passive sediment collectors [Big Springs Number Eight (BSNE ® ), Custom Products, Big Springs, TX, USA] mounted at 0.05, 0.10, 0.20, 0.55 and 1.00 m heights above ground immediately following containment of the fire.Collections were made approximately monthly from October 2007 to October 2008 in the burned area and less frequently (approximately once every two months) in the unburned area because of limited aeolian activity.Samples collected from the burned and unburned area during the same month in October 2007(Fall 2007), April 2008(Spring 2008), August 2008(Summer 2008) and October 2008(Fall 2008) were analyzed and presented here.
Sediment was collected at each height, placed in paper bags, dried at 105 • C for 48 h and weighed to a precision of 0.1 mg.Sediment was then passed through a series of sieves to separate each sample into particle size classes of: <106, 106-500 and >500 µm.These size classes correspond to different aeolian transport processes, where particles <106 µm in diameter tend to be transported via the suspension process (Lyles, 1988) and particles 106-500 µm in diameter are more likely transported via the saltation process (Fryrear et al., 1991), though we acknowledge these threshold sizes are approximate and can vary among studies and soil types.For this study we defined aeolian sediment as all material collected in BSNE collectors, which included both mineral particles as well as low-density particles of organic material.

Sediment transport calculations
The amount of sediment in each size class was determined for the five different heights (0.05, 0.10, 0.20, 0.55 and 1.0 m) in order to calculate rates of sediment transport (sediment flux and sediment discharge).Rates of sediment transport were calculated as described in Van Donk et al. (2003).Briefly, aeolian sediment flux (kg m −2 d −1 ) for each tower was first calculated for each sediment sampling interval and height as: where mass is kg of sediment, area is the size of the opening on the BSNE© sampler (0.0002 m 2 at heights 0.05 and 0.10 m, and 0.001 m 2 at 0.20, 0.55 and 1.0 m), and time is the sampling interval in days.Sediment flux was then plotted as a function of sample height for each tower and sampling interval.A power model was fitted to each plot and integrated over 2.0 m to calculate sediment discharge for each tower (n = 5 per site) and sampling interval with units of kg m We omitted sediment discharge estimates for which the r 2 value of the power model was less than 0.80 from further statistical analyses, so sample size in the burned and unburned area ranged between three and five height-integrated flux measurements for each time period.It is important to note that sediment discharge is calculated from the integration across a vertical profile, and therefore represents the horizontal flux of sediment with units of kg m −1 d −1 .

Nutrient analyses
For each sampling date and location, a subsample of aeolian sediment (0.5 g) from each particle size class was taken for total C and N analyses.Samples for nutrient analyses were ground with a Wig-L Bug grinding mill (International Crystal Lab, Garfield, New Jersey, USA) to a fine powder.Samples were analyzed for total C and N using an elemental analyzer (Model ESC 4010, Costech, Valencia, CA, USA).Horizontal fluxes of C and N were calculated by multiplying nutrient concentrations (g C or N kg −1 sediment) by sediment discharge rates (kg sediment m −1 d −1 ) for each size class and sampling date, which resulted in estimates of horizontal nutrient fluxes in units of g C or N m −1 d −1 .In early November 2008, soil samples were collected from the soil surface for total C and N analyses.Within both the burned and unburned area, 20 random soil samples were collected from the top 1 cm of the soil profile.Total C and N concentrations were determined as described above for aeolian sediment.Inorganic C was previously determined for a set of surface soils in both the burned and unburned area (Hoover, 2010).Using these estimates of inorganic C, organic C of surface soils was calculated as the difference between total C and inorganic C. Inorganic C could not be determined in the aeolian sediment because the amount of sediment collected was often too small to perform both total C and inorganic C analyses.

Statistical analysis
To evaluate the effect of fire on sediment discharge and horizontal nutrient fluxes in the year following the fire we used a two-way ANOVA with area (burned or unburned) and time (Fall 2007, Spring 2008, Summer 2008, and Fall 2008) as fixed factors.To detect differences among size classes, nutrient concentrations, and discharge of sediment we used a three-way ANOVA with area, time, and particle size class as fixed factors.We followed these analyses with post hoc tests using Bonferroni corrections to determine significant differences.Sediment discharge and nutrient data were transformed prior to being subjected to ANOVA (though values presented are mean ± SE of non-transformed data).All statistical analyses were performed using SPSS statistical software (SPSS Inc., v 16.0, 2007).

Total sediment discharge
Total sediment discharge was nearly three orders of magnitude greater in the burned relative to unburned area in Fall 2007, but did not differ thereafter as sediment discharge in the burned area significantly decreased to levels comparable to the unburned area (Fig. 2a).In the unburned area, sediment discharge was negligible and did not differ among sampling periods (Fig. 2a).Average sediment flux rates (kg m −2 d −1 ) at the different heights and total sediment discharge (kg m −1 d −1 ) for the burned and unburned area are shown for the different sampling periods in Appendix A.
Sediment discharge of the suspension-and saltation-sized particles was significantly greater in the burned relative to the unburned area in Fall 2007, but not thereafter as discharge in the burned area significantly decreased during the study (Fig. 2b, c).Sediment discharge of the largest size class did not differ between the burned and unburned area during the study (Fig. 2d).Average flux rates (kg m −2 d −1 ) at the different heights and total sediment discharge (kg m −1 d −1 ) for the different particle size classes in the burned and unburned area are shown for the different sampling periods in Appendix B and C, respectively.

Nutrient concentrations
Overall, total C concentrations of aeolian sediment differed significantly among particle size classes, with the highest concentrations in the largest (>500 µm) size class (22.52 ± 1.05 %), intermediate concentrations in the saltation (106-500 µm) size class (10.46 ± 0.68 %), and the lowest concentrations in the suspension (<106 µm) size class (2.30 ± 0.17 %).There were no differences in C concentrations between the burned and unburned sites within any particle size class.
Total N concentrations of aeolian sediment essentially paralleled that of C. N concentrations of aeolian sediment differed significantly among particle size classes, with the highest N concentrations in the largest size class (1.50 ± 0.11 %), intermediate concentrations in the saltation size class (0.79± 0.06 %), and the lowest concentrations in the suspension size class (0.18 ± 0.01 %).There were no differences in N concentrations between the burned and unburned area within any particle size classes.
At the end of our study period, total N and organic C concentrations in the soil surface (1 cm depth) was significantly lower in the burned (0.21 ± 0.01 % and 2.03 ± 0.12 %; for total N and organic C, respectively) relative to the unburned area (0.26 ± 0.02 % and 2.75 ± 0.31 %).

Horizontal nutrient fluxes
In general, horizontal fluxes of C were nearly three orders of magnitude greater in the burned relative to the unburned area (79.62 ± 35.69 g C m −1 d −1 and 0.13 ± 0.03 g C m −1 d −1 , respectively).Horizontal C fluxes differed significantly between the burned and unburned area in Fall 2007, but not thereafter as horizontal C fluxes in the burned area significantly decreased to rates comparable to the unburned area (Fig. 3a).In contrast, horizontal C fluxes in the unburned area were negligible and did not differ among sampling periods.Overall, horizontal flux of N also differed between the burned and unburned area (7.82 ± 3.38 g N m −1 d −1 and 0.01±0.00g N m −1 d −1 , respectively).As observed for horizontal C fluxes, significant differences in horizontal fluxes of N between the burned and unburned area were only detected in Fall 2007, but not thereafter as horizontal N fluxes in the burned area significantly decreased during the study (Fig. 3b).Horizontal N fluxes in the unburned area were negligible and did not differ among sampling periods.

Horizontal nutrient fluxes among particle size classes
In general, mean horizontal C flux for the saltation size class (29.38 ± 14.45 g C m −1 d −1 ) was significantly greater than both the suspension and largest size classes (4.54 ± 1.78 g C m −1 d −1 and 7.89 ± 4.79 g C m −1 d −1 , respectively).Horizontal C fluxes in the saltation size class differed significantly between the burned and unburned area in Fall 2007, but not thereafter as horizontal C fluxes in the burned area significantly decreased to rates comparable to the unburned area (Fig. 4b).Horizontal C fluxes in the suspension and largest size classes did not differ between the burned and unburned area during the study (Fig. 4a, c).
Horizontal N fluxes essentially paralleled horizontal C fluxes.Mean horizontal N flux for the saltation size class (2.40 ± 1.19 g N m −1 d −1 ) was significantly greater than both the suspension and largest size classes (0.44 ± 0.17 g N m −1 d −1 and 0.50±0.30g N m −1 d −1 , respectively).Horizontal N fluxes in the saltation size class differed significantly between the burned and unburned area in Fall 2007, but not thereafter as horizontal N fluxes in the burned area significantly decreased during the study (Fig. 4e).Horizontal N fluxes in the suspension and largest size classes did not differ between the burned and unburned area during the study (Fig. 4d, f).

Discussion
Previous studies have examined the effect of fire on aeolian transport (Zobeck et al., 1989;Wiggs et al., 1994;Whicker et al., 2002;Vermeire et al., 2005;Ravi et al., 2006Ravi et al., , 2007b;;Sankey et al., 2009aSankey et al., , b, 2010), yet the majority of these studies have focused on factors that contribute to increase rates of aeolian transport, such as changes in vegetation cover and surface roughness (Zobeck et al., 1989;Wiggs et al., 1994  It is important to note that differences in aeolian transport between the burned and unburned area were only detected in the first few months following the fire.This pulse of aeolian transport following the Twin Buttes fire was comparable to post-fire aeolian transport reported in Africa, Australia, and elsewhere in the USA (Wasson and Nanninga, 1986;Zobeck et al., 1989;Wiggs et al., 1994;Whicker et al., 2002;Vermeire et al., 2005), except that background aeolian transport was relatively lower in undisturbed ESRP and the post-fire pulse of aeolian transport was relatively intense and only occurred over a few months.
In addition to increased rates of aeolian transport, we observed significantly higher horizontal nutrient fluxes in the burned area during the first few months following the fire.Previous studies have shown that suspension-sized particles generally contain disproportionately greater amounts of nutrients, and consequently the loss of these particles due to aeolian transport may be an important mechanism leading to nutrient depletion in natural landscapes (Larney et al., 1998;Neff et al., 2005;Li et al., 2009;Harper et al., 2010).However, the majority of studies we are aware of have focused on landscapes with a long history of wind erosion.Unburned sagebrush steppe on the ESRP, on the other hand, has experienced little aeolian losses in recent history and is often characterized by the presence of fine-textured surface soils and the lack of contemporary (i.e., recent decades-centuries) geomorphic evidence of aeolian transport.The impacts of aeolian transport on soil texture and nutrient redistribution could therefore differ substantially at our site compared to landscapes with more frequent episodes of contemporary wind erosion.Interestingly, we observed higher C and N concentrations for the coarser, saltation-sized particles relative to the finer, suspension-sized particles.The higher C and N concentrations for saltation-sized particles at least partly reflects a greater proportion of organic matter in this size class (19.7 ± 1.3 %; determine through mass loss upon combustion) relative to suspension-sized particles (4.0 ± 0.2 %) that consisted mostly of mineral particles with less organic matter (M.Germino, unpublished data).In addition, soil microaggregates (53-250 µm) can contain higher concentrations of nutrients than clay and silt particles (<53 µm; Wick et al., 2008), which could also help explain higher C and N concentrations for the saltation-sized particles.
Such a large pulse of nutrient-rich aeolian sediment following fire appeared to have dramatic effects on near surface soil nutrient pools in this ecosystem.For instance, total N concentrations from the top 1 cm of the soil profile were 19 % lower in the burned relative to the unburned area (0.21 % and 0.26 %, respectively), whereas soil organic C concentrations was 26 % higher in the unburned relative to the burned area (2.75 % and 2.03 %, respectively).The losses reported here are comparable to what Li et al. (2007) found at the Jornada Experimental Range in southern New Mexico, where organic C and N decreased 25 % in the top 5 cm of the soil due to aeolian fluxes.We acknowledge that lower total N and organic C concentrations in the soil surface in the burned area may have, in part, been the result of volatilization during combustion as suggested by slightly lower percent C and N in aeolian sediment immediately after the fire from the burned (7.52 ± 2.19 % and 0.53 ± 0.13%; for C and N respectively) compared to the unburned area (10.93 ± 3.38 % and 0.87 ± 0.32 %).However, given the large differences in horizontal nutrient fluxes between the burned and unburned area, it appears that aeolian transport plays an important role in the redistribution of soil nutrients in recently burned sagebrush steppe.
Horizontal nutrient fluxes, by themselves, do not directly quantify the amount of nutrients lost per unit ground area.We therefore performed a simple calculation to provide an estimate of the amount of C and N lost from the burned area as a result of aeolian transport.Using soil erosion bridges, Sankey et al. (2010) reported a mean rate of surface deflation of 2.1 mm y −1 in the same burned site, which was equal to the loss of 2.63 × 10 4 kg soil ha −1 y −1 .Although there was some indication of local redistribution of sediment within the burned area, the burned area as a whole exhibited net deflation and thus soil loss.By multiplying the amount of sedi-ment loss by the average percent C and N of aeolian sediment from the burned area, we estimate that 1.51 × 10 3 kg C ha −1 and 123 kg N ha −1 might have been lost from the burned area during the first year following fire.For comparison, at a nearby site on the ESRP that had not burned recently, Mc-Gonigle et al. (2005) reported an annual increase in surface soil organic C of 0.5 g kg −1 through litter input, which corresponds to roughly 81 kg C ha −1 y −1 .Thus, the large pulse of aeolian transport during the first year following fire appeared to have caused a rapid loss of soil C, which is roughly equivalent to 19 years of soil C accumulation via litter input in this semi-arid ecosystem.This finding is consistent with other studies that have indentified aeolian transport as an important mechanism by which soil surface can become depleted in C (Delany and Zenchelsky, 1976;Harper et al., 2010).
In addition to the immediate loss of soil C as a result of increase aeolian transport, soil C may be further reduced in the burned area by a reduction in post-fire vegetation recovery.As mentioned above, we estimated a loss of 123 kg N ha −1 from the burned site during the first year following fire.Because soil nutrients are concentrated in the upper soil layers in this ecosystem (McGonigle et al., 2005), losses of N from the soil surface could drastically affect plant productivity and/or lead to changes in species diversity.In the burned area, maximum basal cover in the year following the fire was roughly 5 % and was comprised mainly of grasses and forbs.In contrast, the unburned area had 18 % basal cover and contained an even mix of shrub and herbaceous species (Sankey et al., 2010).This change in species composition and reduction in plant biomass following the fire could, in turn, further reduce soil C in the long-term through a reduction in litter input and root production.Lal (2003) showed how increased rates of aeolian transport can affect soil C storage through the depletion of soil fertility and subsequent decrease in crop production.More detailed studies are needed to better understand the feedbacks between horizontal nutrient fluxes, postfire vegetation recovery and soil C storage in recently burned landscapes.
Although water is a major contributor to soil erosion globally (Middleton and Thomas, 1997), we focused on wind erosion because erosion by wind is the dominant transport process in arid and semi-arid environments (Ravi et al., 2007b;Field et al., 2011).Furthermore, Wilcox et al. (2011) recently showed that runoff and water erosion in sagebrush steppe ecosystems require much steeper slopes than the relatively flat sites we evaluated, and rainfall intensity did not appear conducive to generate the type of runoff needed for water erosion (see Sankey et al., 2009b).We also acknowledge that aeolian transport and associated horizontal nutrient fluxes depend on several other factors, such as fire intensity and duration, post-fire soil moisture conditions as well as chemical and physical controls on post-fire erodibility (as discussed in Sankey et al., 2009b).
Nutrient losses due to aeolian transport are often attributable to the redistribution of suspension-sized particles www.biogeosciences.net/8/3649/2011/Biogeosciences, 8, 3649-3659, 2011  5 cm 0.97 ± 0.54 0.14 ± 0.04 7.4 × 10 −3 ± 2.7 × 10 −3 8 × 10 −3 ± 2.5 × 10 −3 10 cm 0.44 ± 0.09 0.03 ± 0.02 1.9 × 10 −3 ± 7 × 10 −4 7 × 10 −3 ± 3.7 × 10 −3 20 cm 0.18 ± 0.12 0.02 0.18 ± 0.12 0.17 ± 0.13 2.0 × 10 −3 ± 4.0 × 10 −4 2.0 × 10 −3 ± 4.0 × 10 −4 (Larney et al., 1998;Neff et al., 2005;Li et al., 2009;Harper et al., 2010).Vertical fluxes of particles smaller than 10 µm (PM 10 , kg m−2 of ground area) are typically four orders of magnitude less than the associated horizontal sediment fluxes based on ratios of vertical/horizontal fluxes of 0.0001 in Gillette et al. (1997) and 0.0002 for another fire and erosion study using the same BSNE sediment collec-tors on the ESRP (Wagenbrenner et al., 2011).Our study is unique in that it clearly shows how the redistribution of saltation-sized sediment can contribute appreciably to nutrient losses from a recently burned area.Although sediment discharge for suspension-and saltation-sized particles were similar in the burned area, saltation-sized particles contained significantly higher percent C and N. Greater C and N in  saltation-sized particles likely reflects the greater amount of organic matter observed in saltation-sized sediment relative to suspension-sized sediment.Organic matter is characterized by lower particle-density compared to mineral soil, and thus can be more buoyant and potentially capable of transport over longer distances.Additionally, because the burned area was nearly completely void of surface roughness elements during the first few months following the fire (Sankey et al., 2010), there was sufficient amount of time for iterative movement of particles via saltation off of the burned area.This notion of nutrient-rich saltation-sized particles being lost from the burned area is further supported by a reduction in the relative proportion of saltation-sized particles in residual soils in the burned (18 %) compared to the unburned (12 %) area one year after the fire (M.Germino, unpublished data).Understanding the relative importance of organic matter in aeolian transport processes may be crucial for sites that normally do not have much or any wind erosion, but then are suddenly subjected to aeolian processes following a disturbance, such as fire." In conclusion, our results show how an ecological disturbance can exacerbate a geomorphic process, resulting in increased horizontal nutrient fluxes in a semi-arid sagebrush steppe ecosystem.In addition to causing a rapid loss of organic C in the soil surface, increased aeolian transport of nutrient-rich sediment could negatively affect post-fire vegetation recovery, thereby making recently burned areas more prone to further wind erosion as well as reducing soil C in the long-term through a reduction in litter input.Changes in climate along with an increase in invasive annual grasses have lead to an increase in fuel loads and fuel continuity on the ESRP.Consequently, the return intervals of wildfires has drastically shortened from a return interval of 32-70 years or longer before European colonization (Baker, 2006;Keane et al., 2008;Miller and Heyerdahl, 2008) to as short as five years in some places on the ESRP (Whisenant, 1990;Keane et al., 2008).Contemporary increases in fire frequency, therefore, might increase the frequency of post-fire erosion, as well as provide insufficient time for soil C and N stocks to recover between episodes of erosion.Thus, an overarching challenge for ecologists will be to better understand the feedbacks between horizontal nutrient fluxes following fire and post-fire vegetation recovery in order to predict how ecosystem functioning of sagebrush steppe will respond to future climate change scenarios.

Fig. 1 .
Fig. 1. (A) Location of study area relative to Idaho, with inset showing location of Idaho relative to USA.(B) Aerial photograph of the eastern Snake River plain (ESRP), Idaho, showing the area burned by the Twin Buttes wildfire in 2007 (hatched area), and the location of aeolian sediment collectors in the burned (B) and unburned (U) area.

Fig. 2 .
Fig. 2. Mean (±SE) sediment discharge in the burned (filled circles) and unburned (open triangle) area during the first year following the Twin Buttes fire.Total sediment discharge (A), sediment discharge for the suspension (<106 µm) size class (B), saltation (106-500 µm) size class (C) and the largest size class of eroded sediment (D).Values for each sampling period consist of 3-5 height-integrated flux measurements (* denotes significant site differences at P < 0.05).

Fig. 3 .
Fig. 3. Mean (±SE) horizontal fluxes of carbon (A) and nitrogen (B) in the burned (filled circles) and unburned (open triangles) area.Horizontal fluxes of C and N were calculated by multiplying nutrient concentrations (g C or N kg −1 sediment) by sediment discharge rates (kg sediment m −1 d −1 ) for each size class and sampling date (* denotes significant site differences at P < 0.05).

Fig. 4 .
Fig. 4. Mean (±SE) horizontal fluxes of carbon (A)-(C) and nitrogen (D)-(F) among the different particle size classes of sediment in the burned (filled circles) and unburned (open triangles) area (* denotes significant site differences at P < 0.05).

Table A1 .
Van Donk et al. (2003)x (kg m −2 d −1 ) from BSNE collectors located at different heights (0.05, 0.10, 0.20, 0.55 and 1.0 m) in the recently burned and adjacent unburned area.Mean sediment flux was plotted as a function of sample height for each sampling date, and a power model was fitted to each plot to calculate sediment discharge (kg m −1 d −1 ) as described inVan Donk et al. (2003).

Table B1 .
Mean (±SE) sediment flux (kg m −2 d −1 ) and sediment discharge (kg m −1 d −1 ) for different size classes of sediment in the burned area.

Table C1 .
Mean (±SE)sediment flux (kg m −2 d −1 ) and sediment discharge (kg m −1 d −1 ) for different size classes of sediment in the unburned area.