422 



Human Infhu-nccs— Our IJvitiii Rtwource.s 



Trends in Input 



Between 1981 and 1989. 29 precipitation 

 events were highly acidic (pH values less than 

 4.5), with 22 of these events occuning in low- 

 volume summer storm events from May 

 through September. Since 1989 only one storm. 

 in November 1991. had a pH below 4.5. A rise 

 in mean precipitation pH was also recorded dur- 

 ing this period at both the low-elevation and 

 mid-elevation sites. The Elk Creek site mean 

 precipitation pH rose from 5.23 before 1989 to 

 5.68, while the Log Creek site rose from 5.12 to 

 5.39. A beneficial downward trend in the total 

 annual loading of sulfur occurred (Table). 

 which was poorly explained by variation in pre- 

 cipitation (;• = 0.397). The frequency of storm 

 events with very high (> 2 mg L"' ) concentra- 

 tions of sulfur was also reduced from 25 events 

 from 1983 to 1988 to 6 events in the following 

 5 years. No apparent trend for nitrogen was 

 seen. Emerald Lake precipitation chemistry 

 data are insufficient to infer trends. 



lowing the fire, increasing from a pre-bum 

 mean of 0.35 kg ha"' yr'' to over 6.63 kg ha ' 

 yr" ' . Acid-neutralizing capacity also rose from 

 a pre-bum mean of 2 1 . 1 6 kg ha" ' yr" ' to over 

 37.16 kg ha"' yr"'. 



At the Elk Creek site, mean alkalinity (or 

 ANC) expressed as HCO3 was 310.0 pEq L"' 

 with a mean stream pH of 6.61 (neutral). The 

 creek flows intermittently, mostly between 

 January and March. The water is cloudy with 

 suspended clay particles, and debris Hows are 

 common after heavy rains. 



The outflow of Emerald Lake had a mean 

 pH of 6. 17 (neutral) and mean HCO^, of only 

 30 pEq L" ' . Episodic acidification on the order 

 of days to weeks was recorded at Emerald Lake 

 under two scenarios: ( 1 ) during dirty summer 

 storms, when buffering capacity was over- 

 whelmed by low-pH stomiwater flashing into 

 the lake, and (2) durina snowmelt, when NH ,*, 

 NO^. and SO^- were preterentially eluted 

 from the snowpack, causing an acidic pulse 

 (Williams and Melack 1991 )? 



Effects on Stream and Lake 

 Chemistry 



Stream discharge peaks in February at the 

 Elk Creek site, April at Log Creek, and June at 

 Emerald Lake. The first is a direct response to 

 maximum rainfall; the other two discharges 

 reflect the peaks of snowmelt. 



In the mixed-conifer Log Creek site, more 

 than 99% of the nitrogen deposited was con- 

 served. Mean discharge of nitrogen, expressed 

 as NO3, is 0.09 kg ha"' yr"', virtually all of 

 which was derived from the melting snowpack. 

 Seventy-four percent of the annual loading of 

 sulfur was conserved, with a mean discharge, 

 expres.sed as SO_j- , of 0.93 kg ha" ' yr" ' . again 

 mostly derived from melting snowpack. Acid- 

 neutralizing capacity (ANC), expressed as 

 HCO:, (bicarbonate), was many times greater 

 than annual acidic loading at a mean of 320 pEq 

 L" ' . At these levels, most sulfur and nitrogen 

 are being conserved by the biota in the ecosys- 

 tem, and the ecosystem's ability to neutralize 

 acid is generally good except when extreme 

 events occur. 



Tharp's Creek, one of the paired watersheds 

 in the mid-elevation Log Creek site, was burned 

 by prescription in October of 1990, producing 

 striking changes in stream output chemistry that 

 continued through 1993. Although net retention 

 of NO3 continued, discharge of NO, increased 

 from a pre-bum mean of 0.04 kg ha" ' yr" ' to 

 1.59 kg ha"' yr"' in the 3 years following the 

 bum. Thus, the ability to retain nitrogen was 

 decreased, and the system leaked nitrogen and 

 sulfur. The SO^- outputs exceeded inputs fol- 



Discussion 



We have found no long-term chronic acidifi- 

 cation of lakes and streams in our study area, 

 even though the hundreds of lakes in the region 

 are considered to be the most poorly buffered in 

 the western United States (Landers et al. 1987). 



Emerald Lake, typical of subalpine lakes in 

 the region. cuiTcntly generates enough ANC, 

 mostly through cation exchange (Williams et al. 

 1993a), to buffer two to five times the current 

 annual acidic inputs (Sickman and Melack 

 1989). Very clean winter air and the large-vol- 

 ume, extremely dilute snowpack it produced 

 were significant factors in maintaining the 

 buffering capacity. Dirty summer storms and 

 spring snowmelt that concentrate NO, and 

 SOj- and deliver them quickly to the lake have 

 caused episodic acidification, a phenomenon 

 that we are studying. An increase in the fre- 

 quency of storm events during the summer and 

 fall would likely be hamiful to subalpine lake 

 basins if the chemistry of these events were to 

 remain the same. 



The low concentrations of nitrogen and sul- 

 fur in stream water indicate that neither reached 

 saturation in soils and plants of the mixed- 

 conifer forest (Williams et al. 1993b). The 

 buffering capacities of low- and mid-elevation 

 sites were many times greater than acidic 

 inputs. But human-caused nitrogen and sulfur 

 have been shown to stimulate the growth of 

 ponderosa pines (Pimis ponderosa). which in 

 turn increases vulnerability to high local ozone 

 levels and potential for damages (Temple et al. 

 1992). 



