C33C - 1289 Elevational Gradients and Differential Recruitment of Limber Pine (Pinus flexilis) and Bristlecone Pine (Pinus longaeva); White Mountains, California, USA C.I. Millar, R.D. Westfall, and D.L. Delany USDA Forest Service, Sierra Nevada Research Center, Albany, CA Introduction Subalpine tree species and forest communities are assumed to respond to rising temperatures by migrating upslope. This generates a conservation hypothesis that habitat will be lost as area diminishes approaching mountain summits (“elevational squeeze”). Our prior studies of Great Basin forests and historic climate lead us to hypothesize that responses are more complex, with species following individualistic and often surprising paths. We began a study to investigate the effect of ongoing climate change on subalpine limber pine (Pinus flexilis; PiFl) populations in the Great Basin. Here we present results from the White Mtns portion of that study. In this range, bristlecone pine (P. longaeva; PiLo) currently and historically dominates the highest elevations and on dolomite soils (Fig. 1), PiFl occupies the middle subalpine range on non-dolomitic soils (Fig. 2), while singleleaf pinyon pine (P. monophylla; PiMn) historically occurred disjunctly below the other conifers (Fig. 3). Fig 1. PiLo at upper treeline Fig 2. PiFl at mid elevation Current Treelines (m) Bristlecone Pine Limber Pine Pinyon Pine Upper 3400 3300 2800 Lower 3000 3000 1000 Fig 3. PiMn zone at low elevation below PiLo and PiFl Study Question What is the nature of limber pine recruitment (past 120 years) in the White Mtns relative to: -- elevation (lower treeline, middle elevations, and upper treeline) -- vegetation structure (live forest, meadow, relict forest zone) -- slope, aspect, and soils (esp dolomite vs non-dolomite substrates) -- species (limber pine, bristlecone pine, pinyon pine) Study Sites Seven study sites were chosen in the White Mtns between Sierra View and Mt. Barcroft including three at upper PiLo treeline (3341-3570m), two at middle elevations (3184-3223m), and two at lower PiFl/PiLo treeline (2946-3114m) (Fig. 4). These represent a range of slopes, aspects, elevations, vegetation communities, and substrates (Table 1). Fig 4. Study area with high (red), mid (blue), and low (yellow) elevation sample sites Table 1. Physical attributes of the study sites Methods At each site, we established 30m-wide belt transects that traversed 17-100 m elevation. The belts were divided at 30m intervals, establishing 30x30m plots along the transects. We tallied all live trees and dead wood within the plots for tree species and diameter. In addition we estimated age of live limber pines in the plots using whorl counts (trees < 2 m ht) and increment coring (trees > 2 m ht). We designated PiFl trees that were ≤120 yrs old and PiLo trees <20cm dia at 0.5 m height as recruits. We assessed PiFl recruitment age versus elevation with least squares analyses (SAS 2006). To test relationships of climate and year of recruitment, we first developed composite climate indices (see Millar et al. 2004) from four instrumental records (Independence, CA, Mina, NV for temperature; Independence, Mina, Tahoe, CA and Yosemite National Pk, CA for precipitation; periods of record were 1927-2005, 1897-2005, 1909-2005, and 1906-2005, respectively, WRCC 2006; HCN 2006). We analyzed simple linear correlations (SAS 2006) as well as nonlinear relationships. For the latter, we used a second-order least squares response-surface model of recruitment versus minimum and maximum annual temperature, respectively, annual precipitation, as well as standard indices of the Pacific Decadal Oscillation (PDO), North Atlantic Oscillation (NAO), and Atlantic Multidecadal Oscillation (AMO). We evaluated the behavior of these variables in second-order response surfaces. Results PiFl exceeded PiLo in abundance of recruits at all sites except the Relay Ridge middle elevation location (Table 2). The excess ranged from 2803700% and was most prominent at upper treeline locations (Fig. 5). PiFl recruitment was greatest on dolomite substrates at upper elevations, occurred on all aspects, and to elevations above current live PiLo or PiFl, up to 300m above current PiFl treeline. Recruitment of PiFl and PiLo occurred primarily in areas of current or relictual forest, except at middle elevations where PiFl was recruiting into sagebrush (Artemesia tridentata) meadows (Fig. 6). In the lower treeline zone, PiFl recruitment occurred in narrow west- and north-facing ravines on sandstone substrates and along transient watercourses (Fig. 7). Fig 5. PiFl recruits in relict PiLo above upper treeline Table 2. Density and age statistcs Average age of PiFl trees at high elevation sites (22.3 yrs) was lower than at mid- (35.7 yrs), and low elevation sites (33.8 yrs) (Table 2, Fig 8). Age was significantly and negatively correlated with elevation at high elevation sites (Fig. 9A) indicating that younger trees were higher in elevation above upper treeline. At middle elevations, significant correlations indicated younger trees in meadow compared to forest locations (Fig. 9B). At lower treeline, tree age was not significantly correlated with elevation in three of the four transects (Fig. 9C). Recruitment in PiFl during the past 120 years at all sites was highly episodic and clustered between 1975-1991 (Figs. 8, 10A). Recruitment was very low before 1945, increased slightly from 1945-1975, pulsed very high between 19751991, and decreased to a low level between 1991-2006. Fig 6. PiFl recruiting into sage meadows, mid elevation Fig 7. PiFl recruiting in ravines, lower treeline Significant partial correlations of PIFL recruitment levels with climate were in linear and in higher-order interaction trends: Min T AMO PDO x AMO AMO x Min T p=0.01 p=0.04 p=0.04 p=0.06 Fig. 10B illustrates the correlation of recruitment with multi-decadal periods of AMO over the past 120 years: recruitment was higher during “cold ocean” AMO periods and lower during “warm ocen” AMO periods, although asymmetry is strong during the 20th century. Fig 8. Mean plot density PiFl by age class at 7 sites Fig 9. Correlation of PiFl recruit age with elevation at A) high, B) mid, and C) low elevation sites Discussion Discussion and Summary Fig 10. A) Number of PiFl recruits established/ per year (overall dataset) B) Standardized values of mean annual temp, precip, and AMO We find that PiFl is recruiting abundantly and unexpectedly at low to high elevations in the White Mtns in diverse habitats, including cool ravines at low elevations, into sagebrush meadows at mid-elevations, and extending above upper treeline in areas of relict bristlecone pine forests. Contrary to assumptions, PiFl recruitment is greatly exceeding PiLo above current upper PiLo treeline and 300m above current PiFl treeline, and on high-elevation dolomitic soils. Recruitment in PiFl has been episodic during the past 120 years, concentrated during 1975-1991, with much lower recruitment in the past 15 years and earlier in the 20th century. Significant interactions with climate variables suggest that recruitment passed a threshold during the 20th century, whereby recruitment was triggered by warmer and wetter conditions, but that the favorable combination of conditions did not persist during the past 15 years. Recruitment correlates with multi-decadal periods of the AMO, with weak but positive recruitment in the cool early 20th century AMO phase, and strong recruitment in the cool late 20th century phase. Warm phases of AMO have been associated with drought in SW US, and in the White Mtns may relate to timing of precipitation as well as overall amount. Warming temperatures during the second half of the 20th century may have increased the likelihood of precipitation thresholds being crossed at these high elevations such that more abundant recruitment occurred between 1975-1991. PiFl appears to be advancing in dominance over PiLo throughout their elevation ranges in the White Mountains, finding favorable conditions above upper treeline (warm, wet), mid elevation meadows (release from inversion?), and in cool damp ravines at low elevations. References Millar, Westfall, King, Graumlich, Delany. 2004. Arctic, Antarctic, and Alpine Research. 36(2):181-200 WRCC 2006. Western Regional Climate Center. Historical Climate Archives. http://www.wrcc.dri.edu/ HCN 2006. U.S. Historical Climate Network, NOAA. http://www.ncdc.noaa.gov/oa/climate/research/ushcn/ushcn.html SAS 2006. SAS Institute Inc. JMP Statistics and Graphics Guide, version 6.03.