Pattern formation in ecosystems via self-organization is an important area of investigation in the field of ecology. Self-organization is the process whereby short-range facilitation and long-range inhibition lead to patterns in ecosystems at varying scales. Can biotic agents, such as key ecosystem engineers, be responsible for patterns of self-organization? We sought to investigate this question on the tundra of Pikes Peak outside Colorado Springs, CO. Aerial images of Pikes Peak reveal distinct patches of alpine avens (Geum rossii) dotting the tundra. Are there any patterns in the distribution and characteristics of these alpine avens patches? Closer examination reveals that evidence of northern pocket gopher (Thomomys talpoides) soil disturbance also speckles the tundra. Are there any links between gopher disturbance and alpine avens patches? We sought to answer these broad questions through a series of three investigations. We examined the large-scale spatial distribution of the avens patches relative to each other, surface gopher disturbance in relation to individual avens patches, and the underground characteristics of the tundra below the patches. Our findings indicate that while a link can be established between gopher disturbance and avens patches, it is not the complete picture. We found that contrary to the expectation that avens patches would follow a regular distribution at smaller distances, they were in fact randomly distributed at small distances and clumped at greater distances as shown by Ripley’s K tests. In line with our hypotheses, we found that gopher disturbance was clumped, and occurred more often within avens patches than would be expected given disturbance frequency across the tundra, p=0.001 (chi-squared=10.88, DF=1) for quadrant one, and p=0.0001 (chi-squared=306.96, DF=1) for quadrant two. Finally, we discovered an interesting pattern of what appears to be disintegrated bedrock beneath the avens patches, which may have implications for avens patch resilience on the tundra. In t-tests comparing mean resistivity of soil underground inside and outside the patch, p<0.05 for all depths except the lowest depth in one patch. In sum, it appears from our findings that while gopher disturbance may be necessary for avens patches on the Pikes Peak tundra, it is not sufficient. This is given the fact that gopher disturbance occurs in areas where avens patches do not, and avens patch boundaries are crisply defined while gopher disturbance is diffuse. Evidence does seem to point to self-organization on the tundra, with gopher disturbance creating short-range facilitation for alpine avens, and some mechanism of long-range inhibition preventing avens patches from occurring everywhere on the tundra.
The tree line is a climatic boundary, however its ability to respond to changing climate seems to be constrained by the spatial distribution of trees at the leading edge; compared to abrupt or krummholz tree lines, diffuse tree lines are moving upslope much more readily in response to recent anthropogenic warming. Here we report on the micrometeorological processes that result from the diffuse leading edge of a moving tree line on Pikes Peak, Colorado, USA, and on the impacts these processes have on tree temperatures. We focus on the layering and movement of air in the lower 10m of the atmosphere including the height of the displacement of the zero velocity plane. Our experimental design consisted of 300m upslope transects through the tree line into the alpine tundra where we measured: (1) height of the zero plane displacement using handheld anemometers, (2) temperature of 10cm tall seedlings, 3-5m tall trees, and tundra grasses using an IR camera, (3) temperature and relative humidity at 2.5cm an 2m using Kestrel hand held weather stations, (4) the vertical atmospheric profiles using 10m towers equipped with 8 anemometers at 5 different elevations, (5) vertical movement of air using a bubble-blowing machine. Our results show that (1) the zero plane height decreased exponentially with increasing elevation (R2=0.432, N=57, p<0.0005) from approximately 25cm within the tree line to 2.5cm in the tundra above. The spatial variability of the zero plane height also decreased with elevation. (2) The temperature of small seedlings was (3) closely coupled to the ground vegetation (paired t-test t= 2.213, df=10, p=0.051),but seedlings were on average 3.88°C warmer than trees (paired t-test t= 5.808, df=10, p<0.0005), and trees were 6.1°C colder that the tundra (paired t-test t= 6.617, df=10, p<0.0005). (3) Compared to the air at 2m, the air layer at 2cm had higher temperature (+2.5°C, paired t-test t= 7.205, df=19, p<0.0005), and higher relative humidity higher (+29%, paired t-test t= 9.657, df=19, p<0.0005). (4) The vertical wind profile had a simple and smooth slow down to the zero plane at 2.5cm in the alpine tundra. However the profile was complex in all locations where trees were present: It showed an initial slow down to a very low speed at 3-4m, increase in velocity at 2m, and final slow down to the zero plane at 25cm. Qualitative and quantitative analysis of bubble movement (5) showed that the upper boundary layer was turbulent.
Spatially-organized patches primarily composed of Alpine avens (Geum rossii) on Pikes Peak, CO give the tundra of the 14,000 ft. mountain a freckled appearance. The mechanisms causing formation and maintenance of these patches were examined using parameters such as vegetation height, species abundance, micro-topography, C:N ratios of soils and plants, and soil moisture. This study focused on eight patches by evaluating the above parameters along eight, horizontal 15-18 m transects that ran through the centers of the patches. Shockingly, vegetation height was two times greater within the patch compared to open tundra. This suggests nutrient accumulation within the patch parameters. In this thesis we analyze abiotic, top-down and bottom-up processes, to evaluate these patches. We conclude that this ecosystem is a bi-stable dynamical structure (Lotka-Volterra). In addition, scale-dependent feedback mechanisms (short-distance facilitation and long-distance inhibition) may be a primary contributor to patch formation and maintenance.