Floristic dynamics across a semi-arid volcanic chronosequence in Northern Arizona.
This study examined the role of edaphic factors on both plant community composition and on the spatial distribution of floristic assemblages across a Pinyon-Juniper woodland in Northern Arizona. As a result of the area's volcanic history, differently aged ecosystems, i.e., those at different stages of soil development and thus with drastically different soil properties, are spatially segregated across the landscape in sequential fashion. Edaphic factors known to influence plant community composition, which include soil texture, pH, organic carbon, total nitrogen, and phosphorus, varied across the study area. Consequently, the lower San Francisco volcanic field (San Francisco VF) is an ideal venue in which to examine the role of edaphic drivers on local-scale floristic dynamics.
The San Francisco VF of northern Arizona was created by at least seven major eruptive events and is one of the major basaltic volcanic fields on the Colorado Plateau (Cooley, 1962; Tanaka et al., 1986). Basaltic volcanism began roughly six million years ago and has occurred continuously for the past three million years (Moore et al., 1976). The volcanic field includes over 600 Pliocene, Pleistocene, and Holocene volcanic vents, volcanoes, cinder cones, and their associated sheet deposits and lava flows (Tanaka et al., 1986) and covers approximately 5,000 km2 (Moore et al., 1976). The westward drift of the North American plate coupled with local magnetism at lithospheric junctions explains a general eastward migration of volcanism across the San Francisco VF and has resulted in a dramatic ecosystems age gradient (Tanaka et al., 1986). Ecosystems of the "lower" San Francisco VF (those dominated by Pinyon-Juniper woodlands as opposed to higher-elevation coniferous forests) range from less than one thousand years old at the eastern edge of the volcanic field to over four million years old at the southwestern edge, which is only 100 km away (Tanaka et al., 1986; Moore and Wolfe, 1987; Wolfe et al., 1987a, Wolfe et al., 1987b; Fig. 1).
Geologic history, parent material, elevation, climate, and major vegetation type are consistent across the lower San Francisco VF, yet the degree of soil development varies dramatically between sites. Soil chemistry, including total soil nitrogen, soil organic carbon, and plant-available phosphorus, varies differentially with ecosystem age (Selmants, 2007; Selmants and Hart, 2008). Soil texture varies with ecosystem age as well; relatively old sites (over three million years) of the western San Francisco VF have fine-textured soil while relatively young sites (less than 150,000 years) of the eastern San Francisco VF have coarsely-textured soil with very little clay content (Moore and Wolfe, 1987; Wolfe et al., 1987a; Wolfe et al., 1987b; Selmants, 2007; Selmants and Hart, 2008). Total soil nitrogen and soil organic carbon increased following volcanism for roughly one million years and then subsequently decreased (Selmants and Hart, 2008). Plant-available phosphorus peaked at roughly 100,000 years, stayed relatively constant until one million years after volcanism, and then decreased (Selmants, 2007; Selmants and Hart, 2010). Table 1 provides a profile of soil characteristics for the San Francisco VF chronosequence (data from Selmants, 2007). Table 2 provides a climate summary for the study area.
Edaphic characteristics vary spatially across the lower San Francisco VF due to its volcanic history and, therefore, the study system lends itself to a chronosequence approach. Chronosequences, or ecosystem age gradients, act as spatial surrogates for temporal processes (Hugget, 1998). The ideal chronosequence consists of a series of sites or ecosystems in which biotic and abiotic influences are consistent enough so that differences among sites can be solely attributed to the influence of time. Parent geology, climatic factors, disturbance regimes, major vegetation type, and elevation must be consistent for this approach to be effective (Phillips, 2004). The concept of using an ecosystem age gradient as a method to understand plant-substrate interaction seems to have originated in 1899 with Cowles' research on the floristic dynamics of sand dunes near Lake Michigan (Cowles, 1899), and it remains an effective technique to study soil development, soil geography, and plant ecology (e.g., Kardol et al., 2006; Turner et al., 2012). Despite criticisms of the chronosequence approach (see Johnson and Miyanishi, 2008), Walker et al. (2010) reviewed some of its misapplications and concluded that careful use of the technique, particularly in low-biodiversity, low-disturbance, long-term systems, remains a valid way to assess the effects of soil development on community processes.
Numerous studies have documented a correlation between soil characteristics and plant community composition; however, no consistent trends emerge across regions or ecosystems. Both soil texture and soil fertility have been variously shown to drive differences in community composition. Crews et al. (1995) and Paoli et al. (2006) indicated that sites of different ages, and thus those with different soil characteristics, supported different plant communities in tropical rainforests. Kroel-Dulay et al. (2004) showed that soil texture drove compositional differences between two grassland communities in the Chihuahuan Desert, while soil texture can also drive grassland-forest boundaries in tropical wetlands (Zeilhofer and Schessel, 1999). Van der Welle et al. (2003) illustrated that soil moisture and pH influenced community composition in tundra wetlands in Alaska. Alternatively, soil fertility can also largely drive compositional differences between sites. Total soil nitrogen and C:N ratios influence differentiation between grassland and shrubland vegetation types in a Mediterranean climate. Chiarucci (2001) found that soils with higher total nitrogen and higher C:N ratios supported shrublands while soils with lower nitrogen and lower C:N supported grasslands. Nitrogen availability has also been shown to drive compositional differences in successional woodlands and forests (Prevosto, 2004). Soil texture and soil chemistry often likely interact to influence plant community composition. Thwaites and Cowling (1998) documented that the synergy of soil characteristics, including nutrient content, texture, and pH, drove compositional differences in fynbos ecosystems. Similarly, a combination of soil salinity, clay content, and soil organic matter strongly explained the plant-environment relationship in an extreme desert ecosystem in Egypt (El-Ghani, 2000).
No clear correlation between ecosystem age (and resulting soil characteristics) and species richness has emerged in the literature. Across long-term chronosequences in Hawaii, Kitayama and Mueller-Dombois (1995) found no linear trend in species richness across the age gradient, while Crews et al. (1995) showed that species richness increased with substrate age. Conversely, species richness decreased with ecosystem age across a one million-year age gradient in the Alaskan tundra (Vander Welle et al., 2003). Species richness is often higher on poorly-drained (fine-textured) substrates as compared to well-drained soils (Gerhardt and Foster, 2002; Hardtle et al., 2003). In an analysis of six long-term chronosequences in boreal, temperate, and sub-tropical regions, Wardle et al. (2008) found evidence that species richness typically varied across soil fertility gradients. Their evidence suggests that as ecosystems age and soil phosphorous declines, tree diversity decreases and understory species richness increases.
Thus, while no consistent pattern emerges, it has been strongly documented that differences in soil characteristics shape plant community composition and species richness patterns. The purpose of this study was to contribute findings from the unique edaphic landscape of the San Francisco VF to this discussion and to determine if differently aged substrates within the lower San Francisco VF harbored distinct floristic assemblages. It was hypothesized that 1) differently aged substrates of the lower San Francisco VF would support distinct plant communities; and 2) species richness would increase with substrate age.
Methods--The lower San Francisco VF was divided into an ecosystems age gradient or chronosequence consisting of five age classes. Ecosystem age and parent geology were determined from U. S. Geological Survey geologic maps (Moore and Wolfe, 1987; Newhall et al., 1987; Ulrich and Bailey, 1987; Wolfe et al., 1987a; Wolfe et al., 1987b). Five age classes were named Sunset, O'Neill, Red, Hobble, and Cedar based upon representative cinder cones. An effort was made to characterize geographically and chronologically distinct age classes based upon a single geologic unit as defined by U. S. Geological Survey geologic maps. Age classes were essentially mutually exclusive and defined by a single geologic type (basalt flows and cinder cones); however, a chronological overlap did occur between the Red and Hobble age classes from approximately 0.96 to 0.73 million years ago (Table 1). In addition to basaltic flows and cinder cones, the Sunset class included associated pyroclastic sheet deposits, and the O'Neill class included basaltic andesite flows and cinder cones.
Using a geographically explicit input boundary of the lower San Francisco VF and soil age data from U. S. Geological Survey geologic maps, 500 randomly generated Universal Transverse Mercator coordinates were input into a Geographic Information System in order to determine potential study site locations. Universal Transverse Mercator coordinates falling outside of geologic unit or age class parameters were rejected, while the first four Universal Transverse Mercator locations falling within each age class were accepted as plot locations. Four replicate 50 m x 50 m (0.25 ha) study plots were created for each of the five age class treatments (Fig. 1). The southwest corner of each plot was placed at the randomly generated Universal Transverse Mercator location and western plot edges were aligned with a 0[degrees] azimuth.
In order to compare the floristic composition of these ecosystems, plots were surveyed between 20 August 2005 and 16 September 2005. This was the optimal survey window, as both spring-flowering and monsoonal species were present and identifiable at this time. A complete list of vascular plant species occurrence was recorded for each plot. Abundance and frequency were not measured and each species was given equal weight based upon occurrence. A multiple response permutation procedure (MRPP) was conducted in PC-ORD (McCune and Mefford, 1999) to test the null hypothesis of no difference in floristic composition between differently aged sites. Sorensen distances and default group weightings were used in the MRPP analysis. Sorensen distance was chosen over Euclidian distance because its de-emphasis of outliers is better suited for assessing vegetation composition data (Faith et al., 1987). MRPP analyses effectively assess differences among a priori groups and entail lax requirements on the data structure (Zimmerman et al., 1985). The function on an MRPP test is essentially the same as a one-way analysis of variance (ANOVA); however, it does not require the assumptions of normally distributed populations or homogeneous variances because the P-value is derived from a permutation of the actual sample data (Zimmerman et al., 1985). A nonmetric multidimensional scaling (NMDS) ordination in two dimensions using Bray-Curtis distance and binary input was created to visually depict dissimilarity structure of the data using the vegan package (Oksanen et al., 2012) in R (R Core Team, 2012),.
An MRPP begins by computing individual distance measures for each pair-wise comparison within groups and then computes the weighted average pair-wise distance for each group (Zimmerman et al., 1985). The calculated distance measures are compared to a pool of all (including between-group) potential permutations of the input data, which preserves the group sample structure within the combined data. The P-value of an individual statistic is determined by its relative position on an ordered list of possible permutation-derived distance statistics. Because even modest experiments entail enormous permutations, the actual test statistic (T) is computed through a continuous probability function using a Pearson type III distribution (Zimmerman et al., 1985). Lastly, an MRPP calculates an A statistic (chance corrected within group agreement) to measure group similarity. If A = 1 then all samples within a group are identical, while if A = 0 then within-group dissimilarity is the same as expected by chance alone (McCune and Mefford, 1999). As the initial MRPP was significant, subsequent pair-wise MRPP comparisons were made between all age classes.
An indicator species analysis was conducted to determine characteristic species for each of the differently aged ecosystems of the SF Volcanic Field using PC-ORD (McCune and Mefford, 1999) following the methods of Dufrene and Legendre (1997). The significance of species indicator values was assessed using a Monte Carlo test with default settings and 1,000 permutations (McCune and Mefford, 1999). The approach of Dufrene and Legendre (1997) determines indicator values based upon a combination of how faithful a species is to a particular group and how frequently it occurs within that group. Species that occur uniquely within a single age class, and frequently within that age class, are excellent indicators whereas species that occur throughout different age classes, and infrequently within any given age class, are poor indicators. Following the results from the pair-wise MRPP tests, the species indicator analysis was based upon two groups of age classes: one group consisting of the younger sites (Sunset and O'Neill) and another consisting of the older sites (Red, Hobble, and Cedar). Species were determined to be significant indicators if they had indicator values of 50 or greater and P-values of 0.05 or less.
A Shapiro-Wilk test was used to test the equal variance assumption, and then a null hypothesis of no difference in species richness between age classes was tested with a one-way ANOVA. A linear regression of species richness on ecosystem age was created in R.
Results--The initial MRPP test rejected the null hypothesis and indicated a significant difference in floristic composition between differently aged sites (A = 0.40, P < 0.0001). A permutational MANOVA using the adonis function in the vegan package in R corroborated this result (P < 0.001). Subsequent pair-wise MRPP tests indicated no significant difference in floristic composition between Sunset and O'Neill (A = 0.01), Red and Hobble (A = 0.01), Red and Cedar (A = 0.1), and Hobble and Cedar (A = 0.03). There was a significant difference (P < 0.01) between Sunset and age classes Red (A = 0.44), Hobble (A = 0.44), and Cedar (A = 0.43) and also between O'Neill and age classes Red (A = 0.34), Hobble (A = 0.29), and Cedar (A = 0.26).
Pair-wise MRPP tests suggested that two distinct, essentially nonoverlapping floristic assemblages existed within the lower San Francisco VF. The Sunset and O'Neill age classes (those ecosystems younger than 150,00 years) formed a distinct floristic group while the Red, Hobble, and Cedar age classes (those ecosystems 350,000 to 4.4 million years old) formed another coherent floristic assemblage. An NMDS ordination of the data corroborates this pattern (Fig. 2). A secondary NMDS ordination using Raup-Crick dissimilarity to account for varying species richness between differently aged sites was robust to the original pattern. Significant environmental vectors including total soil nitrogen, soil organic carbon, and clay content (soil texture) are depicted by loading arrows in Figure 2; pH and soil phosphorous were nonsignificant. The stress statistics (S) for numerous runs of the ordination were consistently between 0.11 and 0.14, indicating a good to very acceptable goodness of fit for the NMDS ordination.
The two distinct floristic elements determined via pairwise MRPP tests, and supported by NMDS ordinations, were used for indicator species analysis. Apache plume (Fallugia paradoxa), wax current (Ribes cereum), skunkbush sumac (Rhus trilobata), Diamond Valley suncup (Camissonia gouldii), and California brickellbush (Brickellia californica) were most characteristic of young ecosystems; while blue grama grass (Bouteloua gracilis), squirreltail (Elymus elymoides), broom snakeweed (Gutierrezia sarothrae), rose heath (Chaetopappa ericoides), Abert's creeping zinnia (Sanvitalia abertii), and pingue rubberweed (Hymenoxys richardsonii) were most characteristic of older ecosystems (Table 3).
Lastly, a one-way ANOVA indicated a significant difference in species richness between the five differently aged ecosystems of the chronosequence (P < 0.01). A second one-way ANOVA between the "young" (Sunset and O'Neill) and the "old" (Red, Hobble, and Cedar) ecosystems was also significant (P < 0.001). Linear regression showed a positive correlation between ecosystem age and species richness (P < 0.01, [R.sup.2] = 0.32; Fig. 3).
Discussion--The vegetation-environment relationship, and specifically how soil characteristics influence floristic composition, is a complex multidimensional problem. individual species have unique ecological and physiological tolerances; thus, it is often difficult to explain how the synergy of environmental variables drives patterns of species distribution or community composition. it is often possible to observe a pattern of floristic differences between sites; however, it can be difficult to infer causality from the pattern or to extend inferences from one ecosystem to another. Variation in soil chemistry often promotes site specialization at the regional level in fynbos and desert ecosystems (Thwaites and Cowling, 1988; ElGhani and Amer, 2003). Nitrogen availability can primarily shape floristic composition in Mediterranean woodlands (Prevosto et al., 2004), while phosphorus and micronutrient availability can primarily drive vegetation composition in ecosystems ranging from semidisturbed grasslands in Europe to tropical rainforests in Indonesia (McCrea et al., 2001; Paoli et al., 2006). Similarly, soil texture and its associated water-holding capacity can drive floristic composition in arid or semiarid environments (Chiarucci et al., 2001; Kroel-Dulay et al., 2004). Often a combination of soil chemistry and soil structure drives floristic differentiation (El-Ghani, 2000).
While it may be impossible to discover generality on the species-environment conundrum, this study provides a needed example from a temperate coniferous woodland of the Southwestern United States. In the lower San Francisco VF, the soil texture, soil organic carbon, and soil nitrogen associated with soil development all influence floristic composition and local-scale plant geography. Young ecosystems (those on the eastern side of the San Francisco VF) with very coarse soil and low water-holding capacity support shrubs and some characteristic annuals, while older ecosystems (those on the western side of the San Francisco VF) with fine-textured soil and higher water-holding capacity primarily support grasses and perennial herbs (Table 3). The extensive and relatively deep root systems of shrubs, and the opportunistic nature of monsoonal annuals, maximize the plants' respective abilities to acquire water in xeric soil environments. This pattern has been noted in extreme desert systems, as shrubs can access moisture at various depths while grasses are essentially restricted to moisture in the topmost layer of soil (El-Ghani, 2000). As soils age, weathering reduces particle size (i.e., increases clay content) and subsequently increases water-holding capacity (Selmants, 2007). At some threshold of available soil moisture, grasses (specifically B. gracilis and E. elymoides) are able to colonize local ecosystems. Other highly competitive perennial herbs such as G. sarothrae and H. richardsonii soon follow. Soil texture acting as the primary driver of floristic composition is consistent with other studies of vegetation-soil relationships in grasslands, savannahs, and xeric environments, although an experimental approach is required to directly test this prediction. Floristic composition among differently aged sites in the lower San Francisco VF seems to quickly homogenize once soil texture becomes adequately fine. At this point climatic factors and regional species pools, as opposed to the qualities of soil structure or chemistry, seem to become the chief drivers of floristic composition.
In the San Francisco VF, a shift from widely spaced shrubs and low overall vegetation cover, to relatively abundant grasses and perennial herbs and higher cover, occurred prior to 150,000 years of soil development. The O'Neill age class, defined by Moore and Wolfe (1987) as extrusive rocks of the Holocene and youngest Pleistocene, encompasses sites from 1,000 to 150,000 years old. Two of the plots sampled within this age class appeared similar to the younger, shrub-dominated Holocene ecosystems found in the Sunset age class, while the two other O'Neill plots showed floristic composition more similar to the older grass and perennial-dominated Pleistocene and Pliocene sites. Figure 2 shows the transitional nature of O'Neill plots 1 and 2 between the "young" and "old" floristic elements.
As younger ecosystems age, their floristic composition will likely become similar to the rest of the volcanic field. Grasses and perennial herbs will presumably colonize weathered soils and eventually out-compete many of the currently dominant shrubs and characteristic annuals of these young and barren ecosystems. This competition will likely have little effect on many of the shrubs (e.g., R. cereum, R. trilobata, B. californica) at the community scale, as they occur in other scattered microhabitats throughout the greater San Francisco VF ecosystem. Such a shift, however, may have profound consequences on the local cinder-loving endemics. Several edaphically limited, and endemic or near-endemic species (Sunset Crater beardtongue [Penstemon clutei], saw phacelia [Phacelia serrata], C. gouldii, and Sunset Crater blazingstar [Mentzelia collomiae]) have exploited a suitable, competition-free habitat in the barren soils of the youngest substrates of the San Francisco VF. The future of these narrowly restricted edaphic endemics balances precariously on an uncertain volcanic future. Substrates of extant ecosystems will inexorably weather and undergo soil development; yet additional volcanic eruptions are predicted to occur at three- to four thousand-year intervals in the San Francisco VF (Wolfe et al., 1983) and perhaps will continue to create suitable barren habitats for these cinder-loving species.
In the lower San Francisco VF species richness is positively correlated with ecosystem age (Fig. 3). This finding is consistent with the findings of Wardle et al. (2008) across other chronosequences and provides needed support of the pattern in temperate ecosystems. Interestingly, species richness in the San Francisco VF qualitatively decreases as nutrients are leached from the soil in the oldest ecosystems. This suggests that soil richness may track nutrient availability, although an experimental approach is required to test this hypothesis.
Differential characteristics of soil texture and macronutrient availability associated with soil age strongly drive both the floristic composition and the spatial distribution of plant assemblages in the lower San Francisco VF. These findings contribute to the body of knowledge on the species-environment relationship and corroborate broad findings that soil characteristics often strongly drive plant community composition. Despite the inconsistency of species-environment relationships in different ecosystems, this study suggests an interesting line for further investigation. In the San Francisco VF chronosequence, over 25% of the increase in clay content occurs within the first 5% of chronological soil development, and this 150,000-year boundary demarcates the two locally distinct floristic assemblages. No additional compositional differentiation seems to occur within local plant communities during the next four million years of soil development despite major changes in soil fertility. This pattern might imply that soil texture is actually the primary driver of plant community composition in the xeric San Francisco VF. Additional research on edaphic-plant relationships in both xeric and mesic ecosystems will test the generality of this pattern and help to determine the relative importance of soil texture and soil chemistry in shaping plant communities across different environments and moisture regimes.
I thank T. Ayers for her mentorship, M. Kearsley for his statistical guidance and sense of humor, and R. Scott for his advice on teaching and life. P. Selmants introduced me to the biotic and abiotic patterns of the San Francisco VF and shared with me much of his nutrient data. Thanks to H. D. Hammond for his selfless dedication to the Deaver herbarium and his sharp eye for editing. Three anonymous reviewers greatly improved the quality of this manuscript through their comments.
Chiarucci, A., D. Rocchini, C. Leonzio, and V. De Dominicis. 2001. A test of vegetation-environment relationship in serpentine soils of Tuscany, Italy. Ecological Research 16:627-639.
Cooley, M. E. 1962. Geomorphology and the age of volcanic rocks in northeastern Arizona. Arizona Geological Society Digest 5:97-115.
Cowles, H. C. 1899. The ecological relations of the vegetation on the sand dunes of Lake Michigan. Part I. Geographical relations of the dune floras. Botanical Gazette 27:95-117.
Crews, T. E., K. Kitayama,J. H. Fownes, R. H. Riley, D. A. Herbert, D. Mueller-Dombois, and P. Vitousek. 1995. Changes in soil phosphorus fractions and ecosystems dynamics across a long chronosequence in Hawaii. Ecology 76:1407-1424.
Dufrene, M., and P. Legendre. 1997. Species assemblages and indicator species: the need for a flexible asymmetrical approach. Ecological Monographs 67:345-366.
El-Ghani, M. M. A. 2000. Floristics and environmental relations in two extreme desert zones of western Egypt. Global Ecology and Biogeography 9:499-516.
El-Ghani, M. M. A., and W. M. Amer. 2003. Soil-vegetation relationships in a coastal desert plain of southern Sinai, Egypt. Journal of Arid Environments 55:607-628.
Faith, D. P., P. R. Minchin, and L. Belbin. 1987. Compositional dissimilarity as a robust measure of ecological distance. Vegetatio 69:57-68.
Gerhardt, F., and D. R. Foster. 2002. Physiographical and historical effects on forest vegetation in central New England, USA. Journal of Biogeography 29:1421-1437.
Good, D. O. 1931. A theory of plant geography. New Phytologist 30:149-171.
Hardtle, W., G. von Oheimb, and C. Westphal. 2003. The effects of light and soil conditions on the species richness of the ground vegetation of deciduous forests in northern Germany (Schleswig-Holstein). Forest Ecology and Management 182:327-338.
Hooper, D. U., D. E. Bignell, V. K. Brown, L. Brussard, J. M. Dangerfield, D. H. Wall, D. A. Wardle, D. C. Coleman, K. E. Giller, P. Lavelle, W. H. Van Der Putten, P. C. De Ruiter, J. Rusek, W. L. Silver, J. M. Tiedje, and V. Wolters. 2000. interactions between aboveground and belowground biodiversity in terrestrial ecosystems: patterns, mechanisms, and feedbacks. BioScience 50:1049-1061.
Hugget, R. J. 1998. Soil chronosequences, soil development, and soil evolution: A critical review. Catena 32:155-172.
Johnson, E. A., and K. Miyanishi. 2008. Testing the assumptions of chronosequences in succession. Ecology Letters 11:419-431.
Kardol, P., T. M. Bezemer, and W. H. van der Putten. 2006. Temporal variation in plant-soil feedback controls succession. Ecology Letters 9:1080-1088.
Kitayama, K., and D. Mueller-Dombois. 1995. Vegetation changes along gradients of long-term soil development in the Hawaiian montane rainforest zone. Vegetatio 120:1-20.
Kroel-Dulay, G., P. (Odor, D. P. C. Peters, and T. Hochstrasser. 2004. Distribution of plant species at a biome transition zone in New Mexico. Journal of Vegetation Science 15:531-538.
McCrea, A. R., I. C. Trueman, M. A, Fullen, M. D. Atkinson, and L. Besenyei. 2001. Relationships between soil characteristics and species richness in two botanically heterogeneous created meadows in the urban English West Midlands. Biological Conservation 97:171-180.
McCune, B., and M. J. Mefford. 1999. PC-ORD. Multivariate Analysis of Ecological Data. Version 4.34. MjM Software, Gleneden Beach, oregon.
Moore, R. B., and E. W. Wolfe. 1987. Geologic map of the east part of the San Francisco volcanic field, north-central Arizona. Miscellaneous Field Studies Map (MF-1960). Department of the Interior. U.S. Geological Survey.
Moore, R. B., E. W. Wolfe, and G. E. Ulrich. 1976. Volcanic rocks of the eastern and northern parts of the San Francisco volcanic field, Arizona. Journal of Research of the U.S. Geological Survey 4:549-560.
Newhall, C. G., G. E. Ulrich, and E. W. Wolfe. 1987. Geologic map of the southwest part of the San Francisco volcanic field, north-central Arizona. Miscellaneous Field Studies Map (MF-1958). Department of the Interior. U.S. Geological Survey.
Oksanen, J., F. Guillaume Blanchet, R. Kindt, P. Legendre, P. R. Minchin, R. B. O'Hara, G. L. Simpson, P. Solymos, M. M. H. Stevens, and H. Wagner. 2012. vegan: Community Ecology Package. R package version 2.0-5. http://CRAN.R-project. org/package=vegan
Paoli, G. D., L. M. Curran, and D. R. Zak. 2006. Soil nutrients and beta diversity in the Bornean Dipterocarpaceae: evidence for niche partitioning by tropical rain forest trees. Journal of Ecology 94:157-170.
Phillips, J. D. 2004. Divergence, sensitivity, and non-equilibrium in ecosystems. Geographical Analysis 36:369-383.
Prevosto, B., E. Dambrine, C. Moares, and T. Curt. 2004. Effects of volcanic ash chemistry and former agricultural use on the soils and vegetation of naturally regenerated woodlands in the Massif Central, France. Catena 56:239-261.
R Core Team. 2012. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0. Available at: http://www. R-project.org/.
Ricklefs, R. E. 1987. Community diversity: relative roles of local and regional processes. Science 235:167-171.
Selmants, P. C. 2007. Carbon, nitrogen, and phosphorus dynamics across a three million yr. substrate age gradient in northern Arizona, USA. Ph.D dissertation. Northern Arizona University, Flagstaff.
Selmants, P. C., and S. C. Hart. 2008. Substrate age and tree islands influence carbon and nitrogen dynamics across a semiarid retrogressive chronosequence. Global Biogeochemical Cycles 22: GB1021, doi:10.1029/2007GB003062.
Selmants, P. C., and S. C. Hart. 2010. Phosphorus and soil development: does the Walker and Syers model apply to semiarid ecosystems? Ecology 91:474-484.
Siefert, A., C. Ravenscroft, D. Althoff, J. C. Alvarez-Yepiz, B. E. Carter, K. L. Glennon,J. M. Herberling, I. S. Jo, A. Pontes, A. Sauer, A. Willis, and J. D. Fridley. 2012. Scale dependence of vegetation-environment relationships: a meta-analysis of multivariate data. Journal of Vegetation Science 23:942-951.
Tanaka, K. L., E. M. Shoemaker, G. E. Ulrich, and E. W. Wolfe. 1986. Migration of volcanism in the San Francisco volcanic field, Arizona. Geological Society of America Bulletin 97:129-141.
Thwaites, R. N., and R. M. Cowling. 1988. Soil-vegetation relationships on the Agulhas Plain, South Africa. Catena 15:333-345.
Turner, B. L., A. Wells, K. M. Anderson, and L. M. Condron. 2012. Patterns of tree community composition along a coastal dune chronosequence in lowland temperate rain forest in New Zealand. Plant Ecology 213:1525-1541.
Ulrich, G. E., and N. G. Bailey. 1987. Geologic map of the SP mountain part of the San Francisco volcanic field, northcentral Arizona. Miscellaneous Field Studies Map (MF-1956). Department of the Interior. U.S. Geological Survey.
Van der Welle, M. E. W., P. J. Vermeulen, G. R. Shaver, and F. Berendse. 2003. Factors determining plant species richness in Alaskan arctic tundra. Journal ofVegetation Science 14:711-720.
Vivian-Smith, G. 1997. Microtopographic heterogeneity and floristic diversity in experimental wetland communities. The Journal of Ecology 85:71-82.
Walker, L., D. A. Wardle, R. D. Bardgett, and B. D. Clarkson. 2010. The use ofchronosequences in studies ofecological succession and soil development. Journal of Ecology 98:725-736.
Wardle, D. A., R. D. Bardgett, R. R. Walker, D. A. Peltzer, and A. Lagerstro. 2008. The response of plant diversity to ecosystem retrogression: evidence from contrasting long-term chronosequences. Oikos 117:93-103.
Westoby, M., and I. J. Wright. 2006. Land-plant ecology on the basis of functional traits. Trends in Ecology and Evolution 21:261-268.
Wolfe E. W., G. E. Ulrich, and R. B Moore. 1983. San Francisco volcanic field, Arizona. Volcano News 13:1-3.
Wolfe, E. W., G. E. Ulrich, and C. G. Newhall. 1987a. Geologic map of the northwest part of the San Francisco volcanic field, north-central Arizona. Miscellaneous Field Studies Map (MF1957). Department of the Interior. U.S. Geological Survey.
Wolfe, E. W., G. E. Ulrich, R. F. Holm, R. B. Moore, and C. G. Newhall. 1987b. Geologic map of the central part of the San Francisco volcanic field, north-central Arizona. Miscellaneous Field Studies Map (MF-1959). Department of the Interior. U.S. Geological Survey.
Zeilhofer, P., and M. Schessel. 1999. Relationship between vegetation and environmental conditions in northern Pantanal of Mato Grosso, Brazil. Journal of Biogeography 27:159-168.
Zimmerman, G. M., H. Goetz, and P. W. Mielke. 1985. Use of an improved statistical method for group comparisons to study effects of prairie fire. Ecology 66:606-611.
Submitted 7 August 2011.
Acceptance recommended by Associate Editor, Florence M. Oxley, 5 December 2013.
Deaver Herbarium, Department of Biological Sciences, Northern Arizona University, Flagstaff, AZ 86011-5640
Table 1--Soil characteristics of San Francisco Volcanic Field chronosequence. Site Ecosystem age Approx. age of pH Clay range ([10.sup.3] soil samples (g/kg) years) used for analysis ([10.sup.3] years) Sunset 0.9-1.1 1 6.74 11.30 O'Neill 1-150 55 6.56 83.80 Red 350-1,000 750 6.70 318.40 Hobble (a) 750-2,500 1625 6.49 360.00 Cedar 2,800-4,400 3000 6.17 426.20 Site Soil organic Total Labile C (g/kg) soil N P (mg/kg) (g/kg) Sunset 3.10 0.20 11.12 O'Neill 9.60 0.99 62.50 Red 23.20 1.40 51.85 Hobble (a) 19.10 1.28 41.50 Cedar 12.80 1.10 25.61 * Nutrient data unavailable for the Hobble age class; values extrapolated. Table 2--Climate summary of the San Francisco Volcanic Field chronosequence based on data collected from 2002-2005. (a) Site Approx. site Weather Weather Elevation age ([10.sup.3] station station (m) years) latitude (b) longitude (b) Sunset 1 35.39389 -111.42389 1,905 O'Neill 55 35.24611 -111.45889 1,941 Red 750 35.53806 -111.86694 2,073 Cedar 3,000 35.37444 -111.14111 2,003 Site MAP (c) MAT (c) PET (c) (mm) ([degrees]C) (mm/yr) Sunset 328 12 1,325 O'Neill 352 11 1,328 Red 325 11 1,334 Cedar 338 11 1,324 (a) Adapted from Selmants and Hart, 2010. (b) Weather station unavailable for Hobble site. (c) MAP = mean annual precipitation; MAT = mean annual temperature; PET = potential evapotranspiration. Table 3--Summary of indicator species analysis. Species Group Indicator Life form P-value (c) (a) value (b) (%) Fallugia paradoxa young 100 Shrub 0.001 Ribes cereum young 67.5 Shrub 0.002 Rhus trilobata young 68.1 Shrub 0.014 Camissonia gouldii young 50 Annual 0.018 Brickellia californica young 61.4 Shrub 0.019 Ephedra viridis * young 37.5 * Shrub 0.052 * Bouteloua gracilis old 80.7 Grass 0.001 Elymus elymoides old 80 Grass 0.002 Gutierrezia sarothrae old 80 Perennial 0.002 Chaetopappa ericoides old 72.5 Perennial 0.004 Sanvitalia abertii old 66.7 Annual 0.004 Hymenoxys richardsonii old 58.3 Perennial 0.014 Heliomeris longifolia old 58.3 Annual 0.016 var. annua Mahonia fremontii old 58.3 Shrub 0.016 Erigeron divergens old 58.3 Biennial 0.017 Sphaeralcea fendleri old 58.3 Perennial 0.019 (a) The young group includes the Sunset and O'Neill sites while the old group includes the Red, Hobble, and Cedar sites. (b) Indicator value indicates % of perfect indication based upon fidelity to a given group combined with frequency within that group. (b) P-values derived from a Monte Carlo randomization test based upon 1,000 permutations. * Not statistically significant but a good qualitative indicator.
|Printer friendly Cite/link Email Feedback|
|Date:||Sep 1, 2014|
|Previous Article:||Distribution and nesting success of ferruginous hawks and Swainson's hawks on an agricultural landscape in the great plains.|
|Next Article:||Mountain lion habitat selection in Arizona.|