Southeastern Naturalist 467 A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 22001166 SOUTHEASTERN NATURALIST 1V5o(3l.) :1456,7 N–4o8. 73 Field Performance and Common-Garden Differentiation in Response to Resource Availability in Helianthus porteri (A. Gray) Pruski, a Granite-Outcrop Endemic Alan W. Bowsher1,*, Scott D. Gevaert1,2, and Lisa A. Donovan1 Abstract - Plant communities on granite outcrops are prone to water and nutrient limitations because of the shallow soils and high summer-temperatures characteristic of that habitat. Thus, resource availability is expected to influence plant performance on these sites. Here, we sought to determine if low resource-availability (water and nutrients) acts as a selective agent in natural populations of Helianthus porteri (Porter’s Sunflower), a granite-outcrop endemic. We conducted field observations of growth and survival for 3 years in 3 populations spanning the species’ range. We found that survival to flowering was correlated with soil-water availability (estimated by plant predawn water-potentials; Ψpd), and that drought severity (estimated by survival and Ψpd) differed among populations, suggesting the populations could be adaptively differentiated for resource-use traits. However, in a greenhouse common-garden experiment, the population that experienced the greatest drought severity in the field did not exhibit traits associated with greater drought resistance, and all 3 populations responded similarly to water and nutrient limitations. Thus, although drought influences survival in Porter’s Sunflower, populations do not appear to be adaptively differentiated with respect to resource availability. Introduction Granite outcrops in the southeastern US are geographically isolated habitats that are home to a variety of endemic plant species. These outcrops, which occur from eastern Alabama to Virginia, vary from small (several square meters) rock islands surrounded by forests to extremely large (>100 ha) formations that tower above the surrounding landscape (McVaugh 1943). Plant communities on these outcrops occupy shallow soils that accumulate in depressions on the rock surface (commonly referred to as soil islands) and at the margins of the outcrops (Burbanck and Platt 1964, McVaugh 1943). Despite being in a relatively moist climatic region, severe, recurrent drought-stress is a hallmark of these communities given the shallow soil depth, high irradiance, and high surface-temperatures of rock outcrops (Cumming 1969, McCormick et al. 1974, Sharitz and McCormick 1973, Shure and Ragsdale 1977, Wyatt 1997). In addition, water stress is often accompanied by nutrient stress because drought typically reduces both the mobility (Chapin 1991, Lambers et al. 2008) and biological transformations of nutrients in soil (Fierer and Schimel 2002, Sanaullah et al. 2012, Schimel et al. 2007). Thus, very limited availability of resources is expected to be a major selective pressure on survival and local adaptation 1Department of Plant Biology, University of Georgia, Athens, GA, 30602. 2Current address - Biology Department, St. Louis Community College-Florissant Valley, St. Louis, MO 63135. *Corresponding author - alanwbowsher@gmail.com. Manuscript Editor: Justin Hart Southeastern Naturalist A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 468 on granite outcrops (Chapman and Jones 1975, McCormick and Platt 1964, Mellinger 1972, Shelton 1963). One strategy by which plants adapt to scarcity of water and nutrients is through avoidance of resource limitation (i.e., the prevention of internal resource stress by maximizing resource uptake and/or minimizing loss; Verslues et al. 2006). For example, high root:total biomass ratio (RMR) can increase uptake of water and nutrients per unit of plant mass (Aerts and Chapin 1999), while low stomatal conductance (g) helps to conserve water by reducing transpirational water-loss (Héroult et al. 2013). Similarly, smaller, thicker leaves decrease transpirational water-loss because of their reduced boundary layer (McDonald et al. 2003, Smith 1978), and lower specific leaf-area (leaf area mass-1) often results in longer leaflifespan, conferring greater conservation of water and nutrients (Reich et al. 1992, 1997; Wright et al. 2004). Conservative use of resources through greater water-use efficiency (WUE, ratio of carbon gain to water loss) and photosynthetic nitrogenuse efficiency (PNUE; ratio of carbon gain per leaf N content) are also expected to contribute to success in low-resource environments (Donovan and Ehleringer 1994, Funk and Vitousek 2007). However, although such traits are understood to improve plant performance in conditions where resurces are limited, constitutive expression of stress-avoidance traits may come at the cost of plant performance in conditions where resources are abundant (Sambatti and Caylor 2007). For example, high WUE resulting from low stomatal conductance can limit plant growth under well-watered conditions by limiting carbon assimilation (Héroult et al. 2013). In habitats with large temporal fluctuations in resource availability, such as granite outcrops, plant success may therefore be associated with the capacity for rapid growth following rainfall, paired with the ability to shift towards a slow-growing, stress-avoiding phenotype (e.g., higher RMR or WUE in response to drought) when drought conditions return (Lugo 1969, Lugo and McCormick 1981, Shure and Ragsdale 1977). Helianthus porteri (A. Gray) Pruski (Porter’s Sunflower or Confederate Daisy), formerly Viguiera porteri (Pruski 1998), is an endemic annual sunflower found on granite outcrops throughout the southeastern US (Kartesz 2015). Although Porter’s Sunflower often dominates the annual–perennial successional zone (maximum soil depth of 14–39 cm; Burbanck and Platt 1964, McCormick et al. 1974, Mellinger 1972, Shelton 1963), field studies have reported substantial variation in the species’ performance within and among outcrops and among years (Burbanck and Phillips 1983, Cumming 1969, Houle and Phillips 1989, Mellinger 1972, Shure and Ragsdale 1977). Both field observations (Burbanck and Platt 1964, Houle and Phillips 1989, McCormick 1959, Mellinger 1972) and studies of simulated outcrops (Cumming 1969, McCormick 1959, Mellinger 1972) suggest that drought strongly influences productivity and survival of Porter’s Sunflower on the outcrops. However, to our knowledge, no studies have linked survival directly to soil-water availability (assessed as plant predawn water potential; Ψpd) for native popoulations of this species. Such a relationship would confirm drought as a major selective pressure on granite outcrops, and would suggest water- and nutrient-use traits are under strong selection in Porter’s Sunflower. Southeastern Naturalist 469 A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 In addition, although water and nutrient availability within granite outcrops is temporally variable, studies have also reported variation in these resources among outcrops throughout the Southeast (McCormick and Platt 1964, Shure and Ragsdale 1977). Given its wide geographic range within the region, differences in resource availability among Porter’s Sunflower populations could drive adaptive differentiation in resource-use traits (i.e., local adaptation). For example, severe drought stress would be expected to select for greater RMR and WUE in populations experiencing that stress than in populations experiencing less-severe water limitations. Although the relatively low genetic differentiation observed among populations of Porter’s Sunflower suggests that high levels of gene flow (Gevaert et al. 2013) could counteract adaptive divergence among populations (Lenormand 2002, Slatkin 1987), a common-garden experiment showed that disjunct populations of Porter’s Sunflower exhibited genetic differentiation for several traits (seed biomass, height, flower diameter; Mellinger 1972). Additional studies examining water- and nutrient-use traits, and their responses to conditions where resources are limited would shed light on whether varying levels of resource availability across the range of Porter’s Sunflower has resulted in adaptive differentiation among populations. We examined field performance (growth and survival) and resource-use traits of Porter’s Sunflower in 3 populations spanning the species’ range in Georgia. Specifically, we asked: what is the extent of variation in field performance at multiple scales (within and among populations, and among years), and is there evidence of water and/or nutrient limitation as a selective agent in natural populations? In a 3-y field study, we found that the 3 populations differed in both plant performance and resource availability, suggesting the potential for resource availability to drive adaptive differentiation among populations of Porter’s Sunflower. Therefore, we also conducted a common-garden greenhouse study under a fullfactorial combination of high and low levels of water and nutrients to examine whether these populations are genetically differentiated for growth and resourceuse traits in a manner consistent with adaptations for growth and survival with respect to resource availability. Field-site Description Granite outcrops, which result from the exposure of bedrock by erosion, cover over 3200 ha (8000 ac) of land area from Alabama to Virginia, with the majority occurring in Georgia (McVaugh 1943). Plant communities establish in depressions on the rock surface in 4 major successional stages, as proposed by Burbanck and Platt (1964), based on maximum soil depth and vegetation present: the Diamorpha community (soil depth = 2–6 cm), the lichen–annual-herb community (7–15 cm), the annual–perennial-herb community (16–39 cm), and the perennial–shrub community (40–50 cm). Soils at these sites are generally sandy, and accumulate greater organic matter content and cation-exchange capacity as soil depth increases (Braun 1969, Burbanck and Platt 1964, Shure and Ragsdale 1977). Plant communities are often only several meters in length and entirely surrounded by rock, forming “islands” within the granite outcrops (Burbanck and Platt 1964). Although Southeastern Naturalist A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 470 these communities sometimes form in drainage paths that receive runoff water from higher elevations on the rock surface (McCormick et al. 1974), the shallow soils generally dry quickly between precipitation events, leading to frequent, severe drought-stress (Cumming 1969, Houle and Phillips 1989, McCormick 1959, Sharitz and McCormick 1973, Shure and Ragsdale 1977). We conducted field observations of 3 Porter’s Sunflower populations located on granite outcrops in northern Georgia: Camp Meeting Rock (CMR; 33°14'51''N, 8°58'4''W), Panola Mountain State Park (PM; 33°38'9''N, 84°10'1''W), and Heggie’s Rock (HR; 33°32'36''N, 82°16'1''W). We chose these sites because they span the range of Porter’s Sunflower in Georgia, are protected either by the state of Georgia (PM; Georgia Department of Natural Resources) or by the Nature Conservancy (CMR and HR; Atlanta, GA), and include both relatively wet and dry microhabitats within each population. We classified habitats as either wet or dry based on the presence or absence of standing or flowing water when plots were established. The distinction between wet and dry habitats was consistent across all 3 years of the study. The wet habitats generally included Selaginella rupestris (L.) Spring (Northern Selaginella) and various lichens. Dry habitats generally included Andropogon virginicus L. (Broomsedge Bluestem), Opuntia humifosa (Raf.) Raf. (Eastern Prickly Pear), Yucca filamentosa L. (Adam’s Needle or Spanish Bayonet), and Hypericum gentianoides (L.) Britton, Sterns. & Poggenb. (Orangegrass). Both wet and dry habitats included Liatris microcephala (Small) K. Schum. (Smallhead Blazing Star), Oenothera fruticosa L. (Narrowleaf Evening Primrose), Nuttallanthus canadensis (L.) D.A. Sutton (Canada Toadflax), and Tradescantia hirsuticaulis Small (Hairystem Spiderwort). The 3 Porter’s Sunflower populations are separated by at least 100 km, forming an west–east transect (CMR is the western-most, and HR is the eastern-most) across northern Georgia. Each of the 3 populations had more than 1000 Porter’s Sunflower individuals. Methods Field study: Study design and site resource-availability We conducted field observations at the 3 populations from 2008 to 2010 in 2 habitats within each population (wet and dry) with 6 replicate plots (1 m x 2 m) within each habitat. In spring 2008, we classified habitats as relatively wet or dry, and all plots were separated by at least 20 m of bare rock, with no plots set along the forest margins. Although density varied, there were at least 150 Porter’s Sunflower seedlings in all plots. Each spring, we randomly selected and tagged 20 seedlings with 4–6 true leaves in each plot (28–29 March in 2008, 27–29 March in 2009, and 17–19 April 2010). We averaged all within-plot measurements (described below) before further analyses, and we conducted all statistical analyses in SAS (v. 9.2, SAS Institute Inc., Cary, NC). Data were transformed as needed to better-approximate statistical assumptions of normality and homogeneity of variance. We obtained precipitation data for each population from the weather station nearest to each population. For CMR, the long-term average (1904–2010) was from Carrollton, GA (station 091640, 5 km from the site) and the 2008–2010 data were Southeastern Naturalist 471 A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 from Roopville, GA (Georgia Automated Environmental Monitoring Network, 24 km from the site). For PM, the long-term average (1940–2010) was from Jonesboro, GA (station 094700, 18 km from the site) and the 2008–2010 data were an average of available data from station 094700 and a Georgia Automated Environmental Monitoring Network also in Jonesboro, GA. For HR, we obtained both the long-term average (1961–2010) and the 2008–2010 data from Appling, GA (station 090311, 3 km from the site). We measured soil depth for each plot in March 2008. In May 2010, we collected 3 soil samples at a depth of 6–11 cm (depending on plot depth) from each plot and combined them to create a single sample. The samples were dried at 60 °C, ground to fine powder, and analyzed for nitrogen (N) and carbon (C) via Micro-Dumas combustion (NA1500 C/H/N Analyzer, Carlo Erba Strumentazione, Milan, Itlay) at the Stable Isotope Laboratory, University of Georgia, Athens, GA. Soils were analyzed for pH, available P, and extractable Ca, K, Mg, and Mn at the UGA Agricultural and Environmental Services Laboratory, Athens, GA. We analyzed soil data using a two-way ANOVA with population (CMR, PM, HR), habitat (wet, dry), and their interaction as explanatory variables. Field study: Plant measurements We made monthly assessments of height and survival of tagged plants in each plot from May to September. Beginning on 6 September 2008, 5 September 2009, and 15 September 2010, we visited plots twice a week to record the date when the first flower appeared on each surviving plant. We considered the flowers to be open when the first ray flower ligule of the first flowering head was fully expanded. We measured predawn leaf water-potential (Ψpd; MPa) with a pressure chamber (PMS Instrument Company, Albany, OR) 2–3 h before sunrise to estimate soil-water availability (Donovan et al. 2001, Ritchie and Hinckley 1975). Although nighttime transpiration and other processes can affect Ψpd, the extent is usually small for Helianthus (sunflowers), particularly when soil water is limiting (Donovan et al. 2001, Howard and Donovan 2007, Ludwig et al. 2006). W e d e t e r m i n e d Ψpd for 2 individual plants either located in, or within 0.5 m of, each plot in May, July, and September each year of the study. It was not possible to sample all 3 populations on the same day; thus, we sampled populations on 3 consecutive days during an interval with no precipitation (an average of 3 days after a measurable rainfall (≥0.5 in). For each plot and year, we designated the most negative Ψpd value for May or July (the main growing season) as MinΨpd. We collected the leaves assessed for Ψpd in May 2009 and May 2010. Each year, we bulked the collected leaves by plot. The samples were then, dried at 60 °C, ground to fine powder, and analyzed for leaf N, C, and δ13C using Micro-Dumas combustion followed by isotope-ratio mass-spectrometry (Delta V with Conflo III interface, Thermo Scientific, Bremen, Germany) at the Stable Isotope Laboratory. We analyzed all plant measurements by three-way ANOVA, with population, habitat, year, and their interactions as explanatory variables, and then conducted a Tukey test for multiple comparisons. In addition, outliers were assessed using Cook’s D (Cook 1977). We considered as outliers any data points that exhibited Cook’s D values considerably higher than the remaining values (Quinn and Keough 2002), which we identified here as any Southeastern Naturalist A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 472 data point(s) whose Cook’s D value was at least 2-fold greater than the majority of values. Results are presented both including and excluding these outliers. Greenhouse study: Study design In 2009, we collected and stored at 4 °C seeds from both the wet and dry plots of HR, PM, and CMR. We weighed 10 seeds per maternal plant to obtain an estimate of seed mass to account for potential maternal effects during statistical analyses. On 30 September 2009, we placed the seeds on moist filter paper in Petri dishes and kept them in the dark. After 24 h, we scarified the seeds by excising the blunt end of each seed, and we removed seed coats after an additional 24 h. When the hypocotyl had formed and root hairs were present, we placed the Petri dishes under florescent light (80–90 μmol m-2 s -1) with a 12-h photoperiod. After 1 week, we transplanted 40 seedlings per population into 3.8-L pots containing 3:1 sand:Turface (Profile Products, Buffalo Grove, IL) in the UGA Plant Biology greenhouses. We placed the pots in a randomized complete-block design with 3 blocks; seedlings were misted with water twice daily for 1 week, then watered to field capacity with nutrient solution (Peter’s Plant Starter 9-45-15, 100 ppm, Scotts Company, Marysville, OH) 3 times per week thereafter. We provided supplemental lighting to simulate the progression of photoperiod experienced by native populations. Greenhouse study: Seedling maximum relative growth rate (RGRmax) At 4 and 7 weeks following scarification (28 October 2009 and 18 November 2009, respectively), we harvested 8 randomly selected plants from each population for determination of seedling maximum relative growth rate (RGRmax). Due to seedling mortality in the CMR population, the 2 harvests included 7 and 5 individuals, respectively, for that population. We dried the biomass samples at 60 °C, weighed them, and calculated RGRmax (Hunt 1990). We evaluated population differences in seedling RGRmax as the population x harvest interaction term of a two-way ANOVA, with log-transformed plant biomass as the dependent variable (Poorter and Lewis 1986). Greenhouse study: Response to resource treatments Following the second RGRmax harvest, we randomly assigned the remaining plants to 1 of 4 resource treatments (5–6 plants per treatment per population). The well-watered high-nutrient (HWHN) and well-watered low-nutrient (HWLN) plants received 20 g and 1 g slow-release fertilizer (Osmocote Classic 8–9 month release, Everris US Ltd, Marysville, OH), respectively, and were irrigated daily to field capacity (35% soil moisture). The water-limited high-nutrient (LWHN) and water-limited low-nutrient (LWLN) plants received 20 g and 1 g slow-release fertilizer, respectively, but we withheld water until at least half of the pots measured 15% soil moisture at which time we then watered them to field capacity. We used a soil-moisture sensor (ML2X ThetaProbe, Dynamax, Houston, TX) to take daily measurements for all pots in the LW treatments initially, then half of the pots after the first 6 weeks of treatments. During weeks 1–5, 6–9, 10–13 and 14–15 of the resource treatments, the dry-down of soil moisture from field capacity to 15% occurred over Southeastern Naturalist 473 A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 an average of 5, 4, 3, and 2 days, respectively. We did not observe population differences in the rate of dry-down until the final weeks of treatment prior to harvest. At the conclusion of the very last 2-d period of dry-down, plants at PM had lower soil-moisture (8.59% ± 0.32; LS-Means ± 1 SE) than HR (9.83% ± 0.32), though neither differed significantly from CMR (9.05% ± 0.35). From 20–21 December 2009 (week 5 of the resource treatments), we assessed gas exchange and leaf traits during days 3 and 4 of a dry-down (prior to re-watering on day 5). We measured leaf photosynthetic rate per unit area (Aarea) and stomatal conductance (g) with a LI-6400 portable photosynthesis system (Licor Biosciences, Lincoln, NE) on the most recently fully expanded leaf produced after initiation of the resource treatments. Chamber conditions were 1500 μmol m-2 s-1 photosynthetically active radiation and 380 ppm CO2, with block temperature and relative humidity adjusted to ambient conditions on the date measurements were taken (27 °C, 52–57%, respectively). We loosely tied a string around the leaf opposite the gas-exchange leaf, monitored leaf senescence every other day, and recorded when a marked leaf reached 25% yellow as a measure of leaf lifespan. Following gas-exchange measurements, we imaged the gas-exchange leaf, calculated leaf area inside the chamber (to calculate Aarea for leaves smaller than the chamber) and total leaf area, then dried at 60 °C and weighed the samples. We calculated specific leaf-area (leaf area mass-1) for the gas-exchange leaf and used that value to calculate photosynthetic rate per unit mass (Amass) from Aarea and leaf N per unit area (Narea) from leaf N per unit mass. We assessed leaf-level WUE at 2 time scales: instantaneous WUE, (calculated as Aarea/g), and WUE integrated over the lifetime of the leaf (estimated by leaf carbon-isotope ratio; δ13C; Donovan and Ehleringer 1994, Farquhar et al. 1989). We calculated photosynthetic nitrogen-use efficiency (PNUE) as Aarea/Narea. For each plant, we recorded days to first bud and first flower. We harvested plants on 5–9 March 2010, ~4 weeks after the last plant in the experiment flowered; biomass was sorted into aboveground (leaves and stems), belowground (roots), and reproductive components (buds, flowers, and senesced flower heads), dried at 60 °C, and weighed. We employed two-way ANOVA, with population, treatment, their interaction, and block as explanatory variables followed by Tukey tests for multiple comparisons to assess differences in traits and biomass variables. We included initial seed mass as a covariate to account for potential maternal effects, and significant ANOVA effects were largely congruent whether or not the covariate was included. In addition, we evaluated outliers in the analyses of each trait and biomass variable using Cook’s D as described above. Results Field study: Precipitation and soil characteristics The 3 populations differed slightly in long-term mean annual precipitation, forming a west–east gradient of decreasing precipitation across the range of Porter’s Sunflower in northern Georgia (CMR > PM > HR; Fig. 1). However, these differences in long-term means were small compared to temporal variation in precipitation within each population, both within and among the 3 study years. In Southeastern Naturalist A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 474 general, 2008 precipitation was below average, particularly for the HR population during the main growing interval (April–July). Annual precipitation in 2009 was above average for all populations, with peaks in spring and autumn. In 2010, annual precipitation was below the long-term average, but with above-average precipitation in May for both the CMR and PM sites (Fig. 1). Soil depth averaged 10.75 cm (± 0.42) and did not differ among populations or between habitats within populations, although there was a trend for soils to be deeper in wet habitats (P = 0.08; Table 1). Soil-nutrient availability differed among populations; PM generally had the highest overall nutrient availability (Table 1). For soil N, P, K, and Mg, the PM population had among the highest values, but soil P was also high for HR, and soil K and Mg levels were high for CMR. In addition, soil pH was significantly lower for the PM population than the HR population (Table 1). Figure 1. Monthly precipitation at the 3 Helianthus porteri (Porter’s Sunflower) field sites (CMR, PM, HR) in Georgia, 2008–2010 (black line), in relation to the long-term mean (gray area). The long-term annual mean is listed with the site name, and the mean for each year of the study is listed in the bottom corner of each panel. Southeastern Naturalist 475 A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 Table 1. Soil depth, nitrogen (N), carbon (C), available phosphorus (P); exchangeable potassium (K), calcium (Ca), magnesium (Mg), and manganese (Mn); pH means (± SE) and F-values for soils collected at wet and dry microhabitats in 3 Helianthus porteri (Porter’s Sunflower) field sites (CMR, PM, HR) in 2010. Values with different letters within the same row indicate a significant statis tical difference. * indicates P < 0.05. Soil-nutrient availability F-value Population Habitat Pop x Habitat Soil variables CMR wet CMR dry PM wet PM dry HR wet HR dry (F2, 30) (F1, 30) (F2, 30) Soil depth (cm) 11.3 ± 0.9 9.9 ± 1.0 12.4 ± 1.1 9.5 ± 1.0 10.8 ± 0.5 10.6 ± 1.4 0.1 3.2 0.9 Soil N (%) 0.56 ± 0.16C 0.63 ± 0.16BC 1.19 ± 0.14A 0.92 ± 0.14AB 0.63 ± 0.10BC 0.47 ± 0.1C 7.8* 1.1 0.8 Soil C (%) 11.2 ± 4.4B 13.1 ± 3.4AB 16.2 ± 0.8A 15.8 ± 2.8A 7.8 ± 1.3B 6.8 ± 1.4B 6.6* 0.1 0.7 P (kg/ha) 14.0 ± 1.2B 17.1 ± 3.0B 29.4 ± 5.3A 29.9 ± 3.1A 30.0 ± 6.3A 33.7 ± 3.7A 9.2* 0.5 0.1 K (kg/ha) 51.6 ± 6.1ABC 52.8 ± 11.7ABC 67.3 ± 8.2A 55.9 ± 3.8AB 39.9 ± 1.1BC 40.8 ± 6.2C 5.8* 0.5 0.2 Ca (kg/ha) 292.7 ± 117.0 93.3 ± 22.1 146.5 ± 31.2 148.4 ± 25.1 98.4 ± 21.9 65.5 ± 12.7 2.5 2.3 0.8 Mg (kg/ha) 27.5 ± 7.3AB 22.1 ± 6.2ABC 29.7 ± 5.2A 20.6 ± 1.7ABC 16.7 ± 1.5BC 14.1 ± 2.5C 3.6* 2.4 0.1 Mn (kg/ha) 43.3 ± 18.5 8.1 ± 3.3 35.0 ± 13.9 19.6 ± 5.7 21.5 ± 7.6 37.1 ± 10.0 0.1 1.7 2.7 pH 4.8 ± 0.2A 4.1±0.1C 4.4 ± 0.1B 4.4 ± 0.1B 4.8 ± 0.1A 4.8 ± 0.1A 9.3* 6.2* 8.6* Southeastern Naturalist A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 476 Field study: Plant performance Plant survival to first flower differed by population, habitat, and year, with an interaction between population and year (Table 2). The driest year of the study, 2008, was associated with the most dramatic effect on survival to first flower; there was 0% survival to flowering at HR in both the wet and dry habitats (Fig. 2a). In that same year, survival declined more slowly at CMR and PM, and was ~20% at first flower in late summer. In 2009, the wettest year, all of the populations and habitats had a similar seasonal profile of declining survival to 20–50% at first flower (Fig. 2b). In 2010, plants at PM tended towards higher survival through the latter part of the growing season (40–60%) as compared to CMR and HR (15–30%; Fig. 2c). When survival differed between habitats within populations, wet habitats had higher survival than dry habitats (Table 2, Fig. 2a–c). The surviving plants increased in height as the growing season progressed in all years and for all populations and habitats (Fig. 2d–f), and plant height in September differed by population, habitat, and year (Table 2). Across years, PM plants tended to be among the tallest, and HR plants tended to be among the shortest (Fig. 2d–f). Phenology of surviving plants, as indicated by date of first flower, differed by population, habitat, and year, with a significant population x year interaction (Table 2). The most prominent effect was that all 3 populations initiated flowering ~10–15 d later in 2010 compared to 2008 and 2009 (Fig. 2a–c). Leaf N differed by population and year with a significant interaction between population, habitat, and year, while integrated leaf-level WUE (estimated by leaf δ13C) did not differ by population, habitat, or year (Table 2). Seasonal patterns of soil-water availability, estimated with Ψpd, differed by population and year (Fig. 2g–i). In 2008, the HR Ψpd was already relatively low in May, and we documented 0% survival by the July sampling date. MinΨpd, the most negative value observed in the main growth interval (May through July), differed by population, habitat, and year, and had a significant population x year interaction (Table 2, Fig. 2g–i). Plants at HR exhibited lower MinΨpd than those at PM in all 3 years (2008: t90 = 3.90, P = 0.0055; 2009: t90 = 7.29, P < 0.0001; 2010: t90 = 4.18, P = Table 2. F-values for Helianthus porteri (Porter’s Sunflower) plant traits sampled in the field from 2008 to 2010, except leaf traits, which were sampled only in 2009 and 2010. Degrees of freedom for the numerator (ndf) and denominator (ddf) are as follows: Survival to first flower: ndf = 2,1,2,2,4,2,4, ddf = 90; September height: ndf = 2,1,2,2,4,2,4, ddf = 78; MinΨpd: ndf = 2,1,2,2,4,2,4, ddf = 89; Date of first flower: ndf = 2,1,2,2,3,2,3, ddf = 76–78); Leaf N and δ13C ndf = 2,1,1,2,2,1,2), ddf = 59. * indicates P < 0.05. § indicates became significant when 2 outliers were exclud ed. F-values Pop x Pop x Habitat x Pop x Plant trait Population Habitat Year Habitat Year Year Habitat x Year Survival to first flower 36.6* 19.1* 25.6* 0.2 14.2* 2.9 0.2 Sept. Height 9.1* 20.3* 5.7* 2.1 0.8 1.1 0.2 MinΨpd 39.5* 62.7* 6.9* 2.0§ 2.6* 0.8 0.9 Date of first flower 46.6* 24.8* 24.5* 2.2 8.6* 0.2 0.7 Leaf N 11.4* 1.9 39.6* 3.8* 8.3* 1.1 5.8* Leaf δ13C 1.4 0.1 0.9 0.6 1.8 1.5 0.3 Southeastern Naturalist 477 A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 Figure 2. Helianthus porteri (Porter’s Sunflower): (a–c) survival, (d-f) plant height, and (g-i) predawn water-potential assessed at 3 field sites (CMR, PM, HR) in Georgia on sampling dates (number of days following 1 January) in 2008–2010. In panels (a-c) the detached points show survival at flowering. HR data was only available for the first 2 sampling dates in panel (d) and the first sampling date in panel (g) due to 0% survival in that population. Southeastern Naturalist A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 478 0.0021). When we pooled data from all populations, habitats, and years, MinΨpd was positively correlated with survival to first flowering (r = 0.33, P < 0.001, n = 108, Fig. 3). Greenhouse study: Seedling RGRmax, and growth and resource-use traits in response to resource treatments Under non-limiting water and nutrient conditions, the 3 populations did not differ in maximum seedling relative growth rate (RGRmax), as indicated by a non-significant population x harvest interaction (F2,36 = 0.02, P = 0.98). However, the populations differed in nearly all growth and resource-use traits following the resource-limited treatments. The PM population had greater height (t53 = 4.59, P < 0.0001), stem diameter (t53 = 6.26, P < 0.0001), aboveground biomass (t53 = 3.75, P = 0.0013), reproductive biomass (t53 = 4.06, P = 0.0005), root biomass (t53 = 4.22, P = 0.0003), and total biomass (t53 = 4.05, P = 0.0005), but lower Aarea (t53 = 5.54, P < 0.0001) than plants at HR (Figs. 4, 5; Table 3). In addition, the PM population reached first flower an average of 8.7 d later than HR (t53 = 2.53, P = 0.0376; Fig. 5). However, seed mass was a significant covariate for days to first flower (F1,53 = 4.73, P = 0.0340), and population-level differences in days to first flower were not significant when seed mass was removed as a covariate in ANOVA. For all traits that significantly differed among populations, the Figure 3. Helianthus porteri (Porter’s Sunflower) survival to first flower as a function of minimum predawn plant-water-potential (MinΨpd) from May and July assessed in 3 field sites in Georgia from 2008 to 2010. Southeastern Naturalist 479 A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 differences generally remained consistent regardless of resource treatments, as indicated by the absence of significant population by treatment interaction terms (Table 3). Figure 4. Plant growth and biomass traits (mean ± SE) for 3 populations of Helianthus porteri (Porter’s Sunflower) (CMR, PM, HR) grown in the greenhouse under treatments corresponding to different levels of resource availability: high water and high nutrients (HWHN), high water and low nutrients (HWLN), low water and high nutrients (LWHN), and low water and low nutrients (LWLN). Southeastern Naturalist A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 480 Discussion In comparisons of the 3 Porter’s Sunflower populations across 3 years in the field, the PM population generally performed the best, with among the highest rates of survival in both of the relatively dry years (2008 and 2010), and the tallest plants Figure 5. Mean (± SE) for resource-use traits (leaf photosynthetic rate per unit area [Aarea], specific leaf-area [SLA], water-use efficiency [WUE], photosynthetic nitrogen-use efficiency [PNUE]) for 3 populations of Helianthus porteri (Porter’s Sunflower) (CMR, PM, HR) grown in the greenhouse under 4 treatments corresponding to different levels of resource availability: high water and high nutrients (HWHN), high water and low nutrients (HWLN), low water and high nutrients (LWHN), and low water and low nutrients (LWLN). Southeastern Naturalist 481 A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 in the relatively wet year (2009). Conversely, the HR population had 100% mortality prior to reproduction in one of the drier years (2008), and had some the shortest plants during even the relatively wet year (2009). These findings are consistent with the higher soil-moisture and organic matter reported in communities at the PM site than at other granite outcrops of the southeastern US (Braun 1969, Burbank and Platt 1964, Mellinger 1972, Shure and Ragsdale 1977), and with the lower (more negative) MinΨpd found at HR compared to PM in the present study. We also found that survival to first flower was associated with greater (less negative) MinΨpd during the main growth interval, suggesting that the ability to avoid drought confers a fitness advantage in Porter’s Sunflower. Given that drought-induced mortality varied across populations, we hypothesized that drought is a selective agent driving adaptive differentiation in populations of Porter’s Sunflower for resource-use traits, and that the HR population should exhibit traits consistent with greater avoidance of drought and low nutrient-stress than PM and CMR. We tested these hypotheses by comparing resource-use traits of the 3 populations in a common-garden greenhouse study. Consistent with the findings of Mellinger (1972), we found differences among the populations that can be attributed to genetic differentiation; however, we obtained little evidence that the HR population, which exhibited the greatest drought-induced mortality of the 3 populations in the field, exhibited traits consistent with greater avoidance of resource limitation (i.e., maximizing resource uptake and/or minimizing loss). Compared to the PM and CMR populations, plants at HR did not exhibit greater RMR, PNUE, WUE, or Table 3. F-values for traits assessed in a greenhouse resource-limitation study of 3 populations of Helianthus porteri (Porter’s Sunflower). Resource treatments consisted of a full factorial combination of high or low abundance of water and high or low levels of nutrients (see main text for details). * indicates P < 0.05. F-values Effects Population (F2, 67) Treatment (F3, 67) Pop x treatment (F6, 37) Plant height 11.57* 11.78* 1.64 Stem diameter 19.68* 11.58* 1.58 Aboveground biomass 7.69* 27.92* 1.06 Reproductive biomass 8.77* 32.09* 0.38 Root biomass 9.87* 3.92* 0.98 Root:total biomass 4.02* 7.05* 0.39 Total biomass 8.41* 29.19* 0.80 Days to first bud 3.11 8.16* 2.23 Days to first flower 3.21* 7.91* 2.01 Aarea 17.36* 3.69* 0.75 Amass 4.04* 1.44 1.05 g 2.69 4.66* 0.70 SLA 5.78* 0.52 0.67 Leaf N 3.75* 23.41* 1.77 PNUE 1.19 3.44* 1.28 Instantaneous WUE (A/g) 0.29 8.17* 0.24 Integrated WUE (δ13C) 2.23 0.78 0.49 Leaf lifespan 0.86 49.02* 0.70 Southeastern Naturalist A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 482 lower seedling RGRmax, traits which are typically associated with greater adaptation to resource-limited environments (Aerts and Chapin 1999, Chapin 1980, Lambers and Poorter 1992, Poorter and Remkes 1990). Although HR exhibited lower SLA than PM and CMR, consistent with the expectation that structurally robust tissues that confer greater resource conservation provide an advantage in resource-poor habitats (Chapin 1980, Cunningham et al. 1999), this difference was not statistically significant when we excluded 2 outliers. In contrast to the lack of differentiation for traits associated with drought avoidance, the HR population reached first flower an average of 9 d sooner than plants at PM in the common-garden study. Early reproduction may reduce exposure to drought (i.e., drought escape), which may be advantageous in the more droughtstressed HR site. Numerous studies have found evidence for selection towards earlier flowering in response to stress in annual plants (Brouillette et al. 2014, Franks et al. 2007, Griffith and Watson 2005, Heschel and Riginos 2005, Ivey and Carr 2012). However, this difference among populations in days to first flower was only statistically significant when we included initial seed mass (a proxy for maternal effects) as a covariate in the analysis. The inclusion of seed mass may have allowed us to detect differences in flowering among populations because it truly accounted for the variation due to environmentally induced maternal effects (i.e., variation in seed size due to greater resource availability in the maternal environment). The HR and PM populations significantly differed in initial seed mass (t65=3.53, P = 0.0022), suggesting that environmentally induced maternal effects may have differed between these 2 populations and impacted their flowering phenologies. Alternatively, the populations may be genetically differentiated for seed mass at least partially independent of maternal effects because seed size is influenced by genetic factors in addition to the possibility of environmentally driven maternal effects (Alonso-Blanco et al. 1999). Thus, the difference among populations in days to first flower that we detected only when seed mass was included as a covariate should be interpreted with caution because it could potentially reflect a statistical artifact rather than a true difference in flowering time among populations. We recommend additional common-garden studies established directly on the outcrops to assess whether the flowering phenology and fitness of these populations differ in outcrop conditions, and the result of that work would shed light on the potential for adaptive differentiation in drought escape among these populations. Given the extreme fluctuations in resource availability that characterize granite outcrops (McCormick and Platt 1964, Shure and Ragsdale 1977), we also assessed whether the 3 populations differed in their response to resource (water, nutrients, or both) manipulations, which could be suggestive of local adaptation to resource heterogeneity. For example, a common-garden study showed that ecotypes of Diamorpha cymosa (Nutt.) Britton ex Small (Elf Orpine) collected from 16 outcrops from Alabama to North Carolina differed in their responses to changes in water availability, light intensity, and temperature (McCormick and Platt 1964), leading the authors of that study to suggest that differential heterogeneity of resource availability in the “home outcrops” of these ecotypes could be driving Southeastern Naturalist 483 A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 adaptive differentiation for responses to resource availability. In contrast, although trait responses to resource-limitation treatments in the present study were generally in the direction expected (plants generally exhibited lower overall-growth under low-resource conditions, with higher instantaneous WUE and shorter leaf-lifespan associated with limited available water, and lower leaf N associated with low levels of nutrient availability), the magnitude of these responses did not differ among the 3 Porter’s Sunflower populations. Surprisingly, although plants in the water-limited treatments had higher instantaneous WUE than those in well-watered treatments as expected, integrated WUE (estimated as δ13C) did not differ among the resource treatments. We measured both instantaneous and integrated WUE on the most recently fully expanded leaf that was produced after treatment conditions had been initiated. However, we measured gas-exchange measurements during a dry-down cycle, when stomata were likely closing in response to water limitation, resulting in higher instantaneous WUE in the water-limited treatments. Because δ13C is an integrated measure of WUE, it accounts for all of the carbon assimilated during the leaf’s lifetime (Farquhar et al. 1989). Thus, that we did not detect a resource-limitation treatment-effect for leaf δ13C may reflect that most of the carbon assimilation was taking place when soils were near field capacity and stomata were relatively open (i.e., when plants were re-watered between drought treatments). For all traits that responded to resource manipulations, all 3 Porter’s Sunflower populations had responses that were similar in magnitude, suggesting the populations are not differentiated in their capacity to exploit resource heterogeneity. An important consideration of our resource-limitation treatments is that, although the stress treatments significantly reduced all biomass components and influenced nearly all traits in Porter’s Sunflower, the treatments did not cause the degree of stress that the plants would experience in situ—mortality occurred in the field, but not in the greenhouse. In addition, moderate sample sizes (n = 5–6) may have precluded our ability to detect significant differences among populations. Further studies with more extensive sampling of plant responses to a range of stress severities could reveal differentiation among the populations not detected here. Although we demonstrated that drought is an important factor influencing survival in Porter’s Sunflower, we found little evidence to support the hypothesis that populations are adaptively differentiated in a manner consistent with the resource availability documented in the field. This conclusion is in agreement with the relatively low degree of genetic structure reported for Porter’s Sunflower (Gevaert et al. 2013) which suggests that historical gene flow has homogenized populations and limited the potential for local adaptation (Lenormand 2002, Slatkin 1987). The genetic differentiation among populations for traits assessed here and by Mellinger (1972) may reflect population divergence due to genetic drift or selective pressures other than resource availability. The explanation of genetic drift seems unlikely, given that population sizes were relatively large (thousands of individuals), and population divergence by genetic drift should be counteracted by high gene-flow (Lenormand 2002, Slatkin 1987). Future studies are needed to Southeastern Naturalist A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 484 better understand the factors influencing productivity and survival in Porter’s Sunflower and possible selective pressures driving population differentiation in this granite outcrop endemic. Acknowledgments We thank N. Sherman and M. McKain for assistance with data collection during the first year of the study, and M. Hodges (The Nature Conservancy) and T. Patrick and N. Castleberry (Georgia Department of Natural Resources) for help in acquiring and maintaining permission for the use of the field sites. We are grateful to numerous people who helped during the final harvest, including E. Milton, C. Mason, K. Bettinger, R. Shirk, C. Deen, C. Gormally, B. Foltz, R. Rogers, J. Rabanal, C. Allen, and P. Nowicki, without whose generosity this research would not have been possible. We appreciate the UGA Plant Biology greenhouse staff for assistance in plant germination and caring for the plants during the experiment, and 2 anonymous reviewers whose comments greatly improved this manuscript. This work was funded by the UGA Department of Plant Biology (S.D. Gevaert) and National Science Foundation grants 0614739 and 1122842 (L.A. Donovan ). Literature Cited Aerts, R., and F.S. Chapin III. 1999. The mineral nutrition of wild plants revisited: A reevaluation of processes and patterns. Advances in Ecological Research 30:1–67. Alonso-Blanco, C., H. Blankestijn-de Vries, C.J. Hanhart, and M. Koornneef. 1999. Natural allelic variation at seed-size loci in relation to other life-history traits of Arabidopsis thaliana. Proceedings of the National Academy of Sciences of the United States of America 96:4710–4717. Braun, D.G. 1969. Soil factors and sub-annual nutrient cycling in two types of graniteoutcrop soil-island ecosystems. M.Sc. Thesis. Emory University, Atlanta, GA. 178 pp. Brouillette, L.C., C.M. Mason, R.Y. Shirk, and L.A. Donovan. 2014. Adaptive differentiation of traits related to resource use in a desert annual along a resource gradient. 201:1316–1327. Burbanck, M.P., and D.L. Phillips. 1983. Evidence of plant succession on granite outcrops of the Georgia Piedmont. American Midland Naturalist 109:94–104. Burbanck, M.P., and R.B. Platt. 1964. Granite-outcrop communities of the Piedmont Plateau in Georgia. Ecology 45:292–306. Chapin, F.S., III. 1980. The mineral nutrition of wild plants. Annual Reviews in Ecology and Systematics 11:233–260. Chapin, F.S., III. 1991. Effects of multiple environmental stresses on nutrient availability and use. Pp. 67–88, In H.A. Mooney, W.E. Winner, and E.J. Pell (Eds.). Response of Plants to Multiple Stresses. Academic Press, San Diego, CA. 422 pp. Chapman, R.H., and S.B. Jones. 1975. Ecotypic differentiation in Andropogon virginicus (Gramineae). Bulletin of the Torrey Botanical Club 102:166–171. Cook, R.D. 1977. Detection of influential observations in linear regression. Technometrics 19:15–18. Cumming, F.P. 1969. An experimental design for the analysis of community structure. M.A. Thesis. University of North Carolina at Chapel Hill, Chapel Hill, NC. 28 pp. Cunningham, S.A., B. Summerhayes, and M. Westoby. 1999. Evolutionary divergences in leaf structure and chemistry: Comparing rainfall and soil-nutrient gradients. Ecology 69:569–588. Southeastern Naturalist 485 A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 Donovan, L.A., and J.R. Ehleringer. 1994. Potential for selection on plants for water-use efficiency as estimated by carbon-isotope discrimination. American Journal of Botany 81:927–935. Donovan, L.A., M.J. Linton, and J.H. Richards. 2001. Predawn water potential does not necessarily equilibrate with soil water-potential under well-watered conditions. Oecologia 129:328–335. Farquhar, G.D., J.R. Ehleringer, and K.T. Hubick. 1989. Carbon-isotope discrimination and photosynthesis. Annual Review of Plant Physiology and Plant Molecular Biology 40:503–537. Fierer, N., and J.P. Schimel. 2002. Effects of drying–rewetting frequency on soil carbon and nitrogen transformations. Soil Biology and Biochemistry 34:777–787. Franks, S.J., S. Sim, and A.E. Weis. 2007. Rapid evolution of flowering time by an annual plant in response to a climate fluctuation. Proceedings of the National Academy of Sciences of the United States of America 104:1278–1282. Funk, J.L., and P.M. Vitousek. 2007. Resource-use efficiency and plant invasion in lowresource systems. Nature 446:1079–1081. Gevaert, S.D., J.R. Mandel, J.M. Burke, and L.A. Donovan. 2013. High genetic-diversity and low population-structure in Porter’s Sunflower (Helianthus porteri). Journal of Heredity 104:407–415. Griffith, T.M., and M.A. Watson. 2005. Stress avoidance in a common annual: Reproductive timing is important for local adaptation and geographic distribution. 18:1601–1612. Héroult, A., Y-S Lin, A. Bourne, B.E. Medlyn, and D.S. Ellsworth. 2013. Optimal stomatal conductance in relation to photosynthesis in climatically contrasting Eucalyptus species under drought. Plant, Cell, and Environment 36:262–274. Heschel, M.S., and R. Riginos. 2005. Mechanisms of selection for drought-stress tolerance and avoidance in Impatiens capensis (Balsaminaceae). American Journal of Botany 92:37–44. Houle, G., and D.L. Phillips. 1989. Seasonal variation and annual fluctuation in graniteoutcrop plant communities. Vegetatio 80:25–35. Howard, A.R., and L.A. Donovan. 2007. Helianthus nighttime conductance and transpiration respond to soil water but not nutrient availability. Plant Physiology 143:145–155. Hunt, R. 1990. Basic Growth Analysis: Plant-Growth Analysis for Beginners. Unwin Hyman, London, UK. 112 pp. Ivey, C.T., and D.E. Carr. 2012. Testing for the joint evolution of mating system and drought escape in Mimulus. Annals of Botany 109:583–598. Kartesz, J.T. 2015. The Biota of North America Program (BONAP). North American Plant Atlas. Chapel Hill, NC. Available online at http://bonap.net/napa. Accessed 2 March 2016. Lambers, H., and H. Poorter. 1992. Inherent variation in growth rate between higher plants: A search for physiological causes and ecological consequences. Advances in Ecological Research 23:187–261. Lambers H., F.S. Chapin III, and T.L. Pons. 2008. Plant Physiological Ecology. 2nd Edition. Springer, New York, NY. 605 pp. Lenormand, T. 2002. Gene flow and the limits to natural selection. Trends in Ecology and Evolution 17:183–189. Ludwig, F., R.A. Jewitt, and L.A. Donovan. 2006. Nutrient- and water-addition effects on day- and night-time conductance and transpiration in a C3 desert annual. Oecologia 148:219–225. Lugo, A.E. 1969. Energy, water, and carbon budgets of a granite-outcrop community. Ph.D. Dissertation. University of North Carolina at Chapel Hill, Chap el Hill, NC. 466 pp. Southeastern Naturalist A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 486 Lugo, A.E., and J.F. McCormick. 1981. Influence of environmental stressors upon energy flow in a natural terrestrial ecosystem. Pp. 79–102, In G.W. Barrett and R. Rosenberg (Eds.). Stress Effects on Natural Ecosystems. John Wiley and Sons, New York, NY. 305 pp. McCormick, J.F. 1959. Ecological analysis of selected granite-outcrop communities and their response to chronic gamma irradiation. M.Sc. Thesis. Emory University, Atlanta, GA. 174 pp. McCormick, J.F., and R.B. Platt. 1964. Ecotypic differentiation in Diamorpha cymosa. Botanical Gazette 125:271–279. McCormick, J.F., A.E. Lugo, and R.S. Sharitz. 1974. Experimental analysis of ecosystems. Pp. 148–179, In B.R. Strain and W.D. Billings (Eds.). Handbook of Vegetation Science Part VI: Vegetation and Environment. Dr. W. Junk Publishers, The Hague, Netherlands. 193 pp. McDonald, P.G., C.R. Fonseca, J.M. Overton, and M. Westoby. 2003. Leaf-size divergence along rainfall and soil-nutrient gradients: Is the method of size reduction common among clades? Functional Ecology 17:50–57. McVaugh, R. 1943. The vegetation of the granite flat-rocks of the southeastern United States. Ecological Monographs 13:119–166. Mellinger, A.C. 1972. Ecological life cycle of Viguiera porteri and factors responsible for endemism to granite outcrops of Georgia and Alabama. Ph.D. Dissertation. University of North Carolina at Chapel Hill, Chapel Hill, NC. 214 pp. Poorter, H., and C. Lewis. 1986. Testing differences in relative growth rate: A method avoiding curve-fitting and pairing. Physiology of Plants 67:223–226. Poorter, H., and C. Remkes. 1990. Leaf-area ratio and net-assimilation rate of 24 wild species differing in RGR. Oecologia 83:553–559. Pruski, J.F. 1998. Helianthus porteri (A. Gray) Pruski (Compositae), a new combination validated for Confederate Daisy. Castanea 63:74–75. Quinn, G.P., and M.J. Keough. 2002. Experimental Design and Data Analysis for Biologists. Cambridge University Press, Cambridge, UK. 537 pp. Reich, P.B., M.B. Walters, and D.S. Ellsworth. 1992. Leaf life-span in relation to leaf, plant, and stand characteristics among diverse ecosystems. Ecological Monographs 62:365–392. Reich, P.B., M.B. Walters, and D.S. Ellsworth. 1997. From tropics to tundra: Global convergence in plant functioning. Proceedings of the National Academy of Sciences of the United States of America 94:13,730–13,734. Ritchie, G.A., and T.M. Hinckley. 1975. The pressure chamber as an instrument for ecological research. Advances in Ecological Research 9:165–254. Sambatti, J.M.B., and K.K. Caylor. 2007. When is breeding for drought tolerance optimal if drought is random? New Phytologist 175:70–80. Sanaullah M., C. Rumpel, X. Charrier, and A. Chabbi. 2012. How does drought stress influence the decomposition of plant litter with contrasting quality in a grassland ecosystem? Plant and Soil 352:277–288. Schimel J., T.C. Balser, and M. Wallenstein. 2007. Microbial stress-response physiology and its implications for ecosystem function. Ecology 88:1386–1394 Sharitz, R.R., and J.F. McCormick. 1973. Population dynamics of two competing annual plant species. Ecology 54:723–740. Shelton, L.S.J. 1963. The life history of Viguiera porteri (A. Gray) Blake and factors influencing its endemism to granite outcrops. M.A. Thesis. University of Georgia, Athens, GA. 104 pp. Southeastern Naturalist 487 A.W. Bowsher, S.D. Gevaert, and L.A. Donovan 2016 Vol. 15, No. 3 Shure, D.J., and H.L. Ragsdale. 1977. Patterns of primary succession on granite-outcrop surfaces. Ecology 58:993–1006. Slatkin, M. 1987. Gene flow and the geographic structure of natural populations. Science 15:787–792. Smith, W.K. 1978. Temperatures of desert plants: Another perspective on the adaptability of leaf size. Science 201:614–616. Verslues, P.E., M. Agarwal, S. Katiyar-Agarwal, J. Zhu, and J-K. Zhu. 2006. Methods and concepts in quantifying resistance to drought, salt, and freezing: Abiotic stresses that affect plant water status. The Plant Journal 45:523–539. Wright, I.J., P.B. Reich, M. Westoby, D.D. Ackerly, Z. Baruch, F. Bongers, J. Cavender- Bares, T. Chapin, J.H.C. Cornelissen, D. Diemer, J. Flexas, E. Garnier, P.K. Groom, J. Gulias, K. Hikosaka, B.B. Lamont, T. Lee, W. Lee, C. Lusk, J.J. Midgley, M-L. Navas, U. Niinemets, J. Oleksyn, N. Osada, H. Poorter, P. Poot, L. Prior, V.I. Pyankov, C. Roumet, S.C. Thomas, M.G. Tjoelker, E.J. Veneklaas, and R. Villar. 2004. The worldwide leaf-economics spectrum. Nature 428:821–827. Wyatt, R. 1997. Reproductive ecology of granite-outcrop plants from the southeastern United States. Journal of the Royal Society of Western Australia 80:123–129.