Factors that influence the distribution and growth of white heliotrope (Boraginaceae: Heliotropium tenellum (NUTT.) Torr.).
Heliotropium tenellum (Nutt.) Torr., is found in many parts of the south central United States (United States Department of Agriculture, Natural Resource Conservation Service 2015). It occurs from central Texas east to Georgia, northeast to West Virginia and from Iowa south to the Gulf of Mexico. It seems to be present in high light environments in cedar glades or inter-canopy gaps in central Texas (Van Auken 2000). it appears to be in specific places along spatial gradients of abiotic and biotic factors that changed from under the Juniperus-Quercus canopy into the cedar glades. Spatial gradients have been identified for parts of analogous communities, but did not include the entire glade or surrounding woodland (Van Auken 2000).
Certain species are usually found at particular densities along complex gradients (Begon et al. 2006; Krebs 2009). Factors causing gradients could include light levels, soil moisture, soil type and depth, soil nutrient levels, salinity, competition or some combination of these and other factors (Wilson & Keddy 1986; Woodward 1987; Bush 2006; Smith & Smith 2012). It is relatively simple to determine where a species is found and its density, but the reasons or conditions that cause the pattern are much more difficult to elucidate (Begon et al. 2006; Smith & Smith 2012). The specific factors that determine where H. tenellum is found and govern its growth and development have not been identified.
Central Texas cedar glades where H. tenellum is found are dominated by herbaceous vegetation, mostly annuals with some [C.sub.4] southern grasses and typically low plant cover, relative to the surrounding Juniperus-Quercus woodlands (Terletzky & Van Auken 1996; Van Auken 2000). On fairly level surfaces within the cedar glades, there are bare patches where bedrock reaches the surface and soil is very shallow (Van Auken 2000). These bare patches are a common feature of these cedar glades, but the number or the area covered by these features has never been quantified. The bare patches have a soil depth gradient from low or zero in the center to higher toward the edges and the glades themselves seem to have a highly variable soil depth. The perennial [C.sub.4] southern grasses are in the deeper soil of the glades and the herbaceous annuals are in the shallower soil. Some woody species, succulent species, and nonvascular plants may be present at low density or cover. Herbaceous annual species commonly found within central Texas cedar glades include Heliotropium tenellum (white heliotrope), Centaurium texense (Griseb.) Fern. (Lady Bird's centaury), Evax prolifera DC. (rabbit-tobacco), Croton monanthogynus Michx. (prairie-tea), and Chaetopappa bellidifolia (Gray & Engclm.) Shinners. (dwarfwhite aster) (Van Auken 2000), and there are others (Enquist 1987) as well as various native and non-native C4 grasses.
The purpose of this study was to examine several biotic and abiotic factors that may control the distribution of H. tenellum in central Texas glades and associated woodlands and to determine possible reasons for its presence in these communities. Factors assessed in each quadrat were density of H. tenellum, light level, soil depth (cm), percent cover of H. tenellum, rock, leaf litter, other herbaceous plants, non-canopy woody plants and bare soil. In addition, net photosynthesis, stomatal conductance, transpiration and leaf area of selected H. tenellum plants were measured.
Materials and Methods
Study area.--The study area was a cedar glade in Bandera County in the central Texas Edwards Plateau Region. Cedar glades can be found on well-defined flat hill tops (Fig. 1) or on hillsides (Terletzky & Van Auken 1996). Stair-step topography, where hillside cedar glades alternate with Juniperus-Quercus woodlands, is common within the Balcones Escarpment of the Plateau region (Riskind & Diamond 1988; Terletzky & Van Auken 1996). The soil in central Texas cedar glades is shallow, slightly basic, low in organic matter and high in calcium (Terletzky & Van Auken 1996). The subsurface of this area is mostly bedrock consisting of deeply eroded limestone of the Glen Rose and Edwards Formations although sometimes dolomites are present (Sellards et al. 1932). The vegetation structure appears to be influenced by soil depth and intermittent, low soil moisture, in addition to the type of bedrock, grazing intensity and fire frequency (Kucera & Martin 1957; Terletzky & Van Auken 1996; Van Auken 2000).
Cedar glades, in general, consist of open areas with shallow soils, with deep rock crevices and limestone or dolomite at or near the surface, with plant communities dominated by herbaceous angiosperms and/or cryptogams (Baskin & Baskin 1985; Quarterman et al. 1993). Cedar glades have been reported all across North America (Ware 2002; Schedlbauer & Pistora 2013; Moores 2011) including central Texas (Terletzky & Van Auken 1996; Van Auken 2000). Surface light levels in full sun are high and in summer can be > 2,000 [micro]mol/[m.sup.2]/s (Baskauf 1993). Soil water is frequently below the wilting coefficient, suggesting that plants in these communities experience water stress or have adaptations to compensate for such stress (Baskin and Baskin 1978).
Field Measurements.--On August 16-17, 2010 ten randomly placed transects, 10 m in length, were established from under a Juniperus ashei-Quercus virginiana canopy into the associated glade or intercanopy patch (Van Auken et al. 2005). Transects were placed so 4 m of the transect line was under the canopy and 6 m was in the associated glade with the drip-line between. Ten adjacent one meter by one meter quadrats were placed on each side of the transect line for a total of 20 quadrats per transect and 200 total quadrats.
The density of H. tenellum (number of plants), light level (PAR - [micro]mol/[m.sup.2]/s) and soil depth (cm) were measured in each quadrat. Light level was measured with a LI-COR[R]LI-188 integrating quantum sensor within [+ or -] 60 min of solar noon on clear, cloud-free days. The quantum sensor was leveled just above the soil surface in the center of each quadrat. Soil depth was measured with a piece of rebar (0.9 cm dia. and 36.4 cm long) driven into the soil at the center of each quadrat until it would not penetrate deeper (Larcher 2003). Percent cover of H. tenellum, rock, leaf litter, other herbaceous plants, non-canopy woody plants and bare soil was visually estimated in each quadrat (Van Auken & Bush 1992). Mean values were calculated for each variable per quadrat. Data did not meet assumptions for parametric statistics and transformations did not correct the problem so the van der Waerden non-parametric one-way ANOVA was used (Sail et al. 2012).
Regression Analysis of Transect Data.--Analysis of quadrat data indicated that variables changed from below the woodland canopy into the cedar glade, but not which variable or variables were most important in influencing the density of H. tenellum. Thus, scatter plots between the number of H. tenellum plants in each quadrat (dependent variable), and measured abiotic and biotic factors were constructed. The independent variables were light level, soil depth, and the percent cover of rock, leaf litter, woody plants, other herbaceous plants and bare soil. All variables, except light level, soil depth, and percent cover of rock, had numerous zeros in the data set. To evaluate logarithmic transformations, one was added to all measured values (value + 1). The scatter plots of simple linear regressions were used to determine possible transformations for multiple linear regression analyses. The potential transformations examined were natural logarithm of .Y, square root of x, square of x ([x.sup.2]), reciprocal of x (1/x) and exponential ([e.sup.x]). In order to meet the assumptions for parametric statistics (normally distributed residuals), a logarithmic transformation of soil depth, percent cover of leaf litter, and cover of herbaceous plants was necessary. In addition, a reciprocal transformation of percent rock cover, woody plant cover, and bare soil was necessary. The terms obtained from the simple linear regressions with the highest [R.sup.2] that met the model assumption of normally distributed residuals (Shapiro-Wilk Test) were used in the multiple linear regressions (Sall et al. 2012).
Density classes of H. tenellum were created in units of one plant/[m.sup.2]. The mean light level, soil depth, and the percent cover of rock, leaf litter, woody plants, other herbaceous plants and bare soil was calculated for each density class. There were a total of 69 density classes. The sample size of each class (the number of quadrats with a given density) was used for weighting. A stepwise, weighted, multiple linear regression was completed with the density of H. tenellum in each quadrat as the dependent variable and the mean of the above mentioned factors with the appropriate transformation as the independent variables. The Mallow Cp statistic and [R.sup.2] were used as selective criteria to determine which variables to use and which should be retained in the model (Kleinbaum et al. 1988). The correct reduced model is one where Cp is approximately equal to p. When the correct model is chosen, the parameter estimates are considered unbiased and this is reflected in Cp near p (Daniel & Wood 1980; Mason et al. 1989). Due to collinearity among the independent variables (significant correlation between independent variables), condition indices were examined (Kleinbaum et al. 1988; Belsley et al. 2005). The data was centered, to reduce collinearity, which reduced the condition indices to acceptable limits, as well as reducing potentially incorrect parameter estimates of the regression coefficients, variability, and p-values. It should be noted that centering the data did not change the coefficients of determination (Mason et al. 1989).
Gas Exchange Rates.--Light response curves were generated on July 24, 2009 by measuring gas exchange rates as a function of photosynthetic photon flux density (PAR) for leaves of five H. tenellum plants (five replications). Measurements were made within [+ or -] 3 hr of solar noon with a LI-COR[R]6400 infrared gas analyzer. The light curve program in the LI-CORK6400 was used to generate irradiances with the gas How rate set at 400 [micro]mol/s and C[O.sub.2] concentration set at 400 [micro]mol/mol. The LI-COR[R]6400 was run at approximately mid-late summer, midday temperature (35[degrees]C) and relative humidity (50%) (Gagliardi & Van Auken 2010). One mature, undamaged leaf per replication was used in the 2 x 3 cm chamber. The light response curves were started at a PAR of 2000 [micro]mol/[m.sup.2]/s and then decreased to 1800, 1600, 1200, 1000, 800, 600, 400, 200, 100, 75, 50, 25, 10, 5 and 0 [micro]mol/[m.sup.2]/s. The area of each leaf was measured with a LI-COR[R]3000A portable area meter and used to convert the gas exchange values to standard values of [micro]molC[O.sub.2]/[m.sup.2]/s.
Net photosynthesis, stomatal conductance, and transpiration were measured. A one-way ANOVA was used to determine if gas exchange rates were significantly different between the PAR's measured. A Tukey-Kramer HSD test was then used to determine where the significant differences occurred between the PAR's tested for all three parameters. The Shapiro-Wilk test was used to test for normal distributions and Bartlett's test was used to test for equal variances (Sall et al. 2012). Using PhotosynAssitant[R] software, maximum photosynthesis ([A.sub.max]), PAR at [A.sub.max], transpiration at [A.sub.max], conductance at [A.sub.max], light saturation point ([L.sub.sp]), light compensation point ([L.sub.cp]), dark respiration ([R.sub.d]) and quantum yield efficiency (IS) were determined for each replicate curve and means were calculated.
Water potential.--On July 6, 2010 xylem water potential ([[psi].sub.x]) measurements were made at pre-dawn (5:00-6:00 am) and midday (12:00-1:00 pm) using a Scholander pressure bomb (model 1000 PMS[R]) (Scholander et al. 1965). Sample size was ten H. tenellum plants per treatment (pre-dawn and midday) and [[psi].sub.x] was measured immediately after collection and mean [[psi].sub.x] calculated. The Shapiro-Wilk test was used to test for normal distributions and Bartlett's test was used to test for equal variances (Sall et al. 2012) followed by a one-way ANOVA.
Soil water potential ([[psi].sub.soil]) was measured at pre-dawn and midday on the same day as the xylem water potential with four Watermark[R] soil moisture sensors buried in the cedar glades one week before measurements were made. The sensors were connected to a Watchdog[R] data logger (2 sensors/data logger) that collected measurements every 2 hrs. After making the [[psi].sub.soil] measurements the information was downloaded into Excel' and means were calculated followed by a one-way ANOVA.
Field Measurements-Light levels.--There were significant differences between the mean light levels and quadrat location (Fig. 2A). Light levels were low under the canopy and high in the open (van der Waerden one-way ANOVA; p < 0.0001). Light levels under the canopy were 167 [+ or -] 10 [micro]mol/[m.sup.2]/s (mean [+ or -] 1 SE) and were 1960 [+ or -] 19 ([micro]mol/[m.sup.2]/s in the open glades.
Soil depth- There were significant differences between soil depth and quadrat location (van der Waerden one-way ANOVA; p < 0.0001) (Fig. 2B). Soil was deepest under the canopy (11.2 [+ or -] 0.6 cm) and shallowest in the open (4.8 [+ or -] 0.4 cm).
Density-There were significant differences between the density of H. tenellum and quadrat location (Fig. 2C). Highest density was in the open (48 [+ or -] 3 plants/[m.sup.2]) and lowest below the canopy (3 [+ or -] 1 plants/[m.sup.2]) (van der Waerden one-way ANOVA; p < 0.0001). There were significant differences between H. tenellum cover and quadrat location (van der Waerden one-way ANOVA; p < 0.0001, not shown), with a similar trend as plant density. The mean cover of H. tenellum under the canopy was 2 [+ or -] 0%, while in the open it was 20 [+ or -] 1%. There were significant differences between leaf litter cover and quadrat location (van der Waerden one-way ANOVA; p < 0.0001) (Fig. 2D), with the greatest mean leaf litter cover under the canopy (62 [+ or -] 2%) and very little leaf litter cover in the open (2 [+ or -] 0%).
Other Cover.--There were significant differences between rock cover and quadrat location (not shown). Mean rock cover under the canopy was 22 [+ or -] 1% and in the open it was 30 [+ or -] 1% (van der Waerden one-way ANOVA; p = 0.0002). There were significant differences between non-canopy woody plant cover and quadrat location (van der Waerden one-way ANOVA; p < 0.0001). The greatest mean non woody plant cover was under or within the canopy (4 [+ or -] 1%) and less cover in the open (2 [+ or -] 1%) (not shown). There was no significant difference between other herbaceous plant cover and quadrat location (van der Waerden one-way ANOVA; p = 0.3511) (not shown). Mean cover of other herbaceous plants below the canopy and in the glades was 10 [+ or -] 1% for both locations.
There were significant differences between bare soil cover and quadrat location (van der Waerden one-way ANOVA; p < 0.0001), with the greatest mean cover in the glades (36 [+ or -] 1%) and less below the canopy (2 [+ or -] 0%) (not shown).
Gas exchange rates-There were significant differences in CCh uptake rates of H. tenellum at different light levels (Fig. 3) (one-way ANOVA; p < 0.0001; Tukey-Kramer HSD; p < 0.05). There was a gradual decline in photosynthetic rate from 2000 to 900 [micro]mol/[m.sup.2]/s, but no significant differences. A more pronounced significant decline started at 600 [micro]mol/[m.sup.2]/s - 100 [micro]mol/[m.sup.2]/s, but between 100 and 0 ([micro]mol/[m.sup.2]/s there were no significant differences.
There were no significant differences between stomatal conductance of H. tenellum and the light levels tested (one-way ANOVA; p = 0.1796) (not shown). There were significant differences in mean transpiration rates of H. tenellum (not shown), but only at zero [micro]mol/[m.sup.2]/s (7.12 [+ or -] 1.31 mmol [H.sub.2]O/[m.sup.2]/s), compared to 2000 umol/[m.sup.2]/s (13.81 [+ or -] 1.31 mmol [H.sub.2]O/[m.sup.2]/s) (one-way ANOVA; p = 0.0011; Tukey-Kramer HSD; p < 0.05). Mean maximum photosynthetic rate ([A.sub.max]) for H. tenellum was 34.96 [+ or -] 4.43 [micro]mol C[O.sub.2]/[m.sup.2]/s and the PAR at [A.sub.max] was 2000 [+ or -] 0) [micro]mol/[m.sup.2]/s (Table 1). Light saturation point ([L.sub.sp]), light compensation point ([L.sub.cp]), dark respiration ([R.sub.d]), initial slope (IS), stomatal conductance ([g.sub.s]) and transpiration rate (E) at [A.sub.max] are in Table 1.
Water potential-Mean [psi]x of H. tenellum plants at pre-dawn on July 6, 2010 was -0.84 [+ or -] 0.05 MPa and at midday was -1.79 [+ or -] 0.04 MPa (Fig. 4 A). There was a significant difference between them (one-way ANOVA; p < 0.0001). Soil water potential at pre-dawn was -0.12 [+ or -] 0.00 and at midday was -0.14 [+ or -] 0.01 MPa (Fig. 4B) with a significant difference between them (one-way ANOVA; p = 0.0495).
Regression analysis.--A stepwise, weighted, centered, multiple linear regression was completed to determine which measured factors were important in estimating the density of H. tenellum (Table 2). The best model to predict the density of H. tenellum in the glades was:
Density = [(1/%rock - 0.045) x 264.39] - [(log%litter - 1.013) x (-10.86)] - [(log%herb - 2.296) x (-25.20)] + 46.1
When all three terms were included in the model, the [R.sup.2] was 0.72 (p < 0.0001). Thus, 72% of the variation in density of H. tenellum can be explained by three factors: 59% of the variance was explained by the log of % leaf litter cover, 11% by the log of the % cover of other herbaceous plants and 2% by the reciprocal of % rock cover (Table 2). Light level was important, but it was an all or none factor; light levels below the canopy were too low and H. tenellum was not present and there were no differences in light levels in the glades (see Fig. 2A&C).
We found Heliotropium tenellum almost exclusively in the high light environment of the cedar glade, but not in the surrounding central Texas Juniperus-Quercus woodland (Fig. 2). It is usually found in disturbed habitats with high light levels and rocky soils where competition from the [C.sub.4] grasses is reduced or low (Wilson & Keddy 1986; Enquist 1987; Grace & Tilman 1990; Gurevitch et al. 1990; Van Auken 2000; Grellier et al. 2012; Mazia et al. 2013). It has been reported from other central Texas glades and similar communities in the south central United States (United States Department of Agriculture, Natural Resource Conservation Service 2015).
Using regression analysis, rock cover, leaf litter cover and the cover of other plants were identified as explaining 72 % of the density of H. tenellum in central Texas glades (see Table 2). These characteristics have been shown by others to be important in determining the presence and distribution of various species in glade communities, but not how they related to H. tenellum density or cover (Baskin & Baskin 1973; Belsky et al. 1993; Quarterman et al. 1993; Dye et al. 1995; McClain & Ebinger 2002).
In the current study, overall rock cover in the glades was 30.3 [+ or -] 1.0 %. As cover of rock increased, the amount of soil and the density and cover of H. tenellum decreased. The highest rock cover and density of H. tenellum were both found in the glades in general. This high density of H. tenellum probably occurred because light levels were highest in the glades (Fig. 2) and H. tenellum plants are usually found in disturbed habitats with high light levels and rocky soils where competition from the [C.sub.4] grasses is reduced or low (Gurevitch et al. 1990; Grace & Tilman 1990). Interestingly, when we used regression analysis to show which factors influenced the density of H. tenellum, light level was not a significant factor. This may have occurred because light level in the current study was more of a categorical variable than a numerical variable, an "all or none" condition, where there were either high light levels in the glades, low light levels below the woodland canopy, but none between (Belsley et al. 2005; Sail et al. 2012).
We were convinced that light level was important to explain H. tenellum density or cover, consequently, we examined C[O.sub.2] uptake as a function of light levels. As hypothesized, H. tenellum had a high maximum photosynthetic rate ([A.sub.max]) at high light levels typical of sun plants in open habitats (Valladares & Niinemets 2008). Other measured photosynthetic parameters were high when compared to those of typical shade plants (Fig. 3, Table 1) (Araus et al. 1986; Hull 2002; Larcher 2003; Givnish et al. 2004). The maximum photosynthetic rate ([A.sub.max]) for H. tenellum was 34.96 [+ or -] 4.43 [micro]mol C[O.sub.2]/[m.sub.2]/s (mean [+ or -] 1SE). This value is similar to [A.sub.max] values reported for other species found in open high sun environments (Furuya & Van Auken 2010; Van Auken & Bush 2012). [A.sub.max] values of plants found growing in woodland or forests understories were lower than those of H. tenellum (Hattenschwiler & Korner 1996; Hull 2002; Wayne and Van Auken 2009). The light saturation point, light compensation point and dark respiration of H. tenellum were similar to values reported for sun plants not shade plants (Hull 2002; Wayne and Van Auken 2009).
In the current study, transpiration rates and stomatal conductance were high at the [A.sub.max] (Table 1), which is typical of sun plants. Thus water loss and evaporative cooling of the leaves was likely high as well. This may be the reason that H. tenellum leaf temperature was reported to remain at or below the air temperature (Baskin & Baskin 1978). In many species, transpiration and stomatal conductance occur throughout the hottest periods of the day and are essential for leaf cooling (Laurie et al. 1994). The transpiration and stomatal conductance of H. tenellum were little affected by increased light level but possibly were more affected by increased temperatures through the hottest periods of the day in central Texas cedar glades.
In general, characteristics of cedar glades or woodland gaps include a relatively low density and cover of the dominant herbaceous plants and shallow soils (Baskin & Baskin 1978; Quarterman et al. 1993; Terletzky & Van Auken 1996; Van Auken 2000; Ware 2002). In the current study, overall herbaceous cover other than H. tenellum was 10 [+ or -] 1 % and not significantly different between the open bare patch and under the canopy. Overall cover did not change, but the species causing the cover changed from annuals and [C.sub.4] grasses in the glade to mainly a [C.sub.3] perennial below the canopy (Carex planostachys, cedar sedge) (Wayne & Van Auken 2009). Below the woodland canopy, surface light levels decreased, but soil depth increased as did soil moisture levels (Breshears et al. 1998; Van Auken 2000; Wayne & Van Auken 2009).
Soil water is relatively low, intermittent, and very seasonal in the cedar glades (Wayne & Van Auken 2009). Water would run off and evaporate quickly during hot, dry summers and consequently the soil water potential ([[psi].sub.soil]) would be low (Breshears et al. 1998). Low [[psi].sub.soil] leads to internal water deficits in plants, and consequent stomatal closure may limit both transpiration and photosynthesis (Valladares & Niinemets 2008). In the current study, as midday [[psi].sub.soil] became more negative, the xylem water potential ([[psi].sub.x]) of H. tenellum also became more negative, which has been reported for other species (Fonteyn et al. 1985; Breshears et al. 1998; Clements et al. 2002; Schedlbauer & Pistoia 2013). Mean v|/v at midday did not indicate that H. tenellum was experiencing major water stress, even though the ambient temperature, light intensity, photosynthetic rates, transpiration rates and stomatal conductance were all high. These factors would change during long, hot, dry central Texas summers. In spite of the above, H. tenellum appears well adapted to dry soils, high light and temperature levels of these open central Texas cedar glades. As a result of these comparisons and where H. tenellum has been found, this species should be considered as a sun plant and probably a stress tolerant species.
We wish to thank Kelly Jo Stevens of the Biology Department, the University of Texas at San Antonio for help with figure construction and modification. Two anonymous reviewers offered helpful comments as well.
Araus, J. L., L. Alegre, L. Tapia, R. Calafell & M. D. Serret. 1986. Relationships between photosynthetic capacity and leaf structure in several shade plants. Amer. J. Bot. 73(12): 1760-1770.
Baskauf, C. J. 1993. Comparative population genetics and ecophysiology of a rare and a wide spread species of Echinaceae (Asteraceae), Ph. D. Dissertation. Vanderbilt University. Nashville, TN. 185 p.
Baskin, J. M. & C. C. Baskin. 1973. Observations on the ecology of Sporobolus vaginiflorus in cedar glades. Castanea 38(1):25-35.
Baskin, J. M. & C. C. Baskin. 1978. Leaf temperatures of Heliotropium tenellum and their ecological implications. Amer. Midi. Nat. 100(2):488-492.
Baskin. J. M. & C. C. Baskin. 1985. Life cycle ecology of annual plant species of cedar glades of southeastern United States, pp. 371-398. In J. White [ed.]. The Population Structure of Vegetation. Dr. W. Junk, Dordrecht, Netherlands.
Begon. M., C. R. Townsend, & J. L. Harper. 2006. Ecology: from individuals to ecosystems. Blackwell Publishing, Maldon. MA. 738 p.
Belsky, A. J., S. M. Mwonga, R. G. Amundson, J. M. Duxbury & A. R. Ali. 1993. Comparative effects of isolated trees on their under canopy environments in high- and low-rainfall savannas. J. Appl. Ecol. 30(1): 143-155.
Belsley, D. A., E. Kuh & R. E. Welsch. 2005. Regression Diagnostics. John Wiley and Sons, Belmont, CA. 310 p.
Breshears, D. D., J. W. Nyhan, C. h. Heil & B. P. Wilcox. 1998. Effects of woody plants on microclimate in a semiarid woodland: soil temperature and evaporation in canopy and inter canopy patches. Int. J. Plant Sci. 159(6): 1010-1017.
Bush, J. K. 2006. Abiotic edaphic factors affecting the growth of a threatened North American Sunflower, Helianthus paradoxus (Asteraceae). Plant Ecol. 183(2):215-225.
Clements, R. K., J. M. Baskin & C. C. Baskin. 2002. The comparative biology of the two closely-related species Penstemon tenuiflorus Pennell and P. hirsutus (L.) Willd. (Scrophulariaceae, section Graciles): IV. Effects of shade, drought, and soil type on survival and growth. Castanea 67(2): 177-187.
Daniel. C. & F. Wood. 1980. Fitting Equations to Data. John Wiley and Sons, New York. NY. 458 p.
Dye. K. L., D. N. Ueckert & S. G. Whisenant. 1995. Redberry juniper-herbaceous under story interactions. J. Range Manage. 48(2): 100-107.
Enquist, M. 1987. Wildflowers of the Texas Hill Country. Lone Star Botanical, Austin, TX. 275 p.
Fonteyn. P. J., T. M. McClean & R. E. Akridge. 1985. Xylem pressure potential of three dominant trees of the Edwards Plateau of Texas. Southwest. Nat. 30(1): 141-146.
Furuya. M. & O. W. Van Auken. 2010. Gas exchange rates of three sub-shrubs of central Texas savannas. Madrono 57(3): 170-179.
Gagliardi, J. W. & O. W. Van Auken. 2010. Distribution of Verbesinu virginica (Asteraceae. frost weed) in central Texas and possible causes. Tex. J. Sci. 62(3): 163-182.
Givnish. T.J., R.A. Montgomery & G. Goldstein. 2004. Adaptive radiation of photosynthetic physiology in the Hawaiian lobeliads: Light regimes, static light responses, and whole-plant compensation points. Amer. J. Bot. 91(2):228-246.
Grace. J. B. & D. Tilman. 1990. Perspectives on Plant Competition. Academic Press, New York. NY. 484 p.
Grellier, S., S. Barot, J-L. Janeau & D. Ward. 2012. Grass competition is more important than seed ingestion by livestock for Acacia recruitment in South Africa. Plant Ecol. 213(3):899- 908.
Gurevitch. J., P. Wilson. J. L. Stone, P. Teese & R. J. Stoutenburgh. 1990. Competition among old-field perennials at different levels of soil fertility and available space. J. Ecol. 78(3):727-744.
Hattenschwiler, S. & C. Korner. 1996. Effects of elevated CO: and increased nitrogen deposition on photosynthesis and growth of understory plants in spruce model ecosystems. Oecologia 106(2): 172-180.
Hull, J. C. 2002. Photosynthetic induction dynamics to sunflecks of four deciduous forest understory herbs with different phenologies. Int. J. Plant Sci. 163(6):913-924.
Kleinbaum, D. G., L. L. Kupper & K. E. Mueller. 1988. Applied Regression Analysis and other Multivariate Methods. PWS-KENT Publishing. Boston, MS. 718 p.
Krebs, C. J. 2009. Ecology: The Experimental Analysis of Distribution and Abundance (6th edition). Benjamin Cummings, Menlo Park, CA. 655 p.
Kucera, C. L. & S. C. Martin. 1957. Vegetation and soil relationships in the glade region of the southwestern Missouri Ozarks. Ecology 38(2):285-291.
Larcher, W. 2003. Physiological plant ecology: ecophysiology and stress physiology of functional groups. Springer, New York, NY. 513 p.
Laurie, S.. M. Bradbury & G. R. Stewart. 1994. Relationships between leaf temperature, compatible solutes and anti-transpirant treatment in some desert plants. Plant Sci. 100(2): 147-156.
Mason. R. L., R. F. Gunst & J. L. Hess. 1989. Statistical Design and Analysis of Experiments. John Wiley and Sons, New York, NY. 760 p.
Mazia, N., P. M. Tognetti & E. D. Cirino. 2013. Patch identity and the spatial heterogeneity of woody encroachment in exotic-dominated old-field grasslands Plant Ecol. 214(2):267-277.
McClain, W. E. & J. E. Ebinger. 2002. A comparison of the vegetation of three limestone glades in Calhoun county, Illinois. Southeast. Nat. 1 (2): 179-188.
Moores. E. M. 2011. Serpentinites and other ultramalfic rocks: why they are important for earth's history and possibly its future, pp. 3-28. In S. Harrison and N. Rajakaruna [eds.]. Serpentine: the evolution and ecology of a model system. University of California Press, Berkeley, CA.
Quarterman, E., M. P. Burbanck & D. J. Shure. 1993. Rock outcrop communities: Limestone, sandstone, and granite, pp. 35-86. In W. 11. Martin, S. G. Boyce, and A. C. Echternacht [eds.]. Biodiversity of Southeastern United States/ Upland Terrestrial Communities. John Whiley and Sons. New York, NY.
Riskind, D. H. & D. D. Diamond. 1988. An introduction to environments and vegetation, pp. 1-15. hi B. Amos and F. R. Gehlbach [eds.], Edwards Plateau vegetation: plant ecological studies in central Texas. Baylor University Press, Waco, TX.
Sail. J., L. Creighton & A. Lehman. 2012. JMP Start Statistics, 3rd ed. SAS Institute, Inc. Brooks/Cole-Thompson Learning, Belmont, CA. 560 p.
Schedlbauer, J. L. & V. L. Pistoia. 2013. Water relations of an encroaching vine and two dominant [C.sub.4] grasses in the serpentine barrens of southeastern Pennsylvania. J. Torrey Bot. Soc. 140(4):493-505.
Scholander, P. F., H. T. Hammel. E. D. Bradstreet & F. A. Hemmingsen. 1965. Sap pressure in vascular plants. Science 148(3668):339-346.
Sellards, E. H., W. S. Adkins & F. B. Plummer. 1932. The geology of Texas. Vol. 1 Stratigraphy. University of Texas Bulletin 3232, Austin, TX. 1007 p.
Smith, T. M. & R. L. Smith. 2012. Elements of ecology. Benjamin Cummings, New York, NY. 704 p.
Terletzky, P. A. & O. W. Van Auken. 1996. Comparison of cedar glades and associated woodlands of the southern Edwards Plateau. Tex. J. Sci. 48(1):55-67.
United States Department of Agriculture, Natural Resource Conservation Service. 2015. The PLANTS Database. Retrieved March 2015 from National Data Team, Greensboro, NC.
Valladares. F. & U. Niinemets. 2008. Shade tolerance, a key plant feature of complex nature and consequence. Annu. Rev. Ecol. Syst. 39:237-257.
Van Auken, O. W. 2000. Characteristics of intercanopy bare patches in Juniperas woodlands of the southern Edwards Plateau, Texas. Southwest. Nat. 45(2):95-110.
Van Auken, O. W. & J. K. Bush. 1992. Factors affecting the density and distribution of Selaginella lepidophylla in the Black Gap area of the Chihuahuan Desert of western Texas. Southwest. Nat. 37(3):274-279.
Van Auken, O. W. & J. K. Bush. 2012. Photosynthetic rates of two species of Malvaceae, Malvaviscas arboreas var. drummondii (WaxMallow) and Abutilon therphrasti (Velveltleaf). Southwest. Nat. 56(3):325-332.
Van Auken, O. W., J. K. Bush & S. A. Elliott. 2005. Ecology Laboratory Manual. Pearson, Boston, MA. 171 p.
Ware, S. 2002. Rock outcrop plant communities (glades) in the Ozarks: a synthesis. Southwest. Nat. 47(4):585-597.
Wayne, E. R. & O. W. Van Auken. 2009. Light responses of Carex planostachys from various microsites in Juniperus community. J. Arid Env. 73(4-5):435-443.
Wilson, S. D. & P. A. Keddy. 1986. Species competitive ability and position along a natural stress/disturbance gradient. Ecology 67(5): 1236-1242.
Woodward, F. I. 1987. Climate and Plant Distribution. Cambridge University Press, Cambridge, MA. 174 p.
OWVA at: oscar.vanauken[R] utsa.edu
A. K. Boeck and O. W. Van Auken
Department of Biology, University of Texas at San Antonio One UTSA Circle, San Antonio, TX 78249-0662
Caption: Figure 1. Part of a cedar glade in the Edwards Plateau region of Bandera County in central Texas. Juniperus ashei and Quercus virginiana trees surround the edge of the cedar glade with some Opuntia macrohiza plants in the mid-ground (Photo by A. K. Boeck July 6, 2010).
Caption: Figure 2. Presented are A. mean light levels (PAR - [micro]mol/[m.sup.2]/s). B. soil depth (cm.), C. density of Heliotropium tenellum (plants/nr), and U. percent cover of leaf litter along transects from below a Juniperus-Quercus canopy into a cedar glade. Negative numbers along the x-axis indicate quadrats below the canopy, while positive numbers indicate quadrats in the cedar glade. Quadrat location zero on the x-axis indicates the tree drip line. All four factors are significantly different (van der Waerden one-way ANOVA; p < 0.0001). Errors bars (not shown) are approximately 10% of the mean.
Caption: Figure 3. Mean photosynthetic rates ([micro]mol/[m.sup.2]/s) of Heliotropium tenellum were significantly different (one-way ANOVA; p < 0.0001) when exposed to various light levels. Inset graph displays the lower portion of the line graph from 0-100 PAR. Different letters between light levels indicate significant differences (Tukey-Kramer HSD; p < 0.05). Error bar is an example and represent [+ or -] one standard error of the mean.
Caption: Figure 4. Mean xylem water potential A. of Heliotropium tenellum in central Texas cedar glades at different times of the day on July 6, 2010 (one-way ANOVA; p < 0.0001) and mean soil water potential B. in central Texas cedar glades at different times of the day on July 6, 2010 (one-way ANOVA; p = 0.0495). Numbers below each bar are the actual mean values. One standard error is approximately 10% of the mean.
Table 1. Mean [+ or -] one standard error for Heliotropium tenellum photosynthetic characteristics including maximum net photosynthetic rate ([A.sub.max]), light level at the [A.sub.max], light saturation (Lsat), light compensation point (Lcp), dark respiration rate (Rd), initial slope or quantum yield efficiency (IS), stomatal conductance at Amax (gs) and transpiration rate at Amax (E). Parameter Mean [+ or -] SE [A.sub.max] ([micro]mol C[O.sub.2]/[m.sup.2]/s) 34.96 [+ or -] 4.43 PAR at [A.sub.max]([micro]mol/[m.sup.2]/s) 2000 [+ or -] 0 [L.sub.sat] ([micro]mol/[m.sup.2]/s) 591 [+ or -] 122 [L.sub.cp] ([micro]mol/[m.sup.2]/s) 38 [+ or -] 3 [R.sub.d] ([micro]mol C[O.sub.2]/[m.sup.2]/s) 2.63 [+ or -] 0.38 IS ([micro]mol C[O.sub.2]/[micro]mol x quanta) 0.07 [+ or -] 0.01 [g.sub.s] (mol [H.sub.2]O/[m.sup.2]/s) 0.44 [+ or -] 0.08 E (mmol [H.sub.2]O/[m.sup.2]/s) 13.81 [+ or -] 1.62 Table 2. Results of stepwise, weighted, centered, multiple linear regression of density of Heliotropium tenellum on seven independent variables (PAR. log of soil depth, reciprocal of relative rock cover, log of relative leaf litter cover, reciprocal of relative woody plant cover, log of other relative herbaceous plant cover, and reciprocal of relative cover of bare soil). Presented are the Analysis of Variance with the overall p-value and fit of the model, as well as the [R.sup.2] and the adjusted [R.sup.2]. The variables, degrees of freedom (DF), parameter estimates, standard errors, t-values, and p-values are also presented. Analysis of Variance for Density Source DF Sum of Mean Square F Value p > F Squares Model 7 142748.00 20393.00 23.79 <0.001 Error 61 52294.00 857.27 -- -- Corrected Total 68 195042.00 -- -- -- Regression Analysis Root MSE 29.28 [R.sup.2] 0.72 Dependent Mean 29.93 Adjusted [R.sup.2] 0.70 Coefficient of Variation 97.84 -- -- Parameter Estimates Parameter Standard Variable DF Estimate Error t-values p > t Intercept 1 46.10 2.46 18.77 <0.0001 Log % herb 1 -25.20 5.76 -4.37 <0.0001 Log % litter 1 -10.86 3.56 -3.05 0.0033 l/% rock 1 264.39 99.88 2.65 0.0103 1/% woody 1 11.13 7.76 1.43 0.1564 1/% bare soil 1 -2.39 11.76 -0.20 0.8393 PAR 1 0.00119 0.00982 0.12 0.9038 Log soil depth 1 -0.37 6.38 -0.06 0.9545
|Printer friendly Cite/link Email Feedback|
|Author:||Boeck, A.K.; Van Auken, O.W.|
|Publication:||The Texas Journal of Science|
|Date:||Feb 1, 2016|
|Previous Article:||Albino raccoons (Procyon lotor) from League City, Texas.|
|Next Article:||Range extensions and county records for vascular plants from Mills County, Texas.|