See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/318318929
Evaluating Removal of Competition on Morphology of Endangered Slender Rush-Pea ( Hoffmannseggia tenella...
Article in Natural Areas Journal · July 2017 DOI: 10.3375/043.037.0311
CITATIONS READS 0 10
4 authors, including:
Sandra Rideout-Hanzak David B Wester Texas A&M University - Kingsville Texas A&M University - Kingsville
20 PUBLICATIONS 91 CITATIONS 143 PUBLICATIONS 1,752 CITATIONS
SEE PROFILE SEE PROFILE
Weimin Xi Texas A&M University - Kingsville
62 PUBLICATIONS 423 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Restoration Planning and Evaluation Following Damage by the Southern Pine Beetle in a Sustainable Forest Management Context View project
Comanche - Faith Ranches Research Project View project
All content following this page was uploaded by Sandra Rideout-Hanzak on 13 November 2017.
The user has requested enhancement of the downloaded file. Evaluating Removal of Competition on Morphology of Endangered Slender Rush- Pea (Hoffmannseggia tenella) Endemic to Southern Texas, USA Author(s): Ashley C. McCloughan, Sandra Rideout-Hanzak, David B. Wester and Weimin Xi Source: Natural Areas Journal, 37(3):382-393. Published By: Natural Areas Association https://doi.org/10.3375/043.037.0311 URL: http://www.bioone.org/doi/full/10.3375/043.037.0311
BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.
BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research. R E S E A R C H A R T I C L E ABSTRACT: Slender rush-pea (Hoffmannseggia tenella) is an endangered herbaceous legume endemic to southern Texas. Agricultural expansion and invasive nonnative grasses have caused reductions in its • range, threatening existing populations. We compared three management techniques that control com- petition to no management (control) to determine effects on slender rush-pea survival and morphology. We treated some slender rush-pea plants with summer prescribed burning. In the mechanical treatment Evaluating Removal we controlled above-ground competition surrounding target slender rush-pea plants with a string trimmer and we removed both above- and below-ground competition with herbicide in the chemical treatment. of Competition No slender rush-pea plants died during our two-year study. Plants in each competition removal treatment produced more main stems than control plants at various times throughout the study. Summer prescribed on Morphology burning produced longer stems, while both summer burning and chemical treatment resulted in more leaves on the longest stem and more flowers than control plants at various times throughout the study. of Endangered Slender rush-pea individuals with more main stems, longer stems, and more leaves may be more robust and resilient to environmental stress. We speculate increased flower production may indicate enhanced Slender Rush-Pea reproduction. At the pasture scale herbicide treatment is not practical as it may eradicate remaining native plants. Summer prescribed burning may be the most effective method for increasing short-term robustness of slender rush-pea; mowing could be a temporary substitute when fire is impractical. Results (Hoffmannseggia indicate that slender rush-pea benefits from active management and control of competition from nonnative tenella) Endemic to grasses, and if managed properly slender rush-pea could persist in their presence. Index terms: endangered, herbaceous legume, Hoffmannseggia tenella, prescribed burning, removal of Southern Texas, USA competition
Ashley C. McCloughan1 INTRODUCTION must quickly adapt—on the population and potentially the individual level. Among the 1Actagro Soil and Plant Health Presently, grasslands are some of the most problematic species in southern Texas is Technology Solutions threatened or endangered ecosystems Kleberg bluestem (Dichanthium annulatum 2282 Sugar Grove Rd. worldwide because of habitat loss and [Forssk.] Stapf), a C4 perennial bunchgrass Bowling Green, KY 42101 destruction through agriculture, urbaniza- native to India, China, and Northern Africa tion, and introduction of nonnative species, (Ortega et al. 2007; Ruffner and Barnes 2,4 Sandra Rideout-Hanzak which sometimes become invasive in nature 2012). It is a prolific seeder and displays David B. Wester2 (Sampson and Knopf 1994; Ruffner and a high tolerance for drought and cold, as Weimin Xi3 Barnes 2012). Recent unprecedented rates well as some salinity tolerance (Ortega et of spread of nonnative species into new al. 2007; TIPPC 2011). First introduced to habitats and ecosystems have forced native the central and southern Great Plains in the 2Caesar Kleberg Wildlife Research flora and fauna into direct competition with early 1900s for the purpose of improving Institute and Department of Animal, these new, and sometimes competitively pasture and rangeland forage, it is actually Rangeland and Wildlife Sciences superior, species—often to the detriment of lower forage quality than previously Texas A&M University-Kingsville of native communities (D’Antonio and thought, especially in southern Texas MSC 218 Vitousek 1992; Rice et al. 1997). Perennial (Ortega et al. 2007; Ruffner and Barnes Kingsville, TX 78363 grasses in particular are often introduced 2012). Not only has it degraded habitat and to new areas for either livestock forage lowered forage quality, its abilities to with- 3Department of Biological and Health (i.e., because of their tolerance for ex- stand grazing pressure, high temperatures, Sciences treme conditions) or to stabilize soils and and drought have allowed this nonnative Texas A&M University-Kingsville MSC 158, TAMUK prevent erosion (D’Antonio and Vitousek grass to spread unchecked across Texas Kingsville, TX 78363 1992; Ruffner and Barnes 2012). These and northern Mexico landscapes (Ortega introduced grasses present native plants et al. 2007). with novel competition for resources (e.g., • light, water, and nutrients) and may even- This invasion is of particular concern for tually eliminate native species, lowering a federally endangered species endemic overall biotic diversity of the community to southern Texas—slender rush-pea 4 Corresponding author: [email protected]; (D’Antonio and Vitousek 1992; McDonald (Hoffmannseggia tenella Tharp & L.O. 361-593-4546 and McPherson 2011). Williams). In the Fabaceae family, the flowering plant genus Hoffmannseggia Southern Texas hosts several highly in- currently consists of 21 identified species Natural Areas Journal 37:382–393 vasive nonnative grasses that present a and exhibits an amphitropical distribution challenge for native flora and fauna, which in dry subtropical and warm temperate
382 Natural Areas Journal Volume 37 (3), 2017 regions of North and South America (Simp- An ability to compete at the individual level most recent quantitative estimate was in son et al. 2004, 2005). Slender rush-pea may be expressed through morphological the “thousands” (USFWS 2008). It exists grows in former native short-grass prairie plasticity, which is a mechanism employed within remnant Western Gulf Coastal Plain communities found in clayey soils of by an individual to adjust to changes in its region of southern Texas in a semi-arid cli- blackland prairies and creek banks in the natural environment, allowing it to more mate. The location receives approximately Gulf Coastal Plain of southern Texas—par- successfully compete for niche space 83.6 cm (32.9 in) of annual mean rainfall ticularly in Nueces and Kleberg Counties, (Aarssen 1983; Crick and Grime 1987; with an average temperature fluctuation Texas (USFWS 1988). Craine and Dybzinski 2013). Some plants between 16.1 and 27.7 °C (61 and 82 °F) display this through alteration of pheno- (NOAA 2014). Study plants were located Slender rush-pea is an herbaceous perennial typic traits to suit variable environmental in Victoria clay soil, 1%–3% slope, fine, legume, 8–15 cm tall, with a woody taproot conditions. smectitic, hyperthermic Sodic Haplusterts and reddish color on older stems. Leaves (VcB) (NRCS 2014) just west of Carreta are alternate and bipinnately compound We know of only two studies on slender Creek. The sampling area—roughly 0.8 with 11–13 oblong leaflets, which are rush-pea. One study focused on population ha (2 ac)—is free of gravesites and has slightly hairy on the undersurface (USFWS vigor as determined by population size and been subjected to uniform management of 1985). Flowers are typically salmon color, density (Poole and Janssen 1996). Pressly periodic mowing for several decades. Al- and are produced between early March (unpubl. data) studied survival, number though Kleberg dominates the herbaceous and June, then sporadically thereafter of petioles, pinnae, and seed pods in a community, there are remnants of the native with rainfall (USFWS 1985). Legumes shade-house experiment that simulated grass community present, including buffa- are straight, contain 2–4 seeds, and ma- shade, mowing, and root competition; lograss (Bouteloua dactyloides [Nutt.] J.T. ture March through July (USFWS 1988). however, measurements and methods were Columbus), Texas wintergrass (Nassella Slender rush-pea regenerates through seed; not described and the field portion of this leucotricha [Trin. & Rupr.] Pohl), plains while insect pollination is possible, flowers study was abandoned because of extreme bristlegrass (Setaria leucopila [Scribn. & will self-fertilize (Pressly, unpubl. data). drought conditions and human disturbance Merr.] K. Schum), and hooded windmill- Slender rush-pea is occasionally found (Pressly, unpubl. data). grass (Chloris cucullata Bisch). An array with another endangered plant, South of herbaceous forbs (e.g., wheatspike Texas ambrosia (Ambrosia cheiranthifolia Our goal with this project was to determine scalystem (Elytraria bromoides Oerst.), A. Gray) (USFWS 1994). slender rush-pea’s responses to reduced Gregg’s tube tongue (Justicia pilosella aboveground and/or belowground plant [Nees] Hilsenb.), whitemouth dayflower Today slender rush-pea is confirmed in competition. Specifically, our objective was (Commelina erecta L.), spreading fanpetals several locations in parks, cemeteries, and to determine effects of summer prescribed (Sida abutifolia Mill.), and South Texas highway rights-of-way, with some uncon- fire and neighbor removal by means of ambrosia) are also common associates of firmed reports on privately owned lands. chemical or mechanical control on survival slender rush-pea throughout our study area. Reduction of its native range has been and morphometric response of slender Woody plants, such as honey mesquite predominantly driven by the expansion of rush-pea plants. In our initial assessment (Prosopis glandulosa Torr.) and huisache agriculture (i.e., cropland and cattle graz- of plants in the studied population, we (Acacia farnesiana [L.] Willd.) are present, ing) into the Gulf Coastal Plains grasslands found them to have one to three main stems but less common. (USFWS 1988). Invasion from introduced originating from the woody root. We also pasture grasses such as Kleberg bluestem found that plants exhibited variability in and Bermuda grass (Cynodon dactylon [L.] length of stem and number of leaves, flow- Survey Plants Pers.) have further diminished the range ers, and seed pods. Thus, we chose these of this species (USFWS 1988). These variables along with survival as important To evaluate slender rush-pea’s response conditions have placed slender rush-pea measures of robustness. to competition removal and manipulation, in a vulnerable position and thus, in 1985, we established four treatment groups and the species was classified as endangered measured response variables repeatedly MATERIALS AND METHODS without critical habitat on the Federal over two years. In June 2012, we sur- Register (USFWS 1985). veyed a pasture in the eastern portion of Study Area the cemetery that had not been used for The species appears to persist on disturbed gravesites, and recorded all slender rush- prairie until encroaching species crowd it The slender rush-pea population studied in pea plants with GPS. Using ArcGIS (Esri, out through resource competition (USF- these experiments is at the St. James Cath- Redlands, CA, USA) we randomly chose WS 1985, 1988). Because dominance by olic Cemetery, located in southern Nueces 129 plants for study. Of these, 33 plants nonnatives is novel and can cause rapid County, Texas, USA. It is the largest known were scheduled for herbicide treatment ecosystem changes, native species must be extant population, estimated at more than of neighboring vegetation, 33 plants for able to quickly adjust if they are to contend 10,000 individuals when discovered in mechanical treatment of neighboring for resources (Denslow and Hughes 2004). the mid-1980s (Poole, unpubl. data). The vegetation, and 33 plants were marked as
Volume 37 (3), 2017 Natural Areas Journal 383 control plants. We programmed ArcGIS to stem to the nearest mm with a standard 2012, May, July, and September 2013, and randomly select these 99 plants from the ruler. The ruler was placed at the soil April, June, and July 2014, following data same area within one population that were surface on the north side of the plant for collection on these dates. Also in July 2012, at least at least 3 m apart so that study plants consistency. 33 control plants were randomly selected. could be treated and measured without dis- No manipulation was applied to these turbance to neighboring slender rush-pea plants or their immediately surrounding Number of leaves on the longest stem plants and their surrounding vegetation. We areas during the project period. also randomly chose 30 plants that were at The number of leaves branching off the least 5 m apart for prescribed burning from The study was designed to include a pre- longest main stem. another area within the same population. scribed burning treatment. However, during We separated prescribed burning plants early stages of the study the southern part from others in order to avoid burning or Number of flowers showing color on of Texas was in severe drought resulting mechanically disturbing plants in other the entire plant in a county-wide ban on outdoor burn- treatments. We marked each randomly ing, which included prescribed burning, selected slender rush-pea plant with two The number of flowers on the entire plant in Nueces County. Therefore, the fourth aluminum tree tags, one on the north and that were developed enough to show color treatment, prescribed fire, was delayed until south sides of the plant, placed on a U-pin on the petals. 6 August 2013, when the ban on outdoor that was pushed down into the soil until burning was temporarily lifted following the tag rested on the surface. light rainfall. Initially, a line was mowed Number of fruiting bodies attached to around the north, east, and south sides of the plant the burn population that tied into a cemetery Measurements road on the west side in preparation for The number of seed pods attached to the a southeast wind during burning. Before For each plant, six response variables were entire plant. lighting we watered the mowed area to recorded before treatments began and pe- use as a fireline. We chose to use a mowed riodically throughout the study following Treatments wetline rather than installing a mineral rain events that would have stimulated soil line to contain the fire because of the growth or reproduction: In July 2012, an herbicide treatment was two endangered herbaceous plants present first applied to the neighbors of 33 ran- at the site—slender rush-pea and South Mortality domly selected slender rush-pea plants, Texas ambrosia; we did not want to destroy effectively eliminating both aboveground any plants or disturb the soil facilitating To assess mortality we visually located and belowground competition. To accom- further encroachment of invasive grasses. the tags for an individual plant and de- plish this, a plastic container was placed We first burned a blackline approximately termined if the plant was still present. If over the target slender rush-pea plant and 3-m wide just inside the northern part of no aboveground plant parts were obvious neighbors within 1 m of it were sprayed the mowed wetline. Then, beginning with between the tags, we examined the soil with glyphosate, a nonselective herbicide. the plant farthest downwind, we ignited manually for belowground parts. If be- The second treatment, also initially em- the area surrounding each selected slender lowground parts were encountered where ployed in July 2012, was a mechanical rush-pea plant in a separate ignition for no aboveground parts were present, we treatment and was applied to the neighbors true replication, allowing each fire to pass recorded that individual as alive, but no of 33 randomly selected slender rush-pea over the target plant naturally, and tying further measurements were taken. If no plants. Aboveground parts of competing into the previously burned area downwind. belowground parts were detected, that vegetation within 1 m of each selected Each burn was at least 0.001 ha (10 m2) in individual was recorded as dead. slender rush-pea were eliminated with a size, which yields temperatures no different string trimmer to maintain a local envi- than a 16-ha (40-ac) fire (Wright and Bailey ronment free of shade from aboveground 1982), thus simulating a natural, summer Number of main stems competition; in this treatment, however, fire occurrence. Prescribed burning treat- belowground interactions with neighbors ments removed all aboveground biomass, The number of stems branching from the continued. We chose a 1-m diameter including the aboveground portion of the base of the plant. treatment area because it would be large surveyed slender rush-pea individuals. enough to allow enhanced resources both Length of the longest main stem aboveground and belowground while Prior to burning, we estimated fuel load minimizing impact on neighboring plants. by clipping four 0.25-m2 quadrats placed We measured length of the longest main Both of these treatments were repeated as randomly throughout the area to be burned. stem by manually gathering a plant’s necessary due to rainfall and competitor This was done by placing the quadrat in aboveground parts and holding them erect, regrowth throughout the duration of the vegetation adjacent to the area to be burned, measuring the farthest point on the longest study; plots were re-treated in October clipping and removing all aboveground
384 Natural Areas Journal Volume 37 (3), 2017 biomass rooted within the quadrat, as well degrees of freedom closest to the value of Pre-Treatment Comparisons and as the litter. Samples were placed in a pa- 1 was used). Throughout most of the study Plant Survival per bag and dried for three days at 60 °C. period, few fruits were recorded and thus, Samples were then weighed and converted data were sparse and neither Poisson nor At the beginning of our study (July 2012), to kg/ha. Mean fuel load was 2434 (±316) negative binomial distributions were appro- there were no differences in number of kg/ha. Weather variables of temperature, priate; for this response variable we used main stems (F2,101.6 = 1.81, P = 0.17) relative humidity, and wind speed and a permutation-based analysis of variance (Figure 1), length of longest stem (F2,222 direction were recorded on site every (for the CRD repeated measures analysis) = 0.69, P = 0.50) (Figure 2), number of hour for the 3-hr burning event (Kestrel using Euclidean distance as a resemblance leaves per longest stem (F2,351 = 0.31, P = 4500 Weather Meter, Nielsen-Kellerman, metric (Anderson 2001). The only response 0.74) (Figure 3), or number of flowers per Boothwyn, PA, USA) (Table 1). variable that was measured on a continu- plant (F2,268.7 = 0.44, P = 0.65) (Figure 4) ous scale was length of longest stem. For among chemical removal of neighbors, me- length of longest stem, we used Shapiro– chanical removal of neighbors, and control Statistical Analysis Wilk’s (1965) test to assess normality of treatments. Just before the prescribed burn residuals. For repeated measures analyses, treatment was conducted a year later, there This experiment was set up as a com- patterns of variation and covariation can were no differences among treatments in pletely randomized design with respect vary over time (Littell et al. 2006); we the number of stems (July 2013: F = to mechanical, chemical, and control 3,459.3 used an information-theoretic criterion 0.64, P = 0.59) or flowers F( = 0.65, treatments because these treatments were 3,569 (AIC ; Burnham and Anderson 2002) to P = 0.58) per plant. No target slender randomly assigned to individual plants; c select the best model from among the rush-pea individuals died over the course further, because plants were measured over following structures: compound symmetry, of the study. time, a repeated measures analysis is ap- first-order autoregressive, Toeplitz (and propriate. Therefore, a CRD with repeated heterogeneous forms of these structures), measures analysis (Kirk 2013) was used Number of Stems first-order autoregressive moving average, to test hypotheses that means of response and variance-components; a corrected AIC variables (listed below) were not affected By September 2014, two years after ini- was used because n/k (the ratio of sample by treatment, date, or their interaction. Re- tiating chemical and mechanical removal size to the number of parameters estimated) sponse variables were (1) length of longest of neighbors and one year after prescribed was <40 (Burnham and Anderson (2002). stem, (2) number of leaves, (3) number of burning, there was a significant difference stems, (4) number of flowers, (5) number in the number of stems per plant (F3,351.5 = of fruits, and (6) mortality. Numbers of RESULTS AND DISCUSSION 4.98, P = 0.002) with control plants having stems, leaves, flowers, and fruits—all fewer stems than all other treatments (Fig- count variables—are (by definition) not ure 1). At each sampling period following normally distributed because they are mea- Precipitation prescribed burning in August 2013, plants sured on a discrete scale. Commonly used in both the burn and mechanical removal distributions for count variables include Monthly rainfall averages may explain of neighbors treatments had more stems Poisson and negative binomial distribution; short-term pulses in herbaceous plant (September 2013: t365.4 = 3.85, P < 0.01; following Myers et al. (2002), we analyzed growth (Fravolini et al. 2005) (Figures April 2014: t351.8 = 3.31, P < 0.01; June each of these response variables assuming 1–4). However, rainfall remained below 2014: t379.8 = 3.65, P < 0.01; July 2014: (1) a Poisson, and (2) a negative binomial average throughout the course of this t408.1 = 2.02, P = 0.04) than control plants distribution, and used the generalized chi- study. There were no months in which (Figure 1). Number of stems of plants square/degrees of freedom statistic to select rainfall reached the historical mean. Thus, whose neighbors were removed chemically the most appropriate distribution (i.e., the we assume plant responses to treatments was not consistently greater than control distribution with a generalized chi-square/ were due to treatments. plants throughout the study.
Table 1. Weather variables recorded during prescribed burning on 6 August 2013 at St. James Catholic Cemetery in Nueces County, Texas, USA.
Temperature °C Relative Mean wind speed m/s Wind Time (AM) (°F) humidity % (mph) direction 9:15 31.6 (89) 65 3.6 (8) SE 10:35 33.3 (92) 56 4.0 (9) SE 11:40 35.0 (95) 46 4.5 (10) SE
Volume 37 (3), 2017 Natural Areas Journal 385 Figure 1. Mean number of main stems per plant (±SEM) on slender rush-pea individuals at St. James Catholic Cemetery in Nueces County, Texas, USA. Bars on lower portion of graph represent observed monthly precipitation, whereas the line illustrates monthly historical averages (NOAA 2014). Results from July 2012 through May 2013 were analyzed using data from mechanical, chemical, and control treatments while results from July 2013 to the end included these three as well as prescribed fire treatment data. Prescribed burning was conducted 6 August 2013, indicated by arrow. Treatments within a date followed by the same letter are not different (P > 0.05). Mechanical and chemical treatments were reapplied immediately following each measurement date.
386 Natural Areas Journal Volume 37 (3), 2017 Figure 2. Mean length of longest stem (mm) (±SEM) of slender rush-pea plants at St. James Catholic Cemetery in Nueces County, Texas, USA. Bars on lower portion of graph represent observed monthly precipitation, whereas the line illustrates monthly historical averages (NOAA 2014). Results from July 2012 through May 2013 were analyzed using data from mechanical, chemical, and control treatments while results from July 2013 to the end included these three as well as prescribed fire treatment data. Prescribed burning was conducted 6 August 2013, indicated by arrow. Treatments within a date followed by the same letter are not different (P > 0.05). Mechanical and chemical treatments were reapplied immediately following each measurement date.
Volume 37 (3), 2017 Natural Areas Journal 387 Figure 3. Mean number of leaves on the longest stem (±SEM) of slender rush-pea plants at St. James Catholic Cemetery in Nueces County, Texas, USA. Bars on lower portion of graph represent observed monthly precipitation, whereas the line illustrates monthly historical averages (NOAA 2014). Results from July 2012 through May 2013 were analyzed using data from mechanical, chemical, and control treatments while results from July 2013 to the end included these three as well as prescribed fire treatment data. Prescribed burning was conducted 6 August 2013, indicated by arrow. Treatments within a date followed by the same letter are not different (P > 0.05). Mechanical and chemical treatments were reapplied immediately following each measurement date.
388 Natural Areas Journal Volume 37 (3), 2017 Figure 4. Mean number of flowers showing color (±SEM) on slender rush-pea plants at St. James Catholic Cemetery in Nueces County, Texas, USA. Bars on lower portion of graph represent observed monthly precipitation, whereas the line illustrates monthly historical averages (NOAA 2014). Results from July 2012 through May 2013 were analyzed using data from mechanical, chemical, and control treatments while results from July 2013 to the end included these three as well as prescribed fire treatment data. Prescribed burning was conducted 6 August 2013, indicated by arrow. Treatments within a date followed by the same letter are not different (P > 0.05). Mechanical and chemical treatments were reapplied immediately following each measurement date.
Volume 37 (3), 2017 Natural Areas Journal 389 Length of Longest Stems as compared to 0.4 (± 0.13) for the control responded inconsistently over the course of group (Figure 4). the study. Increased leaf production would At the immediate pre-burn sampling date provide plants with more photosynthetic (July 2013), individuals in the prescribed tissue, augmenting their ability to capture Number of Fruits burn treatment had longer stems (tdf<118 = sunlight, and likely enhancing their overall 3.24, P < 0.01; 161.8 ± 7.63) than any other fitness. However, flowering does not occur The only sampling date on which num- treatment (Figure 2). However, those stems uniformly throughout the year—generally ber of fruits produced was sufficient for were removed with the prescribed burning occurring between early March and June, analysis was April 2014. At that time, treatment, and by September 2013 there then sporadically thereafter with rainfall treatments affected (F = 3.34, P = were no differences in length of longest 3,117 (USFWS 1985). Therefore, it is likely that 0.02) fruit numbers: there were more fruits stem. The following spring, individuals environmental conditions were not condu- (t = 3.21, P = 0.003) in the mechanical in the burn treatment regrew longer stems 60 cive to flowering during those times. But removal treatment (25.2 ± 1.91) than in than in any other treatment (April 2014: at times when flowering did occur, at least the control treatment (6.6 ± 1.97) (data I < 5.29, P < 0.01); they remained one of the treated groups had more flowers df<129 not shown in figure). longer than both neighbor-removal treat- than the control group. This indicates that ment plants, and longer or no different than some degree of neighbor removal provides control plants throughout the remainder of CONCLUSIONS slender rush-pea the opportunity to be more the study (June 2014: Idf<125 < 3.08, P < reproductively active by enhancing its re- 0.01; July 2014: tdf<123 < 5.11, P < 0.01). Mechanical and chemical treatments were productive potential. Lack of differences Additionally, plants in the control group initiated in July 2012, and reapplied in among treatments observed in July 2014 had longer (t122 = 2.89, P = 0.005) stems October 2012 and in May and July 2013; more than likely represents environmental than those in the mechanical removal of implementation of prescribed burning conditions related to long-term lack of neighbors’ treatment by July 2014, likely was delayed until August 2013 because rainfall onsite, combined with heat stress. an etiolation response. of a county-wide burn ban. It is also true, however, that prior to our experiment this In Kansas, Kettle et al. (2000) found as study area had been managed with periodic many as 3.3 times the number of flowering Number of Leaves on Longest Stem mowing for decades. During our study, stems occurred in Mead’s milkweed (Ascle- we recorded plant responses on the same pias meadii Torr ex A. Gray), a long-lived Immediately before summer prescribed dates for all treatments that were in place perennial forb, after the prairie was burned burn was implemented, there was a dif- on those dates. Thus, it is reasonable to as compared to years without burning. ference (July 2013: F = 7.1, P < 0.01; 3,546 compare morphological characteristics of Additionally, flowering stems produced Figure 3) among treatments with burn and our plants on the same date and to attribute more flowers per inflorescence in years chemical treatments having more leaves differences to treatments that plants had when the prairie was burned than years of than mechanical treatment and control experienced prior to each sampling date. no burning. Similarly, Towne and Knapp plants. Until the last sampling date when Finally, our intention was not to develop (1996) found that total legume density was there were no significant differences, at management recommendations, but rather higher in annually burned watersheds than least one treatment group displayed more to assess whether and how slender rush- unburned watersheds in tallgrass prairie. leaves on the longest stem than control pea responds morphologically to neighbor group plants. By June 2014, individuals in removal. A plant’s competitive abilities should the prescribed burning treatment had more increase with increased biomass (Keddy (t < 3.03, P = 0.01) leaves on the df<433.6 Our results suggest that manipulating the 1990; Bonser and Reader 1995). Towne longest stem (27.8 ± 4.4) than any other vegetation matrix in which slender rush-pea and Knapp (1996) found that while total group. One month later, all experimental exists has positive morphological effects on legume biomass was similar between burn groups exhibited a decline in mean number it. There is, however, variability in degree treatments, biomass production of white of leaves on the longest stem, at approxi- and timing of response to treatments among prairie clover (Dalea candida Michx. ex mately five for all groups. the variables we considered. It appears as Willd.) and slimflower scurfpea Psoralea( though summer prescribed burning has a tenuiflora Pursh) was higher in burned Number of Flowers more immediate and consistent effect on than unburned areas, yet other forbs in number of main stems and length of the their study responded negatively to annual Number of flowers showing color was longest stem. Further monitoring could burning. In their study, legumes comprised significantly different at only two sampling elucidate whether these results would be approximately only 11% of forb biomass dates (September 2013: F3,478.1 = 8.11, maintained over a longer term. in unburned tallgrass prairie, but made up P = 0.01; April 2014: F3,136.4 = 13.57, P roughly 25% of forb biomass in annually < 0.01)—most notably April 2014 when Although flowering also responded quickly burned prairie. While we did not evaluate plants with chemically removed neighbors to prescribed fire, flower production and biomass, our results show that slender had an estimated mean 3.9 (± 0.98) flowers number of leaves on the longest stem both rush-pea individuals in the prescribed fire
390 Natural Areas Journal Volume 37 (3), 2017 treatment had longer stems than other other conditions that preclude burning. Slender rush-pea demonstrated morpho- treatments by spring following the burning In a central Texas study that compared logical plasticity under both prescribed treatments. Pre-burn (July 2013) of plants means of controlling a closely related burning and removal of influence from to be burned indicated that they had longer nonnative, invasive grass, King Ranch competitive neighbors. This is an important leaves and more stems than control plants bluestem (Bothriochloa ischaemum [L.] mechanism by which plants adapt to both on that sampling date. Although we have Keng var. songarica [Rupr. ex Fisch and biotic and abiotic changes in environment, no explanation for this initial difference, C.A. Mey.] Celarier and Harlan), canopy and it has the potential to allow for per- it is also true that post-burn differences cover was reduced by both early and late petuation of the adapting species (Craine between burn plants and control plants growing-season prescribed fires, while it and Dybzinski 2013). The ability to exploit generally exceeded this initial difference, had a mixed response to glyphosate based ever-changing resources can also provide suggesting a true fire effect. Unlike the on site and rate, and no response to mowing a means for species coexistence (Aarssen other treatments, both prescribed fire plants (Simmons et al. 2007). 1983). Slender rush-pea’s demonstrated and control plants were still surrounded by ability to alter its growth form in response aboveground neighbors. Thus, they would to manipulation of its environment could Fire was historically a common disturbance benefit by vertically elevating themselves allow it to persist with Kleberg bluestem at on the coastal prairies of South Texas. in the canopy—providing better access to this site. Complete eradication of nonnative Most immediately, prescribed fire prunes sunlight and potentially augmenting their grasses is desirable, but unlikely; control existing foliage and removes litter, but it competitive abilities (Hangarter 1997). of these species is a more reasonable goal also serves to rejuvenate and invigorate and may be sufficient for sustaining rare plant growth—partially by increasing nu- Low mortality among treatments through- native plant populations. out the study indicates that slender rush-pea trients available for plant use for a short likely has a life span longer than three time (Vieira et al. 1996). Survival of plants after fire also depends on the positioning years, and that it is a hardy plant that has ACKNOWLEDGMENTS been able to survive with competition of their underground parts (Rowe 1983). from Kleberg bluestem under a mowing Where fires are common, many plants have the ability to regenerate aboveground The authors wish to thank the St. James regime. Our treatments affected number Catholic Cemetery in Bishop, Texas, for of seed pods on only one sampling date; photosynthetic tissue by sprouting from un- derground organs (Malanson and Trabaud graciously and enthusiastically allowing us however, it is possible that fire might to conduct research on the property, and affect germinability: some members of 1988; Paula and Ojeda 2009). The deep, well-protected buds of some legumes are John Reilley and Shelley Maher of the the Fabaceae family are known to exhibit USDA NRCS Kika de la Garza Plant Ma- enhanced germinability after fire, which ideal for regeneration after fire (McLean 1969; Rowe 1983). Our results represent terials Center for supporting this project. helps break dormancy of their hard seeds We thank Adam Toomey, Mylea Lovell, (Silveira and Overbeck 2013). This project trends only one year after a single fire; fur- ther monitoring and analyses of long-term Michael Golla, Luis Bartolo, John Clark, was not designed to answer such questions; and Anastasia Krainyk for field and techni- we recommend future examination. effects of repeated fires are recommended to determine the ideal fire return interval cal assistance. Drs. Timothy Fulbright and Michael Tewes reviewed and improved an Herbicide as a control of invasive, nonna- (FRI) in the context of climate variability. More than likely, the historical FRI of four earlier draft of this paper. We gratefully tive grasses is not a viable option at this site. acknowledge financial support from Caesar A nonselective herbicide would harm both to eight years (Wright and Bailey 1982) would be ideal, depending on rainfall, fuel Kleberg Wildlife Research Institute, US slender rush-pea and South Texas ambrosia, Fish and Wildlife Service, Rene Barrientos, and a grass-specific herbicide may further conditions, and effects of repeated fire on slender rush-pea populations. and the South Texas Quail Coalition. This reduce or eliminate the remaining native is publication number 15-120 of the Caesar grasses at the site. Historically, the site has Kleberg Wildlife Research Institute. been managed with mowing, which may We also recommend more long-term have benefitted (D’Antonio and Vitousek studies on control of Kleberg bluestem 1992; McDonald and McPherson 2011) throughout the native range of slender slender rush-pea by reducing vertical rush-pea, and safeguarding the genetic competition and biomass. Our results in- stock of slender rush-pea through expan- Ashley McCloughan is a Field Research dicate prescribed fire may be as, or more, sion of propagation begun at NRCS Kika Scientist with Actagro Soil and Plant beneficial to continued health of the slender de la Garza Plant Materials Center in Health Technology Solutions. At the time of rush-pea populations at this or similarly Kingsville, Texas, plus additional efforts this research she was a graduate research invaded sites as mowing. However, both to establish a reliable ex situ population assistant, Caesar Kleberg Wildlife Re- our data and the fact that slender rush-pea with enough genetic diversity to ensure search Institute and Department of Animal, has persisted many years in this site with long-term species survival (Trusty et al. Rangeland and Wildlife Sciences, Texas continued mowing indicate mowing is a 2009) and protection against demographic A&M University-Kingsville, where she viable substitute during burn bans and or genetic collapse. earned an MS degree in Wildlife Science.
Volume 37 (3), 2017 Natural Areas Journal 391 Sandra Rideout-Hanzak is a Research plants as ecosystem dominants. Weed Tech- disturbance in three resprouter species of Scientist and Associate Professor, Caesar nology 18:1283-1287. the genus Erica. Botany 87:253-259. Kleberg Wildlife Research Institute and Fravolini, A., K.R. Hultine, E. Brugnoli, R. Poole, J.M., and G.K. Janssen. 1996. Managing Department of Animal, Rangeland and Fazal, R.B. English, and D.G. Williams. and monitoring rare and endangered plants Wildlife Sciences, Texas A&M Universi- 2005. Precipitation pulse use by an invasive on highway rights-of-way in Texas. Pp. 8-12 woody legume: The role of soil texture and in J. Maschinski, H.D. Hammond, and L. ty-Kingsville, Kingsville, Texas. She is a pulse size. Oecologia 144:618-627. Holter, tech. eds., Southwestern Rare and fire ecologist and restoration ecologist Hangarter, R.P. 1997. Gravity, light and plant Endangered Plants: Proceedings of the who earned a Ph.D. degree from Stephen form. Plant Cell and Environment 20:796- Second Conference, Flagstaff, Arizona, F. Austin State University. 800. September 11–14, 1995. RM-GTR-283, USDA Forest Service, Fort Collins, CO. Keddy, P.A. 1990. Competitive hierarchies and David B. Wester is a Research Scientist centrifugal organization in plant communi- Rice, P.M., J.C. Toney, D.J. Bedunah, and C.E. and Professor, Caesar Kleberg Wildlife Re- ties. Pp. 265-290 in J. Grace, ed., Perspec- Carlson. 1997. Plant community diversity search Institute and Department of Animal, tives on Plant Competition. Academic Press, and growth form responses to herbicide Rangeland and Wildlife Sciences, Texas San Diego, CA. applications for control of Centaurea maculosa. Journal of Applied Ecology A&M University-Kingsville, Kingsville, Kettle, W.D., H.M. Alexander, and G.L. Pittman. 34:1937-1412. Texas. He is a statistical ecologist with a 2000. An 11-year ecological study of a rare Rowe, J.S. 1983. Concepts of fire effects on Ph.D. degree from Texas Tech University. prairie perennial (Asclepias meadii): Impli- cations for monitoring and management. plant individuals and species. Pp. 135-154 American Midland Naturalist 144:66-77. in R.W. Wein and D.A. MacLean, eds., Weiman Xi is an Assistant Professor, The Role of Fire in Northern Circumpolar Department of Biological Sciences, Texas Kirk, R.E. 2013. Experimental Design: Proce- Ecosystems. John Wiley & Sons, New York. A&M University-Kingsville, Kingsville, dures for the Behavioral Sciences, 4th ed. Sage Publishing, Los Angeles, CA. Ruffner, M.E., and T.G. Barnes. 2012. Eval- Texas. His research interest is in plant uation of herbicide and disking to control ecology and plant taxonomy. He earned Littell, R.C., G.A. Milliken, W.W. Stroup, R.D. invasive bluestems in a south Texas coastal Wolfinger, and O. Schabenberger. 2006. SAS a Ph.D. degree from University of North prairie. Rangeland Ecology and Manage- for Mixed Models, 2nd ed. SAS Institute, ment 65:277-285. Carolina at Chapel Hill. Cary, NC. Sampson, F., and F. Knopf. 1994. Prairie con- Malanson, G.P., and L. Trabaud. 1988. Vigour servation in North America. BioScience Quercus coccifera of post-fire resprouting by 44:418-421. L. Journal of Ecology 76:351-365. LITERATURE CITED Shapiro, S.S., and M.B. Wilk. 1965. An analysis McDonald, C.J., and G.R. McPherson. 2011. of variance test for normality (complete Aarssen, L.W. 1983. Ecological combining Absence of a grass/fire cycle in a semiarid samples). Biometrika 52:591-611. ability and competitive combining ability in grassland: Response to prescribed fire and plants: Toward a general evolutionary theory grazing. Rangeland Ecology and Manage- Silveira, F.S., and G.E. Overbeck. 2013. Ef- of coexistence in systems of competition. ment 63:384-393. fect of high temperature on germination American Naturalist 122:707-731. of four legumes from a forest–grassland McLean, A. 1969. Fire resistance of forest spe- mosaic in southern Brazil. Biota Neotropica Anderson, M.J. 2001. A new method for cies as influenced by root systems. Journal 13:331-335. non-parametric multivariate analysis of of Range Management 22:120-122. variance. Austral Ecology 26:32-46. Simmons, M.T., S. Windhager, P. Power, J. Lott, Myers, R.H., D.C. Montgomery, and G.G. R.K. Lyons, and C. Schwope. 2007. Selective Bonser, S.P., and R.J. Reader. 1995. Plant Vining. 2002. Generalized Linear Models and non-selective control of invasive plants: competition and herbivory in relation to with Applications in Engineering and the The short-term effects of growing-season vegetation biomass. Ecology 76:2176-2183. Sciences. John Wiley and Sons, New York. prescribed fire, herbicide, and mowing in Burnham, K.P., and D.R. Anderson. 2002. NOAA. 2014. Precipitation for 78363, Kings- two Texas prairies. Restoration Ecology Model Selection and Multimodel Inference, ville, TX July 2012 through July 2014. 15:662-669. 2nd ed. Springer-Verlag, New York. National Oceanic and Atmospheric Ad- Simpson, B.B., J.A. Tate, and A. Weeks. ministration. Accessed from
392 Natural Areas Journal Volume 37 (3), 2017 Towne, G.E., and A.K. Knapp. 1996. Biomass wildlife and plants; Listing Hoffmannseg- CFR, Part 17.12. and density responses in tallgrass prairie gia tenella as an endangered species. US USFWS. 2008. Slender rush-pea (Hoffmannseg- legumes to annual fire and topographic Fish and Wildlife Service, Federal Register gia tenella), 5-year review: Summary and position. American Journal of Botany 50:45614-45618. evaluation. US Fish and Wildlife Service, 83:175-179. USFWS. 1988. Slender rush-pea (Hoffmannseg- Corpus Christi, TX. gia tenella) Recovery Plan. US Fish and Trusty, J.L., I. Miller, R.S. Boyd, L.R. Goert- Wildlife Service, Albuquerque, NM. Vieira, E.M., I. Andrade, and P.W. Price. 1996. zen, V.C. Pence, and B.L. Plair. 2009. Ex Fire effects on a Palicourea rigida (Rubia- situ conservation of the federally endan- USFWS. 1994. Endangered and threatened ceae) gall midge: A test of the plant vigor wildlife and plants; Determination of endan- gered plant species Clematis socialis Kral hypothesis. Biotropica 28:210-217. gered status for the plants Ayenia limitaris (Ranunculaceae). Natural Areas Journal (Texas Ayenia) and Ambrosia cheiranthifo- Wright, H.A., and A.W. Bailey. 1982. Fire Ecol- 29:376-384. lia (South Texas Ambrosia). US Fish and ogy: United States and Southern Canada. USFWS. 1985. Endangered and threatened Wildlife Service, Federal Register 59:50 John Wiley & Sons, New York.
Volume 37 (3), 2017 Natural Areas Journal 393
View publication stats