Recovery of the Price’s -Bean ( priceana) at Sauta Cave National Wildlife Refuge,

Final Report

Submitted to:

Bill Gates Wheeler National Wildlife Refuge 2700 Headquarters Road Decatur, AL 35603

Prepared by:

Alfred Schotz Alabama Natural Heritage Program 1090 South Donahue Avenue Auburn University, AL 36849

September 30, 2016

TABLE OF CONTENTS

TABLE OF CONTENTS ...... ii INTRODUCTION ...... 1 METHODOLOGY ...... 3 RESULTS AND DISCUSSION ...... 5 LITERATURE CITED ...... 11 APPENDIX A ...... 12 Table 1: Coordinates for corner stakes ...... 12 Table 2: Coordinates for actively reproducing individuals of ...... 13 Table 3: Percent overstory measurements ...... 13

LIST OF FIGURES Figure 1: of Apios priceana ...... 1 Figure 2: Fruit of Apios priceana...... 1 Figure 3: Map: Location of Apios priceana monitoring plots ...... 3 Figure 4: Stake used to mark corners of plots ...... 4 Figure 5: Tag and flag marking location of ...... 4 Figure 6: Chart: Total stem counts for 2015 and 2016 ...... 6 Figure 7: Chart: Total number of reproducing for 2015 and 2016 ...... 7 Figure 8: Chart: Annual percent of reproduction ...... 7 Figure 9: Chart: Percent reproduction for plots Control 1 and Treatment 1 ...... 8 Figure 10: Map: Plots C1 and T1 depicting locations of reproductively active plants and canopy cover measurements ...... 10

Sauta Cave Apios priceana Monitoring Study ii

INTRODUCTION

Auburn University, through the Alabama Natural Heritage Program, entered into an agreement with the U.S. Fish and Wildlife Service to assist with recovery efforts of the Price’s potato-bean (Apios priceana) at Sauta Cave National Wildlife Refuge (SCNWR), in Alabama. The species was listed as federally threatened by the U.S. Fish and Wildlife Service in 1989, and is currently ranked as G3 by NatureServe suggesting it to be globally vulnerable (NatureServe 2016). The species was first brought to the attention of the scientific community by Sadie F. Price from woods and thickets near Bowling Green, , in July 1896 (Robinson 1898). At the time of listing, the species had been collected from 21 sites in Alabama, , Kentucky, , and , with only 10 extant occurrences having been known and with 60% of those threatened by destruction (Woods 2005). The greatest concentration of occurrences was, and continues to be, from Kentucky and Tennessee. Apios priceana is considered historic in Illinois where the species has not been relocated despite repeated searches since it was last observed in the state from Union County, in 1977 (U.S. Fish and Wildlife Service 1993). The species is an herbaceous , climbing to roughly 5 m from a large thickened tuber. The inflorescences occur on 1-3 racemes, each 5-15 cm long, born in the axils. Each is approximately 1 cm long, greenish-white tinged with pinkish-purple or maroon, and quite bean- like (Figure 1). The standard is elongated and in-rolled on itself at the apex, fleshy and beak- like. The are 2-3 dm long, alternate and pinnately compound with 5-9 leaflets; leaflets narrow to broadly ovate, entire and acuminate. The fruit is a cylindrical legume 13-20 cm long, similar in appearance to the garden green bean (Figure 2). Apios priceana can be separated by the similar and more abundant A. americana by its large, single tuber as compared to the smaller multiple tubers produced by A. americana. The relatively large pink flowers typical of A. priceana differ from the smaller, typically maroon, unappendaged flowers of A. americana. The fruit of A. priceana is typically longer, with the smallest equaling the length of the largest fruits produced by A. americana (Robinson 1898, Kral 1983, U.S. Fish and Wildlife Service 1993).

Figure 1. Flowers of Apios priceana. Figure 2. Fruit of Apios priceana.

Sauta Cave Apios priceana Monitoring Study 1

The genus consists of three Asian species and two American species (Woods 1988, Woods 2005). Flowers of A. priceana bloom from late June through August, occasionally as late as September (Kral 1983, Mahler 1970). Early reports by the discoverer of the species, Sadie Price, suggested that the species does not frequently set fruit (Robinson 1898). Potential pollinators include butterflies Papilio glaucus (tiger swallowtail), Eudamus tityrus (silver spot skipper), and Urbanus proteus (long-tailed skipper), as well as honeybees (Apis mellifera), and bumble (Bombus spp.) (Robinson 1898, U.S. Fish and Wildlife Service 1993, Woods 2005, Schotz pers. obs.). Flowers in the genus Apios have a tripping mechanism that causes the keel to coil when triggered by an insect. When the keel coils, it exposes the anthers and pistil, allowing to occur (Bruneau and Anderson 1988). Common names for Apios priceana include Price’s potato-bean, Sadie Price’s potato-bean, potato-bean, and Price’s ground-nut.

Apios priceana is an inhabitant of open, mixed-hardwood forests, forest edges and clearings on river bottoms and ravines, being unable to tolerate deep shade (Kral 1983). The species prefers well-drained loams on calcareous substrates, frequently occurring on rocky slopes in the northern portion of its range. The occurrence at SCNWR falls within the Cumberland Plateau of north Alabama where it inhabits bouldery limestone slopes under a filtered hardwood canopy primarily consisting of Quercus muehlenbergii, Quercus shumardii, Carya glabra, Fraxinus americana, Liriodendron tulipifera, Cercis canadensis, Ulmus rubra, Ostrya virginiana, and Acer barbatum. Eastern red cedar (Juniperus virginiana), the only conifer present, is widely scattered among the hardwoods. Principal shrub and herbaceous associates include Hypericum frondosum, Forestiera ligustrina, Frangula caroliniana, Campanula americana, Spigelia marilandica, Verbesina virginica, Rudbeckia triloba, and Solidago auriculata.

The occurrence at SCNWR was discovered by Al Schotz on August 9, 2002. At the time of discovery, the site contained greater than 150 individuals in relatively undisturbed, high quality habitat. In accordance with NatureServe’s ranking criteria the occurrence was issued a conservation rank of A, the highest rank given suggesting excellent viability and habitat integrity. In 2011, a second series of census data was acquired as part of a statewide status assessment of the species, during which a relatively large number of flowering plants were also reported. However, USFWS biologist Bill Gates has noted through casual observations of the plant for the last five or more years that the number of those flowering is generally low and that some plants have disappeared. Only one individual of 100-125 plants located in 2014 during a half-day joint survey effort between Bill Gates and Tennessee Valley Authority staff produced flowers. The species does not flower every year and this may partially explain the discrepancy in the proportion of flowering and non-flowering plants. For comparison, Bill Gates had examined an occurrence at Redstone Arsenal within two weeks of the survey at SCNWR and observed that the majority of plants within a power line corridor and lightly canopied areas of the adjacent woodland were healthy and profusely flowering. He also observed that inflorescences on plants at the arsenal were also much larger than the one in flower on the refuge. However, it is unknown whether different populations flower in different years.

It was hypothesized that the high level of canopy closure in the location where A. priceana has disappeared on SCNWR is the cause and has also contributed to decreased reproduction. The amount of canopy closure in the area of plant disappearance has been observed numerous times during the course of several years by Bill Gates and has shown a marked increase to nearly 100% coverage. Occurrences having the greatest vigor appear to be often associated with clearings in forests and along rights-of way, including roadsides and power lines. As of this study, it is

Sauta Cave Apios priceana Monitoring Study 2

unknown how much canopy opening is needed to promote optimal growth and reproduction of A. priceana.

The premise of this study is to determine the ideal canopy cover necessary to promote and maintain optimal growth and reproduction of A. priceana at SCNWR. It is anticipated the results of the project will have broader applications, serving as a resource guide as it relates to management across the range of the species.

METHODOLOGY

Four plots within SCNWR known to contain Apios priceana were selected for the study, comprised of two treatment and two control plots (Figure 3) delineated by marking corners with permanent, orange plastic stakes (Figure 4) and recorded using Global Positioning System (GPS) capabilities set at WGS 84 with sub-meter accuracy. Coordinates for corner stakes appear as

Figure 3. Location of Apios priceana monitoring plots at Sauta Cave National Wildlife Refuge. Plots are identified as “C” for control and “T” for treatment.

Table 1 in Appendix A. Each plot consists of 0.25 – 1.0 acres, and was selected to represent similar habitat conditions among the four plots. Treatment plots entailed reducing the canopy by 100% in Treatment 1 (T1) and by 50-75% in Treatment 2 (T2), whereas the control plots remained unchanged. Pre- and post-treatment canopy closure measurements were randomly placed to best represent overall canopy coverage within each plot and were determined using a spherical densiometer with the location of points recorded using GPS technology with sub-meter accuracy to allow for easily replicated, ongoing monitoring. Overstory measurements and

Sauta Cave Apios priceana Monitoring Study 3 coordinates appear in Appendix A as Table 3. Census data were gathered and reported on field forms specifically designed for the study, tabulating non-flowering and flowering plants and further refining levels of reproduction by recording the number of inflorescences and fruit (beans) present. Plants were identified and permanently marked in the field with metallic numbered tags and different colored flags (Figure 5) to denote levels of plant development, where pink = flowering, orange = non-flowering, and red = plants reproducing the previous year that showed no vegetative growth the following year. Flags were switched on some plants the second year of the study to indicate a change of growth status from non-flowering to flowering and from flowering to non-flowering, and to identify flowering individuals that failed to reappear the second year.

Figure 4. Example of stake used to mark Figure 5. Example of tag and flag used to corners of plots. mark location of plants.

History of Treatment Plots (Contributed by Bill Gates, USFWS) Work to remove all of the canopy in T1 was largely completed before this study. A Tennessee Valley Authority contractor removed the majority of the canopy via drum chopping in late winter 2015 as a cooperative project with the refuge. On July 10, 2015, additional small trees (up to 4” DBH) were hacked and squirted (HS) up to the toe of the slope (along the southern border) by refuge staff. In addition, 5-10 larger trees (6-8” DBH) were removed with a chain saw. Refuge staff was trying to create nearly 100% canopy removal from the main gravel road into the refuge on the north, the toe of the slope on the south, and extending from a gravel pile in the west to the western base of the road leading to the upper cave entrance. They decided to leave one very large (20-24” DBH) yellow buckeye (Aesculus flava) in the eastern half of the plot because it was too large to remove with their chain saw and they did not want to create a snag via HS that could fall across the main road. About half of the small trees HS along the toe of the slope died by the time the project had started.

Refuge staff did not succeed in removing the proper amount of canopy from T2 during the entirety of this study (Table 3). On July 10, 2015, ten very large trees (>14” DBH) (Large trees Sauta Cave Apios priceana Monitoring Study 4 were purposely selected to benefit Indiana, and, possibly, other bats – research in the Talladega National Forest indicated Indiana bat summer roosts/maternity colonies were in large, dead hardwood trees with flaking bark. A few Indiana bats hibernate in Sauta Cave and may be using the forest onsite in the summer.) along the unmaintained road leading to the upper cave entrance were HS with 100% AquaStar (glyphosate). This was intended as an effort to determine how canopy removal would affect A. priceana plants growing along the road independent of the two treatment plots (T1 and T2). However, part of the road in this area had to be incorporated into T2 due to the existing distribution of A. priceana. An effort was made to include A. priceana in the two treatment and two control plots of the main study. The canopy of three of the ten large trees HS on July 10 would influence the amount of light that would fall on T2. These ten trees did not die during the course of this study. By the 2016 part of the study some leaves were not present, but this did not impact the amount of light hitting the ground.

On September 24, 2015, additional large trees in T2 were HS with 100% AquaStar. This was after census data were gathered in 2015. By late spring 2016 these trees generally had not died nor lost most of their leaves. On July 19, 2016, nearly all of the trees in T2 treated earlier, including those along the road, were re-treated with Garlon 3A. In addition, larger trees and a few smaller trees were treated. The smaller, subcanopy trees were treated because their canopy would also reduce the amount of light reaching the forest floor. Very few trees, and none of the larger ones, had died nor lost many leaves prior to the end of the study. Thus, percent canopy cover was not impacted.

Supplemental spraying was conducted in T1 during 2015 and 2016 to control Nepalese brown- top grass (Microstegium vimineum), kill small trees that had re-sprouted in the plot, and to control additional competing vegetation such as sericea lespedeza (Lespedeza cuneata). This additional spraying did not affect the canopy and served to prevent canopy closure from re- occurring.

RESULTS AND DISCUSSION

Census data gathered in August 2016 from the four plots report a total of 140 plants, representing an overall decrease of 23% from a total of 182 plants in 2015 (Figure 6). Similarly, an overall decrease of reproductively active (flowering and fruiting) plants was reported, showing a reduction of 53% from 15 stems in 2015 to 7 stems in 2016 (Figure 7), with further calculations showing a decrease in reproduction of 8.2% to 5.0% in relation to the total number of plants for 2015 and 2016, respectively. These calculations mark a striking contrast to census data acquired by Bill Gates in 2014 where one plant was observed reproducing among 100-125 plants, representing a level of reproduction less than 1% (Figure 8). Treatment 1 (T1) experienced the greatest decrease in total number of plants at 55%, whereas T2 was the only plot to experience an increase of plants at 22%. Thirteen plants having reproduced in 2015 did not produce flowers in 2016, including two plants that did not reappear. Two plants have been documented as reproducing both years of the study. Reproduction of A. priceana was documented from only Plots C1 and T1, with T1 having the highest level of reproduction in 2015 and 2016 at 14.0% and 28.6%, respectively, whereas C1 experienced the lowest level of reproduction each year at 8.5% and 1.3% (Figure 9). Figure 10 shows Plots C1 and T1 depicting the locations of reproductively active plants alongside points showing the most recent (as of 2016) canopy cover measurements. No flowering or fruiting plants were observed in Plots C2 and T2. To facilitate relocation of reproducing plants, coordinates are provided in Appendix A, Table 2. The factors contributing to the significant decrease of plants in Plot T1 and the overall decline of plants witnessed in 2016 are unknown at this phase of the study, but may be attributable to regional

Sauta Cave Apios priceana Monitoring Study 5 climatic patterns. Abnormally low precipitation levels, specifically during the growing season, may have fostered high mortality among seedlings and juvenile plants. Further, the significant reduction of plants in T1 may also be partially influenced by high level canopy removal, where, in combination with unusually dry weather during the project period, increased sunlight and higher temperatures may have contributed to atypical levels of mortality.

To evaluate the influence canopy cover has on the reproductive capacity of the species, efforts were made to reduce overstory coverage of the two treatment plots in 2015 and 2016. Canopy coverage in T1 was significantly reduced, largely through drum chopping in early 2015, prior to flowering of A. priceana and field studies of this project. A treatment of glyphosate in 2015 and Garlon 3A in 2016 was applied to T2 in an attempt to reduce foliar coverage to 25-50%. No discernable reduction in the overstory was detected between T2 and the two control plots in 2015 and 2016, as a longer period of time is necessary for the chemicals to take effect. Table 3 in Appendix A presents percent cover measurements recorded at selected waypoints within each of the four monitoring plots.

Figure 6. Total stem counts for 2015 and 2016.

Sauta Cave Apios priceana Monitoring Study 6

Figure 7. Total number of reproducing plants for 2015 and 2016.

Figure 8. Annual percent of reproduction of total population.

Sauta Cave Apios priceana Monitoring Study 7

Figure 9. Percent reproduction for plots Control 1 (C1) and Treatment 1 (T1) in 2015 and 2016.

Success of the Project The overall goal of the project to determine the level of canopy cover for achieving optimal growth and reproduction of Apios priceana, could not be entirely accomplished. The greater number of reproducing plants in T1 relative to C1 during both years of the study and the ratio of reproducing plants to vegetative plants in T1 between 2015 and 2016 presents a significant difference necessitating further explanation. Six reproductively active plants were documented in T1 during each year of the study, highlighting an annual increase of approximately five plants from census data acquired by Bill Gates in 2014 (an approximation is provided, as plot boundaries were not delineated in 2014). This increase in reproducing plants may be correlated to canopy removal efforts in early 2015, prior to the inception of this study. Similarly, increased light levels may have also fostered elevated germination rates of soil seed reserves in 2015, accounting for a higher incidence of A. priceana seedlings and subsequently a higher ratio of reproducing to vegetative plants. Conversely, the abnormally dry weather of 2016 in conjunction with increased solar insolation may have contributed to a high mortality rate among seedlings in T1, substantiating a significantly lower ratio of reproducing to vegetative plants. Moreover, the increase of reproducing plants in T1 in relation to C1 during the project period may also be correlated to canopy reduction efforts.

The level of canopy coverage between T2 and the two control plots in relation to herbicide application showed no discernible difference during the course of this study. Overall, the two- year timeframe allocated for the project allowed insufficient time to assess the full effects of canopy removal upon the viability of the species. While efforts of select overstory removal in 2015 and 2016 are showing the initial phases of mortality to trees having received injections of either glyphosate or Garlon 3A, no detectable leaf drop was apparent during this study. A

Sauta Cave Apios priceana Monitoring Study 8

minimum of five years is recommended to allow for complete mortality and to better determine how a change in percent canopy cover will affect the long-term viability of A. priceana.

Future Needs To acquire a competent understanding how percent canopy cover affects the growth and reproduction of A. priceana, it is encouraged this study continue for a minimum of an additional three years. If extended, a correlation of reproduction in relation to climatic conditions (e.g., drought) can also be examined. As emphasized earlier, the outcome of the project will have broader applications, serving as a resource guide as it relates to management across the range of the species.

Sauta Cave Apios priceana Monitoring Study 9

Figure 10. Plots Control 1 (C1) and Treatment 1 (T1) depicting locations of reproductively active plants and canopy cover measurements.

Sauta Cave Apios priceana Monitoring Study 10

LITERATURE CITED

Boyd, R.S. and N. Paris. 2012. Annual report of research activities (2011) on the Price’s potatobean (Apios priceana) population at Madkin Mountain, Redstone Arsenal, AL. 6 pp.

Bruneau, A. and G.J. Anderson. 1988. Reproductive biology of diploid and triploid (Leguminosae). American Journal of Botany 75: 1876-1883.

Flora of North America Editorial Committee, eds. 1993+. Flora of North America North of Mexico. 15+ vols. New York and Oxford.

Kartesz, J.T. 2011. Draft of BONAP’s Floristic Synthesis of North America. In: Kartesz, J.T. (ed.). Floristic synthesis of North American flora. Version 1.0. Unpublished draft.

Kral, R. 1983. A report on some rare, threatened, or endangered forest-related vascular plants of the South. USFS Tech. Publ. R8-TP 2, Atlanta, GA. Vol. 1: 718 pp.

Mahler, W.F. 1970. Manual of the legumes of Tennessee. J. Tennessee Acad. Sci. 45 (3): 6596.

Medley, M.E. 1980. Status report on Apios priceana. Unpublished USFWS report, Region IV, Newton corners, MA. 24 pp.

NatureServe. 2016. NatureServe Explorer: An online encyclopedia of life [web application]. Version 7.1. NatureServe, Arlington, Virginia. Available http://explorer.natureserve.org. (Accessed: September 7, 2016).

Ostlie, W. 1990. Element Stewardship Abstract for Apios priceana – Price’s potato-bean. Unpublished report for The Nature Conservancy, Arlington, VA. 10 pp.

Robinson, B.L. 1898. A new species of Apios from Kentucky. Bot. Gazette 25: 450-453.

U.S. Fish and Wildlife Service. 1993. Recovery Plan for Apios priceana. Jackson, MS. 43 pp.

Weakley, A.S. 2012. Flora of the southern and mid-Atlantic states (http://www.herbarium.unc.edu/flora.htm). Working draft of 30 November 30 2012. University of North Carolina Herbarium, North Carolina Botanical Garden, Chapel Hill.

Woods, M. 1988. A revision of Apios and Cochlianthus (Leguminosae). Doctoral dissertation. Southern Illinois University, Carbondale, IL.

Woods, M. 2005. A revision of the North American species of Apios (). Castanea 70 (2): 85-100.

Sauta Cave Apios priceana Monitoring Study 11

APPENDIX A

Table 1. Coordinates for corner stakes of each plot established to monitor Apios priceana at Sauta Cave National Wildlife Refuge. Plot Coordinates (WGS 84) of Corner Stakes 34.61642oN, -86.13105oE 34.61628oN, -86.13096oE Control #1 (C1 - Section 1*) 34.61617oN, -86.13109oE 34.61646oN, -86.13116oE 34.61620oN, -86.13082oE 34.61638oN, -86.13077oE Control #1 (C1 – Section 2*) 34.61601oN, -86.13041oE 34.61594oN, -86.13040oE 34.61613oN, -86.13086oE 34.61579oN, -86.13034oE Treatment #1 (T1) 34.61570oN, -86.13043oE 34.61605oN, -86.13093oE 34.61583oN, -86.13123oE 34.61605oN, -86.13145oE Control #2 (C2) 34.61616oN, -86.13123oE 34.61601oN, -86.13106oE 34.61556oN, -86.13007oE 34.61514oN, -86.13018oE Treatment #2 (T2) 34.61515oN, -86.13044oE 34.61556oN, -86.13050oE *Section 1 of Control #1 occupies either side of the walkway to the cave entrance and Section 2 of Control #1 occupies an area along the northeast side of the refuge road, opposite Treatment #1.

Sauta Cave Apios priceana Monitoring Study 12

Table 2. Coordinates for actively reproducing individuals of Apios priceana documented during 2015-2016 in the four study plots at Sauta Cave National Wildlife Refuge. Tag Number Plot Year Recorded Coordinates (NAD 83) 5 T1 2015 34.61605oN, -86.13093oE 9 T1 2015 34.61606oN, -86.13090oE 20 T1 2016 34.61607oN, -86.13087oE 29 T1 2015 34.61600oN, -86.13084oE 36 T1 2015, 2016 34.61602oN, -86.13086oE 37 T1 2016 34.61604oN, -86.13081oE 42 T1 2016 34.61603oN, -86.13084oE 45 T1 2016 34.61602oN, -86.13082oE 46 T1 2015 34.61602oN, -86.13081oE 48 T1 2015 34.61596oN, -86.13076oE 49 T1 2016 34.61598oN, -86.13078oE 3901 C1 2015 34.61623oN, -86.13109oE 3951 C1 2015 34.61623oN, -86.13081oE 3988 C1 2015 34.61618oN, -86.13103oE 3990 C1 2015 34.61624oN, -86.13154oE 3992 C1 2015, 2016 34.61644oN, -86.13114oE 3994 C1 2015 34.61611oN, -86.13102oE 3995 C1 2015 34.61643oN, -86.13112oE

Table 3. Percent overstory measurements for the four Apios priceana monitoring plots at Sauta Cave National Wildlife Refuge. Plot Coordinates (WGS 84) % Cover 2015 2016 34.61640oN, -86.13113oE 80 82 34.61626oN, -86.13109oE 83 82 34.61622oN, -86.13108oE 85 85 Control 1 (C1) 34.61625oN, -86.13076oE 90 90 34.61617oN, -86.13068oE 91 93 34.61611oN, -86.13059oE 90 90 34.61600oN, -86.13041oE 93 92 34.61609oN, -86.13089oE 19 21 34.61607oN, -86.13091oE 66 66 34.61603oN, -86.13078oE 19 20 Treatment 1 (T1) 34.61593oN, -86.13072oE 46 44 34.61585oN, -86.13054oE 26 26 34.61576oN, -86.13037oE 38 38 34.61622oN, -86.13098oE 92 94 Control 2 (C2) 34.61603oN, -86.13159oE 95 96 34.61594oN, -86.13155oE 94 95 34.61547oN, -86.13014oE 84 84 34.61536oN, -86.13037oE 92 92 Treatment 2 (T2) 34.61536oN, -86.13020oE 93 91 34.61524oN, -86.13038oE 96 95 34.61549oN, -86.13037oE 96 94

Sauta Cave Apios priceana Monitoring Study 13