Species Status Assessment of Guadalupe Fescue (Festuca Ligulata Swallen) Version 1.0

SPECIES STATUS ASSESSMENT OF GUADALUPE FESCUE (FESTUCA LIGULATA SWALLEN) VERSION 1.0

August 2016 Southwest Region U.S. Fish and Wildlife Service Albuquerque, NM

Guadalupe Fescue Species Status Assessment – Aug 2016

Prepared by Chris Best, Austin Ecological Services Field Office, with assistance from Hayley Dikeman, Stacey Stanford, and Nathan Allan (U.S. Fish and Wildlife Service), Dr. Joe Sirotnak (U.S. National Park Service), and Dr. Norma Fowler (University of Texas at Austin).

Suggested citation:

U.S. Fish and Wildlife Service. 2016. Species status assessment of Guadalupe Fescue (Festuca ligulata Swallen), Version 1.0. U.S. Fish and Wildlife Service Southwest Region, Albuquerque, New Mexico. 52 pp. + 3 appendices.

Guadalupe Fescue Species Status Assessment – Aug 2016

EXECUTIVE SUMMARY

Guadalupe fescue (Festuca ligulata) is a grass species found only in a few high mountains of the Chihuahuan Desert, in trans-Pecos Texas and Coahuila, Mexico. These “sky island” habitats are conifer-oak woodlands above 1,800 meters (5,905 feet (ft)) elevation. The species has been reported in only six sites. It was first collected in 1931 in the Guadalupe Mountains, Culberson County, Texas, and in the Chisos Mountains, Brewster County, Texas; these sites are now within Guadalupe Mountains National Park and Big Bend National Park, respectively. Guadalupe fescue was documented near Fraile, southern Coahuila, in 1941, in the Sierra la Madera, central Coahuila, in 1977, and two sites in the Maderas del Carmen mountains of northern Coahuila in 1973 and 2003. The last three sites are within Protected Natural Areas (Areas Naturales Protegidas, ANP) designated by the Mexican federal government.

Data collected by Texas Parks and Wildlife Department and Big Bend National Park reveals that Guadalupe fescue is a short-lived perennial with relatively low fecundity. The monitored population at Big Bend National Park has decreased significantly over time; this may be explained by a recruitment rate that is too low to sustain the population. Since grasses are wind- pollinated, small, widely scattered populations produce few if any seeds from out-crossing (pollination by unrelated individuals); the remaining individuals are probably highly inbred. We have not calculated a minimum viable population size for the species. However, by comparison to plants with similar life histories, we estimate that at least 500 to 1,000 individuals are necessary for long-term population viability.

To estimate the amount and distribution of potential Guadalupe fescue habitat, we created maps of conifer-oak forests in the Chihuahuan Desert at elevations greater than 1,800 m; since larger habitat areas may be more suitable, we restricted this model to areas greater than 200 hectares (ha) (494 acres (ac)). This model reveals that northern Mexico has 283 areas of potential habitat totaling 537,998 ha (over 1.3 million ac), compared to 20 such areas totaling 27,881 ha (68,894 ac) in Texas. Thus, about 95 percent of the potential habitat is in Mexico.

Wildfires occurred in the vicinity of the Chisos population at least 10 times between 1770 and 1940; the last major fire there was more than 70 years ago, due to fire suppression within the National Park. Periodic wildfire and leaf litter reduction may be necessary for long-term survival of Guadalupe fescue populations, although this has not been investigated. The long absence of fire and the resulting accumulation of fuels also increase the risk of more intense wildfire and potential loss of the remaining Guadalupe fescue population.

The Service, Big Bend National Park, and Guadalupe Mountains National Park established an updated Candidate Conservation Agreement on August 26, 2008. The objectives of this 10-year agreement include monitoring and surveys, seed and germ plasm banking, fire and invasive species management, trail management, educating staff and visitors, establishing an advisory team of species experts, and cooperation with Mexican agencies and researchers to conserve the known populations and search for new ones. Research objectives include investigations of fire ecology, habitat management, genetic structure, reproductive biology, and reintroduction.

Guadalupe Fescue Species Status Assessment – Aug 2016

Factors that may affect the continued survival of Guadalupe fescue include the genetic and demographic consequences of small population sizes and isolation of its known populations, changes in the wildfire cycle and vegetation structure, livestock grazing, trampling from humans and pack animals, trail runoff, competition from invasive species, the effects of climate change, such as higher temperature and changes in the amount and pattern of rainfall, and fungal infection of seeds.

Of six historically known populations, the species is currently known from two extant populations. The Chisos Mountains population is far smaller than our estimated minimum viable population level, and despite protection, appropriate management, and periodic monitoring by the U.S. National Park Service, it has declined between 1993 and 2014. The other extant population, at ANP Maderas del Carmen in northern Coahuila, Mexico, was probably viable as recently as 2003, and the site is managed for natural resources conservation. The population at Guadalupe Mountains National Park has not been seen in more than 60 years and is presumed extirpated. The status of three other populations, all in Coahuila, is unknown. It is possible that undiscovered populations of Guadalupe fescue exist in northern Mexico and that the overall status of the species may be better than we are currently able to determine. We assess the species’ future viability, under a range of scenarios, and conclude that viability will be influenced by the effectiveness of conservation and management of populations and habitats, public outreach and support, continued surveys to document the species’ existing populations and range, and the severity of climate changes.

iii

Guadalupe Fescue Species Status Assessment – Aug 2016

Table of Contents.

Executive Summary ii I. Introduction. 1 II. Species Information. 3 II.1. Morphology. 3 II.2. Taxonomy. 3 II.3. Genetic Diversity. 3 II.4. Phenology and Reproduction. 4 II.5. Survival Rates, Lifespan, and Demographic Trends. 5 II.6. Habitat. 7 II.7. Geographic Range. 10 II.8. Range of Estimated Potential Habitats. 14 II.9. Ecology and the Wildfire Cycle. 19 III. Summary of Individual, Population, and Species Requirements. 20 III.1. Requirements of Individuals. 20 III.2. Requirements of Populations. 21 III.3. Species Requirements. 22 IV. Factors Affecting the Survival of Guadalupe Fescue: Threats, vulnerabilities, and 24 conservation challenges. V. Conservation Efforts. 35 VI. Current Status and Viability. 36 VI.1. Current Viability. 38 VII. Assessment of Future Viability. 41 VIII. Literature Cited. 46 IX. Photographic Credits. 52 X. Acronyms Used. 52 XI. Scientific Units. 52 Appendix A. Glossary of Scientific and Technical Terms. 53 Appendix B. Analyses of Monitoring Plot Data at Big Bend National Park, Texas. 57 Appendix C. Cause and Effects Evaluations for Guadalupe Fescue. 70

Guadalupe Fescue Species Status Assessment – Aug 2016

List of Tables.

1. Guadalupe fescue population data at Big Bend National Park. 6 2. Plant species associated with Guadalupe fescue. 8 3. Historic and current populations of Guadalupe fescue. 12 4. Estimates of pine, oak, and pine-oak habitats above 1,800 m in the Chihuahuan 16 Desert, Texas and northern Mexico. 5. Minimum viable population guidelines applied to Guadalupe fescue. 21 6. Means and dispersion of 30 climate models for Brewster County Climate 28 Projections: 2050 to 2074 compared to 1950 to 2005, RCP 4.5 scenario. 7. Summary of current status of Guadalupe fescue populations. 37 8. Summary of requirements, factors affecting survival, and current conditions of 39 Guadalupe fescue individuals and populations, and the species’ viability (representation, redundancy, and resilience). 9. Future viability under a range of scenarios. 45

List of Figures.

1. Species Status Assessment Framework. 1 2. Guadalupe fescue at Big Bend National Park: Population size and annual 6 precipitation. 3. Images of Guadalupe fescue. 9 4. Global distribution of Guadalupe fescue. 13 5. Guadalupe fescue habitat factors. 17 6. Potential Guadalupe fescue habitat in Texas and Mexico. 18 7. Projected changes in August maximum temperature (C°), Brewster County, Texas. 30 8. Projected changes in January minimum temperature (C°), Brewster County, Texas. 31 9. Projected changes in annual precipitation (mm/day), Brewster County, Texas. 32 10. Projected changes in annual evaporative deficit (mm/month), Brewster County, 33 Texas.

Guadalupe Fescue Species Status Assessment – Aug 2016

I. Introduction.

Guadalupe fescue (Festuca ligulata) is a grass species found only in a few high mountains of the Chihuahuan desert, in Trans-Pecos Texas and Coahuila, Mexico. The Trans-Pecos is the region of Texas west of the Pecos River, and Coahuila is a state in northern Mexico (see Figure 4 below).

Guadalupe fescue has been a candidate for listing since 1980 (45 FR 82519). We are now reviewing the species for listing under the Endangered Species Act (Act). This Species Status Assessment (SSA) is a comprehensive status review of Guadalupe fescue to inform the listing decision and to guide future conservation efforts. This SSA will also provide the background information to guide future actions and documents, which may include listing rules, recovery plans, 5-year reviews, and section 7 consultations. We will update this SSA as new information becomes available.

The SSA framework (Figure 1; Service 2015, entire) summarizes the information assembled and reviewed by the Service, incorporating the best available scientific and commercial data, to conduct an in-depth review of a species’ biology and threats, evaluate its biological status, and assess its resources and conditions needed to maintain long-term viability. For the purpose of this assessment, we define the viability of Guadalupe fescue as its ability to sustain populations in the wild beyond the end of a specified time period. Using the SSA framework, we consider what the species needs to maintain viability through an assessment of its resilience, redundancy, and representation.

• Resilience refers to the population size necessary to endure stochastic environmental Figure 1. Species Status Assessment variation (Shaffer and Stein 2000, pp. 308– Framework. 310). Resilient populations are better able to recover from losses caused by random variation, such as fluctuations in recruitment (demographic stochasticity), variations in rainfall (environmental stochasticity), or changes in the frequency of wildfires.

• Redundancy refers to the number and geographic distribution of populations or sites necessary to endure catastrophic events (Shaffer and Stein 2000, pp. 308–310). As defined here, catastrophic events are rare occurrences, usually of finite duration, that cause severe impacts to one or more populations. Examples of catastrophic events include tropical storms, floods, prolonged drought, and unusually intense wildfire. Measured by the number of populations,

Guadalupe Fescue Species Status Assessment – Aug 2016

their resiliency, and their distribution (and connectivity), redundancy gauges the probability that the species has a sufficient margin of safety to withstand or recover from catastrophic events.

• Representation refers to the genetic diversity, both within and among populations, necessary to conserve long-term adaptive capability (Shaffer and Stein 2000, pp. 307–308). Representation can be measured by the breadth of genetic or environmental diversity within and among populations and gauges the probability that a species is able to adapt to environmental changes (natural or human caused) and to colonize new sites.

In summary, this SSA is a scientific review of the available information related to the biology and conservation status of Guadalupe fescue. Importantly, it does not provide or pre-determine the Service’s decision that Guadalupe fescue does, or does not, warrant protection under the Act. The Service will make that decision after reviewing this document, along with the supporting analyses, other relevant scientific information, and all applicable laws, regulations, and policies, and the results of the decision will be announced in the Federal Register.

Throughout this document, the first uses of scientific and technical terms are underscored with dashed lines; these terms are defined in the glossary in Appendix A. Appendix B describes our analyses of the 22-year monitoring data from the Chisos Mountains population. Appendix C is a table of cause and effects pathways.

Guadalupe Fescue Species Status Assessment – Aug 2016

II. Species Information.

This chapter compiles the background information about Guadalupe fescue upon which our viability assessment is based. The only Guadalupe fescue population that has been quantitatively monitored is the Chisos Mountains population at Big Bend National Park in Texas. Researchers from Texas Parks and Wildlife Department and Big Bend National Park monitored permanent plots in this population 15 times over a 22-year period. This research and analysis provides the best scientific and commercial data available, and we refer to it throughout the SSA Report. Appendix B provides a description and analyses of this project.

II.1. Morphology.

Guadalupe fescue (Festuca ligulata; Figure 3.1) is a perennial, rhizomatous bunchgrass. The stems range from 40 to 100 centimeters (cm) (16 to 40 inches (in.)) tall, and the leaf blades are less than 3 millimeters (mm) (0.12 in.) broad (Poole et al. 2007, p. 228). The panicles (flower stalks; Figure 3.3) have up to three branches bearing a few awnless spikelets (flower clusters that lack hair-like prolongations of the mid or lateral nerves of glumes, lemmas, or paleas (Shaw 2012)) (Figure 1.3a), each with a cluster of 2 to 3 florets (Figure 3.3a).

II.2 Taxonomy.

Guadalupe fescue, a member of the Poaceae (grass family), was described from specimens collected in 1931 in the Guadalupe Mountains, Culberson County, Texas (Swallen 1932, p. 436). The best available scientific and commercial data recognizes that Festuca ligulata is a valid, distinct species classified in the subgenus Leucopoa (Swallen 1932, p. 436; Aiken and Consaul 1995, pp. 1290, 1292, 1296; Aiken et al. 1996, p. 1; Tropicos 2011, p. 1; Integrated Taxonomic Information System 2011, p. 1; Natural Resources Conservation Service 2011, p. 1). Festuca ligulata is distinguished from most other Festuca species by its longer ligule (3 to 5 mm (0.12 to 0.20 in.); the ligule is a membranous or hairy appendage at the junction of the sheath and blade (Figure 3.2) (Gould 1975, pp. 100–101, 633). Festuca ligulata is distinguished from F. thurberi, its closest relative, by the presence of rhizomes, leaf blades with fewer veins, and shorter lemmas and anthers; F. thurberi is a large bunchgrass of high elevations found from southern Wyoming to central New Mexico (Darbyshire and Pavlick 2015, pp. 4, 16). Throughout the rest of this document, we refer to Festuca ligulata as Guadalupe fescue.

II.3. Genetic Diversity.

Two chloroplast gene sequences have been published for Guadalupe fescue (National Center for Biotechnology Information 2015). However, no other genetic analyses have been conducted, so essentially nothing is known about the species’ genetic viability, within-population and within- species genetic differentiation, chromosome number (Darbyshire and Pavlick 2015, p. 16), or breeding system.

The historical populations of Guadalupe fescue occurred in numerous montane “sky island” habitats separated from each other by expanses of low elevation desert where the species has never been observed. Since the species is wind-pollinated and its seeds do not disperse far, it is

Guadalupe Fescue Species Status Assessment – Aug 2016

unlikely that gene flow between populations would have occurred between neighboring mountain ranges since the end of the Pleistocene geological era. Any extant populations would almost certainly have undergone genetic drift, allowing them to adapt to the specific soils, climates, and biota of isolated mountain ranges.

II.4. Phenology (life-cycle events that are influenced by climate and seasonal change) and Reproduction.

Grasses are predominantly wind-pollinated (fertilization requires the transfer of pollen between compatible flowers by air currents). We believe that Guadalupe fescue is wind-pollinated. Among grass species that have been tested, pollen remains viable for a relatively short time, and viability is further reduced under direct sunlight. Wang et al. (2004, pp. 523–530) found that the viability of tall fescue (Festuca arundinacea) pollen, under sunny conditions, declined to less than 5 percent 30 minutes after release from anthers and was completely non-viable by 90 minutes. Under cloudy conditions, viability fell to 5 percent after 150 minutes and was completely sterile after 240 minutes. In a similar study of switchgrass (Panicum virgatum), the pollen viability half-life in sunlight was 4.9 minutes and pollen was complete sterile after 20 minutes, although the viability half-life was 5 times longer under cloudy conditions (Ge et al. 2011, pp. 1–8). Hence, many wind pollinated plant species are characterized by large, dense populations (Friedman and Barrett 2009, p. 1518) to ensure effective fertilization between individuals that are not too closely related (out-crossing).

The breeding system of Guadalupe fescue is currently unknown. The closely related Festuca thurberi might serve as a surrogate, but the breeding system has not been determined for that species either (Meyer 2009, p.5). However, many perennial grasses, including some Festuca species, are known to be obligate out-crossers (Fryxell 1957 p. 180; Pedersen and Sleper 1993, p. 187; Meyer 2009, p. 5). The flowers of an obligate out-crossing plant can only be fertilized by pollen from another individual to which it is not too closely related. When relatively few individuals of wind-pollinated plants are widely dispersed, there is a lower probability that a viable pollen grain carried by wind can reach the receptive stigma of a different, compatible plant. Widely dispersed grasses may still produce abundant seeds if they are self-fertilized. In contrast, pollen transfer and seed production between isolated plants of meadow fescue (Festuca pratensis), which has a high degree of self-incompatibility, decreases rapidly at distances up to 75 m (246 ft) (Rognli et al. 2000, pp. 550–560).

If Guadalupe fescue is an obligate out-crosser, small, scattered populations would produce few or no seeds. Consequently, recruitment would be lower than mortality, and the populations would be at great risk of extirpation. Conversely, if the species is at least capable of self- fertilization, small populations are very likely to be inbred, and could also be at risk of decline through inbreeding depression. Fowler (2015e), considering the small population size and low population density of Guadalupe fescue in the Chisos Mountains population, speculated that this population would already have died out if the species is an obligate out-crosser. Consequently, Fowler concluded that this population is very likely to be highly inbred as a result of extensive self-fertilization. In either case, the physical clustering of numerous genetically diverse plants in close proximity is probably necessary for effective fertilization, out-crossing, and seed

Guadalupe Fescue Species Status Assessment – Aug 2016

production. The issue of critical density (the spatial distribution of individuals) would be exacerbated if Guadalupe fescue is an obligate out-crossing species.

The total annual production of florets at the Chisos Mountains monitoring plots ranged from 0 to over 19,000 and averaged over 4,000 per year (see Appendix B). However, we have not determined the typical range of seed set (the proportion of florets that produce a viable seed) in Guadalupe fescue. Soil samples collected in 2003 contained several hundred seeds of other plant species, but none of Guadalupe fescue (Big Bend National Park and Service 2008, pp. 3–4). This suggests that the seeds do not persist long in the soil seed bank. Among grasses in general, the proportion of florets that produce viable seeds can be very low, particularly among species that are obligate out-crossers occurring at low population densities (Reilley 2016, all). Therefore, the determination of the breeding system, investigation of the genetic variability, and amount of seed set are high priority research needs for the conservation and management of Guadalupe fescue.

Most Guadalupe fescue seeds probably disperse short distances through the actions of wind, water, and gravity. However, viable Festuca seeds may be dispersed longer distances following consumption and subsequent defecation by large herbivores (Janzen 1984, pp. 338–339). Therefore, we believe that Carmen white-tailed deer (Odocoileus virginianus carminis) are able to disperse viable seeds of Guadalupe fescue to potential habitats throughout their home ranges (see further discussion in Section II.8).

II.5. Survival Rates, Lifespan, and Demographic Trends.

The Chisos Mountains data (Appendix B) indicate that Guadalupe fescue is a short-lived perennial with relatively low fecundity. Estimated annual survival rates (proportion of a population that survives to the following year) of Guadalupe fescue ranged from 0.62 to 0.75, and average lifespans ranged from 3.1 to 3.9 years. About 41 percent of individuals die before they are able to reproduce, and 55 percent reproduce from 1 to 3 times. Plants that reproduced in a given year had a lower survival rate (0.61) than plants that did not (0.71). This may be due to the plant consuming large amounts of its stored nutrients and water to produce reproductive structures and mature, viable seeds. Survival rates and reproductive output (estimated through panicle production) are positively, significantly correlated with precipitation, and survival rates have increased significantly from 1993 to 2010; conversely, the population size has decreased significantly over time (Appendix B). The inverse trends in increasing survival rates and decreasing population size over time may be explained by a recruitment rate that is too low to sustain the population. Low recruitment could be due to low fertilization and seed set and/or low establishment rates; the latter may result from poor germination and/or high seedling mortality. These in turn could be caused by lack of genetic fitness, habitat factors, such as the amount of light reaching the soil surface and depth of detritus layer, changes in temperature and rainfall, pathogens, or parasites, among other possible explanations.

Guadalupe Fescue Species Status Assessment – Aug 2016

Table 1. Guadalupe fescue population data at Big Bend National Park. Shading indicates years when surveys were conducted.

- - - - Y e a r - - - - 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Individuals 125 127 112 117 103 78 60 69 Panicles 60 150 8 352 227 2 39 23 Spikelets 843 2058 146 6567 3147 54 648 498 Total Precipitation* (in) 23.06 15.05 15.6 7.97 8.3 14.45 17.32 13.09 9.55 13.75 7.63 13.83 Total Precipitation (cm) 58.6 38.2 39.6 20.2 21.1 36.7 44.0 33.2 24.3 34.9 19.4 35.1

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Individuals 77 96 97 111 115 47 47 Panicles 147 172 0 18 162 18 84 Spikelets 498 2712 0 322 2471 275 1281 Total Precipitation (in) 15.18 22.01 12.85 10.56 13.8 13.08 10.65 12.11 2.53 11.86 13.54 13.57 Total Precipitation (cm) 38.6 55.9 32.6 26.8 35.1 33.2 27.1 30.8 6.4 30.1 34.4 34.5 * Measured at Panther Junction.

Figure 2. Guadalupe fescue at Big Bend National Park: Population size and annual precipitation. 70.0 140

60.0 120 50.0 100 (cm)*

40.0 80 30.0 60 20.0 40

Precipitation 10.0 20 Number of Individuals 0.0 0

Year Total Precipitation (cm) Individuals

Guadalupe Fescue Species Status Assessment – Aug 2016

II.6. Habitat.

Guadalupe fescue is endemic to a few high mountains of the Chihuahuan Desert of trans-Pecos Texas and Coahuila, Mexico. Annual precipitation averages 44.2 cm (17.4 in.) at the Chisos Mountains site, with the majority falling during the monsoon months of June to October. The average daily high temperature is 30.3 degrees Celsius (°C) (86.6 degrees Fahrenheit (°F)) in June and 14.7°C (58.5°F) in January (Big Bend National Park and Service 2008, p. 4).

The known habitats are rocky or talus soils of partially shaded sites in the understory of conifer- oak woodlands (Figures 3.4 and 3.5) above about 1,800 m (5,905 ft) elevation (Poole 1989, p. 8). Documented populations have occurred in soils of igneous as well as calcareous origins. The associated vegetation of the extant populations in the Chisos Mountains (Texas) and the Coahuilan sites is similar (Table 2). At all known sites, Guadalupe fescue occurs in scattered patches in the understory (Poole 1989, p. 10; Big Bend National Park and Service 2008, p. 4).

We know relatively little about the habitat of specimens from the Fraile area collected in 1941 by Stanford et al. (Tropicos 2011), and botanists have not re-visited the site since then. The herbarium label states that the site was on the south slope of a mountain, 24 km northwest of Fraile, Coahuila, on burned-off land with Juniper and Arbutus (madrone) present.

We do not know if the habitats of these known extant populations represent optimal conditions for Guadalupe fescue, or if these are relict populations that are merely persisting in sub-optimal habitats. In the latter case, we would expect these populations to decline without appropriate habitat management. For example, if the historical frequency of wildfire is necessary to sustain Guadalupe fescue populations, we would recommend prescribed burning of habitats (see Section II.9).

Guadalupe Fescue Species Status Assessment – Aug 2016

Table 2. Plant species associated with Guadalupe fescue. Species (Common Name) Chisos Sierra la Madera, Maderas del Mountains, Coahuila, Carmen, Coahuila, Texas1 Mexico1 Mexico2 Quercus gravesii (Chisos red oak) x x x Q. grisea (gray oak) x Pinus cembroides (Mexican pinyon) x x Juniperus deppeana (alligator x x juniper) J. flaccida (drooping juniper) x x Acer grandidentatum (bigtooth x x maple) Q. laceyi (Lacey oak) x Cupressus arizonica (Arizona x cypress) P. arizonica (Arizona pine) x P. strobiformis (southwestern white x pine) Q. hypoleucoides (silverleaf oak) x (replaces gray oak as a dominant species) 1 Poole et al. 2007, p. 228 2 Big Bend National Park and Service 2008, p. 4

Guadalupe Fescue Species Status Assessment – Aug 2016

II.7. Geographic Range.

II.7.a. Historical Range:

Guadalupe fescue has been documented in only six sites, in west Texas and Coahuila, Mexico (see list and Figure 4, below). The best available information indicates that the historical range extended from the Guadalupe Mountains of Texas, in the north, and at least as far south as Ejido El Fraile, southern Coahuila. However, the species’ historical range may have extended further south in Mexico, where there is extensive, potentially suitable habitat (see discussion in section II.8 and Figures 5 and 6, below).

Texas, U.S.A.

1. J.A. Moore and J.A. Steyermark collected Guadalupe fescue in 1931 in McKittrick Canyon in the Guadalupe Mountains, Culberson County, Texas, near the New Mexico state line (Swallen 1932, p. 436). This site is now within Guadalupe Mountains National Park, which was established in 1972. 2. C.H. Muller also collected the species in 1931 in the Chisos Mountains, Brewster County, Texas (Specimen C.H. Muller #7815) (Texas Natural Diversity Database 2007, pp. 3071– 3072); this area became part of Big Bend National Park in 1944. 3. D.J. Appelhons’ report (1973, p. 40) of Guadalupe fescue from the Franklin Mountains in extreme west Texas is probably a misidentification (Worthington 1991, p. 1; Poole et al. 2007, p. 229).

Coahuila, Mexico.

1. In 1941, L. R. Stanford, K. L. Retherford, and R. D. Northcroft collected Guadalupe fescue 24 kilometers (km) (15 miles (mi)) northwest of Ejido El Fraile, in southern Coahuila (specimen numbers ARIZ 15004, MoBot L. R. Stanford 405) (Tropicos 2011, p. 1). (Ejidos is an area of communal land used for agriculture, on which community members individually farm designated parcels and collectively maintain communal holdings in Mexico). 2. M.C. Johnston collected specimens in the Sierra El Jardín in 1973 (Poole 1989, pp. 8, 12). 3. T. Wendt and J. Valdés-Reyna collected the species in 1977 in the Sierra la Madera (specimen ARIZ 237645) (Poole 1989, pp. 8, 12). 4. Three colonies of several hundred Guadalupe fescue plants were documented in 2003 at the north end of the Maderas del Carmen range at an elevation of 1,981 m (6,500 ft) (Big Bend National Park and Service 2008, p. 3). These colonies apparently are close enough to be considered parts of a single population.

II.7.b. Current Range:

Texas.

The Chisos Mountains population in Big Bend National Park is the only known population remaining in the United States. Botanists have extensively surveyed the limited amount of potential habitat at Big Bend National Park where the elevation exceeds 1,829 m (6,000 ft), as well as most of the potential habitat in the Davis Mountains, Jeff Davis County, Texas (see

Guadalupe Fescue Species Status Assessment – Aug 2016

Figure 4), but have not found additional populations (Big Bend National Park and Service 2008, p. 3). Botanists periodically continue surveys at McKittrick Canyon in Guadalupe Mountains National Park, but Guadalupe fescue was last observed there in 1952 (Texas Natural Diversity Database 2007, pp. 3073–3074) and is presumed extirpated. While undiscovered populations might exist in the New Mexico portion of Guadalupe Mountains National Park where the habitat appears suitable, the best available data does not indicate any other extant populations.

Mexico.

In 2003, Joe Sirotnak (National Park Service) observed three colonies of Guadalupe fescue at the Sierra Maderas del Carmen that each had several hundred plants (Big Bend National Park and Service 2008, p. 3). Valdés-Reyna (2009, pp. 2, 3, 13, 15) collected Guadalupe fescue at this Sierra Maderas del Carmen site in 2007 and September 1–3, 2009, but did not determine the population size. He also searched at the approximate geographic coordinates of Johnston’s 1973 collection site in the Sierra El Jardín, but did not find the species there; however, we do not know if he searched exactly where Johnston found the species, or if this population is extant. Both of these mountain ranges are within the Area de Protección de Flora y Fauna Maderas del Carmen (Maderas del Carmen Protected Area for Plants and Animals), a federally designated natural area owned by CEMEX, a cement manufacturer. In 2008 and 2009, Valdés-Reyna was unable to obtain access to the Sierra la Madera, now a federally designated Area de Protección de los Recursos Naturales (APRN), Zona Protectora Forestal (Protected Natural Resource Area, Protected Forest Zone), where he and Wendt collected the species in 1977 (Valdés-Reyna 2009, p. 13). We do not know if the population northwest of Fraile has been observed since 1941; Valdés-Reyna (2010, p. 1) did not have information on that collection, and no other information is available. Many potentially suitable sites in the mountains of Coahuila and adjacent areas in the Mexican states of Nuevo León, San Luís Potosí, Zacatecas, Durango, and Chihuahua, where other populations may exist, have not been surveyed.

Guadalupe Fescue Species Status Assessment – Aug 2016

Table 3. Historical and current populations of Guadalupe fescue.

Site Name Ownership Location County, State, First Last Observed Current Status Country Record Guadalupe National Park McKittrick Canyon Culberson County, 1931 1952 Extirpated Mountains Service Texas, U.S.A. National Park Big Bend National Park Chisos Mountains Brewster County, 1931 2014 Extant National Park Service Texas, U.S.A. Franklin 31°57’8” to El Paso County, 1973 Misidentified Mountains 31°57’27” N; Texas, U.S.A. 106°29’30” to 106°30’0” W APFF Maderas CEMEX (Private) Maderas del Coahuila, Mexico 2003 2009 Extant del Carmen Carmen APFF Maderas CEMEX (Private) Sierra El Jardín Coahuila, Mexico 1973 1973 Unknown del Carmen Ej. El Fraile Ejido El Fraile 24 km (15 mi) Coahuila, Mexico 1941 1941 Unknown (Communal) northwest of Fraile APRN Sierra Private Sierra la Madera Coahuila, Mexico 1977 unable to Unknown la Madera access in 2009

Guadalupe Fescue Species Status Assessment – Aug 2016

II.8. Range of Estimated Potential Habitat.

Zimmerman and Moir (1998, p. 14), considering the populations reported in the Sierra la Madera and Maderas del Carmen, Coahuila, speculate that the Big Bend National Park and Guadalupe Mountains National Park populations may represent the northern extent of the range of Guadalupe fescue and that the range may be centered further south in Coahuila. In addition, during our Candidate Assessment in 2009, we found herbarium records for the 1941 collection by Stanford et al. 24 km (15 mi) northwest of Fraile, an ejido in southern Coahuila; this collection had been overlooked in previous Candidate Assessments and other information published about the species. Thus, the range and amount of habitat that could potentially support Guadalupe fescue populations may be much greater than previously acknowledged.

What is known about Guadalupe fescue suggests a simple model for its geographic distribution: all documented occurrences are in the Chihuahuan Desert ecological region at elevations above 1,800 m (5,905 ft) in conifer-oak woodlands.

Considering that only 6 populations have been found, we probably do not know the actual elevation tolerance of this species. Many plant species occur at relatively lower elevations in mountains where habitats are relatively cool and moist, such as in narrow ravines, north-facing slopes (in the northern hemisphere), or windward slopes where there is a pronounced rain shadow (higher rainfall on prevailing windward slopes). Optimal elevation ranges for species are usually higher at lower latitudes and vice-versa. Nevertheless, the 1,800 m elevation contour represents the best available information regarding the lower elevation tolerance of this species. We do not know what the upper elevation limit is, and it is possible that there are no mountains in the Chihuahuan Desert region that are too high for Guadalupe fescue.

Conifer-oak woodlands may occur in areas classified as pine, conifer, pine-oak, or conifer-oak, and as forest or woodland, on available vegetation classification maps. The conifer species typically include one or more of the following: Mexican pinyon (Pinus cembroides), Arizona pine (P. arizonica), southwestern white pine (P. strobiformis), alligator juniper (Juniperus deppeana), drooping juniper (J. flaccida), and Arizona cypress (Cupressus arizonica). Characteristic oaks include one or more of the following: Chisos red oak (Quercus gravesii), gray oak (Q. grisea), Lacey oak (Q. laceyi), and silverleaf oak (Q. hypoleucoides). Other broadleaf trees, such as bigtooth maple (Acer grandidentatum), may also occur in this element.

Habitat areas do not need to be contiguous to be considered occupied, provided that they are not separated by low-elevation (<1,000 m elevation) gaps. This rational is based on expected long- distance dispersal of viable seeds of Guadalupe fescue by large mammals (Janzen 1984, pp. 338– 339), such as elk, whitetail deer, mule deer, or pronghorn antelope. The Carmen white-tailed deer is the most common ungulate in the Chisos Mountains (Sirotnak 2016). Krausman and Ables (1981) investigated the ecology of Carmen white-tailed deer in the Chisos Mountains. They report that grass species constitute 3.5 percent of the deer’s overall diet but this amount increases to 12 percent during some seasons (pp. 22–24). Carmen white-tailed deer are sedentary and have similar home range sizes to other white-tailed deer subspecies in Texas (radii of 1.1 to 2.4 km) (p. 25). Their habitat use is closely associated with dense stands of oak and the presence of free-standing water (pp. 4, 17), and their range is restricted to elevations above 906 to 1,220 m (2,970 to 4,000 ft) (p. 4). Hence, we expect that Carmen white-tailed deer are able to

Guadalupe Fescue Species Status Assessment – Aug 2016

disperse viable seeds of Guadalupe fescue to potential habitats that are not separated by gaps that are below about 1,000 m (3,208 ft) and more than 2.4 km (1.5 mi) wide.

We used topographic and vegetation cover shapefiles from INEGI (Instituto Nacional de Estadística y Geografía, México), Digital Elevation Models from U.S. Geological Survey, and vegetation cover shapefiles from Texas Parks and Wildlife Department, to map where these geographic features occur (Figure 5). We created new shapefiles from the intersection of areas within the Chihuahuan Desert that are above 1,800 m (5,905 ft) and have conifer, oak, or conifer- oak forest and woodland. This method is based on fairly broad assumptions but provides at least an estimate of the amount and distribution of potential habitat for Guadalupe fescue (red-tinted polygons in Figure 6, summarized in Table 4). The geographic model reveals large numbers of small areas – individual mountain peaks – that are probably too small and isolated to support viable populations of Guadalupe fescue. We do not know how large an area must be to support a viable population. However, we do know that many plant species in the Chihuahuan Desert have migrated to different elevations and latitudes, or were extirpated, since the end of the late Wisconsinan glaciation (about 11,000 years ago) (Van Devender and Spaulding 1979, pp. 701– 710). Larger habitat areas provide more opportunities for populations to migrate, as plant communities and weather patterns change, and therefore may be more suitable. For this reason, we restricted the model to a provisional minimum area of 200 ha (494 ac) to screen out small areas. This geographic model reveals that northern Mexico has 283 areas of potential habitat totaling 537,998 ha (over 1.3 million ac), compared to 20 such areas totaling 27,881 ha (68,894 ac) in Texas. Thus, about 95 percent of the potential habitat identified by this method is in Mexico.

This estimate suggests several important conclusions: 1) There may be much more potential habitat in northern Mexico than in west Texas; 2) some areas of potential habitat in northern Mexico occur in federally designated Protected Natural Areas (Areas Naturales Protegidas) that are at least nominally managed for conservation, and where it may be easier to obtain permission to access land, conduct surveys, and promote appropriate management practices; 3) many areas of potential habitat in northern Mexico are relatively large, and could support larger, genetically more robust populations of Guadalupe fescue; and 4) wildfires probably occur with greater frequency in many Mexican habitats, which could be beneficial for long-term sustainability of Guadalupe fescue populations.

Guadalupe fescue is a relatively small grass that superficially resembles other Festuca species. Even the few botanists who are familiar with the species have to look carefully to identify it. Potential habitats are remote, rugged, and difficult to access. The proportion of these potential habitats in Mexico that have been thoroughly surveyed by qualified botanists is probably very low. Finally, security concerns in northern Mexico during the last ten years have discouraged botanical exploration of remote areas. Therefore, we conclude that undiscovered viable populations of Guadalupe fescue may exist in northern Mexico. However, the best available data does not indicate any additional populations.

Guadalupe Fescue Species Status Assessment – Aug 2016

Table 4. Estimates of pine, oak, and pine-oak habitats above 1,800 m in the Chihuahuan Desert, Texas and northern Mexico.

Nation Habitat No. Median Average Total No. Total, National Type of Area - Area - Ha Areas Areas Total, All Areas Ha Ha > 200 > 200 Ha Areas Ha > 200 Ha Mexico Pine-Oak 310 215.9 983.2 304,788 162 292,294 537,998 Mexico Oak 242 150.0 962.8 232,986 103 223,308 Mexico Conifer 33 238.6 715.8 23,621 18 22,396 Texas Pine-Oak 135 3.8 193.7 26,152 19 24,048 27,881 Texas Conifer 6 0.3 642.8 3,857 1 3,833 TOTALS 726 591,404 303 565,879

Guadalupe Fescue Species Status Assessment – Aug 2016

II.9. Ecology and the Wildfire Cycle.

The conifer-oak woodlands of the Chisos Mountains experienced relatively frequent, low- intensity wildfires for centuries (Moir and Meents 1981, p. 7; Moir 1982, pp. 90–98; Poole 1989 p. 8; Camp et al. 2006, pp. 3–6, 14–23, 59–61). Fire scars on tree stumps at and near the Chisos site indicate that as many as 10 wildfires have occurred between 1770 and 1940; however, the last wildfire was in 1944 (Zimmerman and Moir 1998, p. 12), more than 70 years ago. Wildfire has been suppressed at Big Bend National Park since the park’s establishment in 1944; no wildfires have occurred at the Guadalupe fescue site since that time (Camp et al. 2006, p. 4). Periodic wildfire and leaf litter reduction may be necessary for long-term survival of Guadalupe fescue populations, although this has not been investigated (Big Bend National Park and Service 2008, pp. 4–5). The absence of wildfire in the Chisos range has led to an increased density of small-diameter trees and a deep accumulation of leaf litter. This high fuel load increases the risk of a much more intense wildfire that could kill all or most of the vegetation and sterilize the soil. The impact of an intense wildfire would potentially be catastrophic to the remaining Guadalupe fescue population. Therefore, we need to learn more about the fire ecology of Guadalupe fescue and the frequency and history of wildfires in the known U.S. and Mexican populations.

Seeds of many fire-dependent plant species are able to remain dormant in the soil seed bank for many years, and are stimulated to germinate after fire has burned surface vegetation (Landis 2000, pp. 25–29; Jefferson et al. 2008, 251–259). However, soil samples collected in 2003 contained several hundred seeds of other plant species, but none of Guadalupe fescue (Big Bend National Park and Service 2008, pp. 3–4). This suggests that the seeds do not persist long in the soil seed bank.

Guadalupe Fescue Species Status Assessment – Aug 2016

III. Summary of Individual, Population, and Species Requirements.

This section summarizes the requirements of individuals, populations, and the species, based on the information presented in Section II and in Appendix B.

III.1. Requirements of Individuals.

Habitats.

Guadalupe fescue grows in rocky or talus soils of partially shaded sites in the understory of conifer-oak forest and woodlands. We do not know if it specifically requires light gaps in forests, but it is likely that the amount of light is important. Suitable habitats occur above about 1,800 m (5,905 ft) in the Chihuahuan Desert of northern Mexico and Texas. Soils may be of either igneous or calcareous origin.

Reproduction.

Guadalupe fescue is wind-pollinated, and therefore, out-crossing is unlikely to occur unless numerous unrelated plants are physically close together. The breeding system has not been determined. If the species is an obligate out-crosser, effective fertilization requires relatively large, dense populations. Conversely, if the species is at least capable of self-fertilization, populations must also be relatively large to conserve genetic diversity and minimize inbreeding. In either case, the physical clustering of numerous plants in close proximity is probably necessary for effective fertilization, out-crossing, and seed production. Colonization of new sites and gene flow between more distant sites may also be possible through seed dispersal in the dung of Carmen white-tailed deer (see Section II.8).

Precipitation.

Panicle production and survival are positively correlated with rainfall. The amount of rainfall over longer periods, such as the previous 21 months, appears to have more influence on panicle production than rainfall during the previous 9 months or the previous February through May.

Lifespan and survival rates.

Guadalupe fescue is a short-lived perennial; our estimates of average lifespan range from 3.2 to 3.9 years. About 41 percent of individuals die before they are able to reproduce, and 55 percent reproduce from 1 to 3 times. Individuals have lower survival rates (61 percent) in years that they reproduce than non-reproductive plants (71 percent).

Guadalupe Fescue Species Status Assessment – Aug 2016

III.2. Requirements of Populations.

Minimum viable population size and density.

Estimate of minimum viable population.

Minimum viable population (MVP) refers to the smallest population size that has a high probability of surviving a prescribed period of time. For example, Mace and Lande (1991, p. 151) propose that species or populations be classified as vulnerable when the probability of persisting 100 years is less than 90 percent. Determinations of MVP usually take into account the effective population size, rather than total number of individuals; 10 genetically identical individuals (for example, clones) would have an effective population size of 1. Population Viability Analysis (PVA) is based on essentially the same type of baseline data and calculation, but addresses the survival probability of a specified population.

Although the MVP has not yet been calculated for Guadalupe fescue, we can estimate its MVP by comparison to species with similar life histories (i.e., surrogates) for which MVPs have been calculated, using the following guideline adapted from Pavlik (1996, p. 137). Species with traits that all fall under column A (Table 5) would have MVPs around 50 individuals. Those with traits that all ascribe to column C would have MVPs around 2,500 individuals. We added an intermediate column (B) to Pavlik’s table to account for species with intermediate or unknown traits. The bold letters in the table indicate values, if known, for Guadalupe fescue. Three factors require more individuals (growth form, fecundity, and survivorship), three are intermediate or unknown (longevity, breeding system, and seed duration), and three require fewer individuals (ramet production, environmental variation, and successional status), suggesting an estimated MVP in the intermediate range (500 to 1,000 individuals). This provisional MVP range may be revised in the future if accumulated data permits a more precise calculation.

Table 5. Minimum viable population guidelines applied to Guadalupe fescue (adapted from Pavlik 1996, p. 137).

Factor A. MVP of 50 B. Intermediate MVP C. MVP of 2,500 individuals for Range for species with individuals for species with intermediate or unknown species with these these traits. traits. traits. Longevity Perennial Short-lived perennial Annual Breeding System Selfing Unknown Outcrossing Growth Form Woody Herbaceous Fecundity High Probably Low Ramet Production Common Rare or None Survivorship High Low Seed Duration Long Unknown Short Environmental Low High Variation Successional Status Climax Seral or Ruderal

Guadalupe Fescue Species Status Assessment – Aug 2016

Due to wind-pollination and the possible effects of inbreeding in small populations, it is likely that populations will decline if the total number of individuals is too low, and if individuals are scattered too widely in habitats. The issue of critical density would be exacerbated if Guadalupe fescue is an obligate out-crossing species. The population at Big Bend National Park has declined over the last 22 years, despite an increasing survival rate over the same time frame. This suggests that recruitment has been insufficient for long-term viability. We do not know why recruitment is low, but one possible cause is low fertilization rates due to low numbers of widely scattered plants. By comparison with other plants that have similar life histories, we provisionally estimate that the minimum viable population size is between 500 and 1,000 individuals.

Population size may be positively correlated with rainfall over 33-month periods. Rainfall (or drought) over shorter time frames appears to have less effect on population size.

Size of habitat patches.

The amount of contiguous area in habitat patches is likely to be important, although we do not know if there is a critical lower limit or what that limit would be. Larger habitats are probably better, since they could include larger populations and greater genetic diversity. Furthermore, since habitat characteristics change over time, due to ecological succession, climate changes, or other causes, larger habitats should have more opportunities for the migration of colonies over time. For example, the optimal elevation zone could rise due to increasing temperatures and/or lower rainfall. Large patches are more likely to have a greater range of elevations, as well as other habitat characteristics that the species requires. To gain an understanding of the geographic distribution of potential habitats, in the absence of data that reveals the minimum or optimum sizes of habitats, we chose a provisional patch size of 200 ha (494 ac), with totals shown in Table 4 (see section II.8).

Fire.

The role of fire on Guadalupe fescue populations is very likely to be important for two reasons. First, many grass and forb understory species are stimulated during the years immediately following wildfire but decline during long periods without fire. Second, the accumulation of fuels and small dead trees can cause very hot, catastrophic fires that kill overstory trees as well as the seeds and beneficial microorganisms in the soil, such as mycorrhizal fungi. Fire is probably inevitable in the pine and pine-oak forests of the Chihuahuan Desert. Frequent, relatively cool fires may be essential for the long-term sustainability of these forested ecosystems; the last fire at the Chisos Mountains population site was more than 70 years ago.

III.3. Species Requirements.

The viability of a species can be assessed in terms of its resilience, redundancy, and representation (Shaffer and Stein 2000, pp. 307–310). Resilience refers to population sizes; larger populations are more likely to endure than small ones. Redundant populations increase the species’ chances of surviving catastrophic events. Representation refers to the breadth of genetic diversity necessary to conserve long-term adaptive capability. With regard to resilience,

Guadalupe Fescue Species Status Assessment – Aug 2016 we estimate, provisionally, that viable populations should have at least 500 to 1,000 individuals. The best available information does not indicate what the minimum viable degree of representation and redundancy should be; it is reasonable to conclude that more is better.

Guadalupe Fescue Species Status Assessment – Aug 2016

IV. Factors Affecting the Survival of Guadalupe Fescue: Threats, Vulnerabilities, and Conservation Challenges.

The following list describes factors that affect the continued survival of Guadalupe fescue. These largely interrelated factors are not listed in order of severity.

Demographic consequences of small population size and density.

Small populations are less able to recover from losses caused by random environmental changes (Shaffer and Stein 2000, pp. 308–310), such as fluctuations in recruitment (demographic stochasticity), variations in rainfall (environmental stochasticity), or changes in the frequency of wildfires. Furthermore, since Guadalupe fescue is wind-pollinated, there is a markedly reduced probability of out-crossing between unrelated individuals of small, scattered populations; if the species requires out-crossing, the remaining plants would produce few viable seeds (see Section II.4), further reducing population recruitment and engendering a downward spiral toward extirpation. The demographic consequences of small population size are compounded by genetic consequences (discussed below), since reduced out-crossing corresponds to increased inbreeding. We have provisionally estimated an MVP of 500 to 1,000 individuals (Section II.10). Population viability must also have an element of minimum viable density to ensure high rates of out-crossing, although we currently have no data to estimate what that minimum density should be.

Genetic consequences of small population sizes (Barrett and Kohn 1991, pp. 3–30).

Small, reproductively isolated populations are susceptible to the loss of genetic diversity, genetic drift, and inbreeding. The loss of genetic diversity may reduce the ability of a species or population to resist pathogens and parasites, to adapt to changing environmental conditions, or to colonize new habitats. Conversely, populations that pass through a “genetic bottleneck” may subsequently benefit through the elimination of harmful alleles. Nevertheless, the net result of loss of the genetic diversity is likely to be a loss of fitness and lower chance of survival of populations and of the species.

Genetic drift is a change in the frequencies of alleles in a population over time. Genetic drift can arise from random differences in founder populations and the random loss of rare alleles in small isolated populations. Genetic drift may have a neutral effect on fitness, but is also a cause of the loss of genetic diversity in small populations. Genetic drift may also result in the adaptation of an isolated population to the climates and soils of specific sites, leading to the development of distinct ecotypes and to speciation.

Inbreeding depression is the loss of fitness among offspring of closely related individuals. While most animal species are susceptible to inbreeding depression, plant species vary greatly in response to inbreeding. It is very likely that the Chisos Mountains population of Guadalupe fescue is highly inbred, and the Maderas del Carmen population, although somewhat larger, may also be inbred. Nevertheless, we do not know to what extent inbreeding has reduced fitness of this population.

Guadalupe Fescue Species Status Assessment – Aug 2016

Changes in wildfire frequency.

Evidence from tree rings indicates that the Chisos Mountains experienced frequent low-intensity wildfire until 1944 (discussed in section II.9). We assume that historically regular wildfire cycles occurred in most if not all montane “sky island” conifer-oak forests of the Chihuahuan Desert. In the U.S., where wildfires have been effectively suppressed for many decades, the reduced frequency of wildfire is likely to be an important factor affecting the survival of Guadalupe fescue for two reasons. First, periodic low-intensity fire may be necessary to restore habitat conditions that are suitable for Guadalupe fescue; this could include reducing competition from woody plants, reducing the depth of the surface litter, and increasing the amount of light reaching the understory of this woodland habitat. Second, since wildfire is probably inevitable in this ecosystem, an eventual fire occurring after a long hiatus will be more severe and may drastically change the woodland community, rendering it unsuitable for Guadalupe fescue; drastic changes could include a crown fire that kills the conifer-oak overstory, sterilizes the soil seed bank, and increases soil erosion.

Wildfires are probably more common in northern Mexico, where prescribed burning has been practiced as a traditional land-management tool since pre-Colombian times. Furthermore, potential habitats in Mexico are remote, rugged, and difficult to access, and federal, state, and local agencies have fewer resources and less training for wildfire suppressions. Consequently, disruption of the natural wildfire cycle is less likely to be a significant factor affecting the survival of Guadalupe fescue in Mexico.

Livestock grazing.

Guadalupe fescue is palatable to livestock and is potentially threatened by grazing animals (Poole 1989, p. 13). Although Big Bend National Park and Maderas del Carmen do not allow livestock grazing, trail crews at Big Bend National Park use horses and mules (Poole 1989, p. 14; Gordon and Poole 2009, p. 2). Introduced animals, including feral burros, horses, hogs, and aoudad (Barbary sheep) have damaged native vegetation and habitats in other areas of Big Bend National Park. Due to the small population sizes at Big Bend National Park, Maderas del Carmen, and perhaps other sites, the loss of even a few individuals could reduce the population size and genetic diversity below the level necessary for long-term survival. Nevertheless, under the terms of a Candidate Conservation Agreement (Big Bend National Park and Service 2008), personnel at Big Bend National Park take precautions to prevent impacts to Guadalupe fescue by pack animals and feral livestock. Therefore, it is unlikely that livestock grazing will affect Guadalupe fescue at Big Bend National Park.

Conversely, many potential habitat areas in Mexico, particularly privately owned and ejido lands, may be subject to severe livestock grazing pressure. Other areas in Mexico may have little or no livestock, due to remote locations, lack of water, and rugged terrain. Thus, livestock grazing may be a more significant concern to the survival of Guadalupe fescue in some potential habitats in Mexico, and the severity of this threat is not likely to diminish in the future.

Guadalupe Fescue Species Status Assessment – Aug 2016

Recreation.

A popular hiking trail from Pinnacles to Boot Spring in Big Bend National Park bisects the Guadalupe fescue population and raises some potential threats. These threats include trampling by hikers straying from the trail, trampling or grazing by pack animals used by Big Bend National Park personnel, and erosion or debris flow caused by trail runoff. In 2005, debris flows below trail switchbacks buried some of the monitored Guadalupe fescue plants up to 20 cm (8 in.) deep; trail runoff may or may not have caused the debris flow (Big Bend National Park and Service 2008, p. 5). However, since the site and its public use are now well-managed at Big Bend National Park (see section V below), it is unlikely that recreation will significantly affect the survival of this population.

We have no information on recreational uses of potential habitats in Mexico. However, considering the remoteness of the potential habitats, low human population density, and limited access to private and ejido land, recreational uses have a low probability of affecting the survival of Guadalupe fescue in Mexico.

Invasive species.

Horehound (Marrubium vulgare), an introduced invasive plant, is present in the Chisos Mountains at Big Bend National Park but has been removed from the vicinity of the Guadalupe fescue site (Big Bend National Park and Service 2008, p. 5). King Ranch bluestem (Bothriochloa ischaemum), an introduced grass of Asian origin, is abundant and very invasive throughout central and west Texas, including at the historic population in Guadalupe Mountains National Park. Horehound, King Ranch bluestem, and other invasive plant species potentially threaten Guadalupe fescue through competition for water, nutrients, and light. The Candidate Conservation Agreement (Big Bend National Park and Service 2008) calls for periodic monitoring of the Guadalupe fescue population and control of invasive species. Big Bend National Park is currently consulting with the Service under section 7 of the Act regarding a proposed programmatic invasive species management plan. Therefore, since invasive species will be carefully monitored and controlled in the Chisos Mountains, the magnitude of this threat is currently low. However, the impacts of invasive species could become more severe in the future as a result of climate changes (discussed below).

It is highly likely that introduced invasive species are widespread throughout the range of potential habitats in Mexico, but we have no information on introduced invasive species in the known Mexican sites or their impacts on Guadalupe fescue.

Climate change.

The Fifth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) (IPCC 2013, p. 23) projects the following changes by the end of the 21st century, relative to the 1986 to 2005 averages: It is virtually certain that most land areas will experience warmer and/or fewer cold days and nights; it is virtually certain that most land areas will experience warmer and/or more frequent hot days and nights; it is very likely that the frequency and/or duration of warm spells and heat waves will increase in most land areas; it is very likely that the frequency, intensity, and/or amount of heavy precipitation will increase in mid-latitude land masses; it is

Guadalupe Fescue Species Status Assessment – Aug 2016

likely that the intensity and/or duration of droughts will increase on a regional to global scale. The magnitude of projected changes varies widely, depending on which scenario of future greenhouse gas emissions is used. These scenarios are called Representative Concentration Pathways (RCPs). Under the best-case scenario of RCP2.6, the combined emissions of carbon dioxide, methane, and nitrous oxide, expressed as the carbon dioxide equivalent, will stabilize at 475 parts per million (ppm) by the year 2100. This figure rises to 630, 800, and 1,313 ppm under the RCP4.5, RCP6.0, and RCP8.5 scenarios, respectively (IPCC 2013, p.22).

To evaluate how the climate of Guadalupe fescue habitats may change, we used the National Climate Change Viewer (U.S. Geological Survey 2015) to compare past and projected future climate conditions for Brewster County, Texas. The baseline for comparison was the observed mean values from 1950 through 2005, and 30 climate models were used to project future conditions for 2050 through 2074. We selected the climate parameters of August maximum temperature, January minimum temperature, annual mean precipitation, and annual mean evaporative deficit, and used both the RCP4.5 and RCP8.5 scenarios to provide a range of projected values. The results are summarized in Table 6 and in Figures 4, 5, 6, and 7. To interpret these results, it is important to consider the means as well as the dispersion of the 30 climate models (Table 6). The historical baseline average annual precipitation is 0.8 mm/day, or 292 mm/year (11.5 in/year). Although the model mean projects no change in rainfall, these models do not simulate well the projected patterns of regional precipitation (IPCC 2013, p. 11); hence, the projection of “no change” reflects a lack of precision, rather than a determination that there will be no change in precipitation. On the other hand, the models do project a greater increase in evaporative deficit due to increasing temperatures. Evaporative deficit, defined as the difference between actual and potential evapotranspiration (U.S. Geological Survey 2014, p. 11), may be a better indicator of plant stress than precipitation alone, since it takes temperature into account. The baseline evaporative deficit for Brewster County is 57.0 mm/month (26.9 in/year). Hence, these models project that plant growth and survival in Brewster County will become more moisture limited, although the degree of change depends on the RCP model. Under the RCP8.5 scenario, the projected changes in temperatures and evaporative deficit are greater, as one would expect. Interestingly, the projected change in annual precipitation under RCP8.5 differs little from the RCP4.5 scenario, due to the lack of model precision in projecting changes in precipitation.

Guadalupe Fescue Species Status Assessment – Aug 2016

Table 6. Means and dispersion of 30 climate models for Brewster County Climate Projections: 2050 to 2074 compared to 1950 to 2005, RCP 4.5 scenario (U.S. Geological Survey 2015).

Climate Parameter Mean of 30 models Range of individual models August maximum Increase of 2.5° C (4.5° F) Increase of 0.5° to 6.4° C (0.9° to 11.5° F) temperature

January minimum Increase of 1.7° C (3.1° F) Increase of 0.3° to 3.0° C (0.5° to 5.4° F) temperature

Average annual No change Decrease of 0.2 mm/day (2.9 in/year) to precipitation increase of 0.2 mm/day (2.9 in/year)

Evaporative Deficit Increase of 13.7 Increase of 2.0 to 25.2 mm/month (0.9 to 11.9 mm/month (6.5 in/year) in/year)

Poulos (2014) and Waring and Schwilk (2014) documented woody plant mortality and dieback in the Chisos Mountains following a prolonged freeze in February 2011 and the exceptional drought and above-average temperatures that prevailed from October 2010 through September 2011. Poulos (2014) reported significant mortality of trees after the drought, with the greatest mortality among trees less than 20 cm (7.9 in) diameter at breast height (p. 5). Piñon pine had higher mortality than alligator juniper or Emory oak (p. 6). Waring and Schwilk (2014) found that woody plant cover was significantly reduced post-drought at elevations ranging from 666 m (2,185 ft) to 1,920 m (6,299 ft) (pp. 5–8), and proportional dieback of succulents and shrub species was significantly lower at higher elevations (p.8). Although these investigators could not determine the extent that freezing and drought contributed to plant mortality, their results confirm that tree canopy cover was significantly reduced and species composition was altered after a single year of extreme weather. If similar extreme weather events occur with increasing frequency in the future, it is likely that the conifer-oak woodland habitat of Guadalupe fescue will continue to change in woody plant cover and composition.

The impacts of ongoing climate changes to plant communities may be judged from vegetation responses to prior climate changes. Van Devender and Spaulding (1979) used data from packrat (Neotoma sp.) middens to determine how vegetation has changed in the Southwestern U.S. during the last 22,000 years. They determined that piñon-juniper woodlands present from 22,000 to 11,000 years before the present (b.p.), in sites ranging from 550 m (1,804 ft) to 1,525 m (5,003 ft) elevation, now support desert scrub (p. 701). In the Guadalupe Mountains, Douglas fir (Pseudotsuga menziesii) and southwestern white pine (Pinus strobiformis) at 2,000 m (6,562 ft) were replaced by juniper grasslands after 11,500 years b.p. (p. 704). Rapid, widespread changes occurred during the middle Holocene, when many woodland plants migrated northward or to higher elevations in mountains and disappeared from lowlands (p. 706), and grasslands present from 1,200 m (3,937 ft) to 2,000 m were replaced by succulent desert scrub (p. 707). However,

Guadalupe Fescue Species Status Assessment – Aug 2016

the ranges of some plants migrated to lower elevations during this period (pp. 707–708); this may be explained by the milder winters during the Late Wisconsinan, caused by the blocking effects of the Cordilleran and Laurentide Ice Sheets (pp. 708–709). After about 8,000 years b.p., precipitation patterns changed: winter precipitation decreased and summer monsoon rainfall increased throughout this region (p. 708). The authors caution that entire plant communities do not migrate synchronously, since individual plant species respond in different ways to changes in temperature and precipitation patterns (p. 709).

Nevertheless, we do not know how Guadalupe fescue responded to prior climate changes, nor can we determine how these projected climate changes, forecast by the range of models and emissions scenarios, will affect the synecology of Guadalupe fescue and its habitat. A warmer climate may adversely affect Guadalupe fescue by raising the species’ optimal elevation range. In this case, the species could persist by migrating upslope, but the potential habitat area diminishes with increasing elevation in mountain ranges; at Big Bend National Park, the Guadalupe fescue population is already near the top of the Chisos range and would not be able to migrate to a higher elevation. Furthermore, it is unlikely that the species could migrate fast enough to match the projected rate of climate change. Climate warming may also improve habitat conditions for competitive invasive plant species, such as Lehmann lovegrass (Eragrostis lehmanniana) and buffelgrass (Pennisetum ciliare) that currently occupy drier, hotter habitat at lower elevations than Guadalupe fescue. Alternatively, increased tree mortality could reduce competition from woody plants. Warmer winters could extend the growing season and improve reproduction and survival of Guadalupe fescue. Thus, although it is likely that the projected climate changes will affect the survival of Guadalupe in infinitely complex ways, we do not currently know what the net result of beneficial and detrimental effects will be.

Guadalupe Fescue Species Status Assessment – Aug 2016

Figure 7. Projected changes in August maximum temperature (C°), Brewster County, Texas. 2050-2074 average compared to 1950-2005 average 10

9 9 8 8 7 7 6 6 5 5 5 4 4 4 3 3 2 2 2 2 1 1 1

Number of Climate = (N Models 30) Climate of Number 1 0

Change in August Maximum Temperature C° RCP4.5 RCP8.5 RCP4.5 Mean: 2.5°; Range 0.5° - 6.4° RCP8.5 Mean: 3.5°; Range 2.0° - 8.3° Data source: U.S. Geological Survey 2015

Guadalupe Fescue Species Status Assessment – Aug 2016

Figure 8. Projected changes in January minimum temperature (C°), Brewster County, Texas. 2050-2074 average compared to 1950-2005 average

14 13

12 10 10 9

8 7

6 5

4 3 3 3 2 2 2 1 1 1 Number of Climate = (N Models 30) Climate of Number

Change in January Minimum Temperature C° RCP4.5 RCP8.5 RCP4.5 Mean: 1.7°; Range: 0.3° - 3.0° RCP8.5 Mean: 2.6°; Range: 0.9° - 3.9° Data source: U.S. Geological Survey 2015

Guadalupe Fescue Species Status Assessment – Aug 2016

Figure 9. Projected changes in annual precipitation (mm/day), Brewster County, Texas. 2050-2074 average compared to 1950-2005 average 30

25 24 20 20

15 10 10 6 5 Number of Climate = (N Models 30) Climate of Number 0 -2.0 to -1.5 -1.5 to -1.0 -1.0 to -0.5 -0.5 to 0 0 to 0.5 0.5 to 1.0 1.0 to 1.5 1.5 to 2.0 Change in Annual Precipitaiton (mm/day) RCP4.5 RCP8.5

RCP4.5 Mean: 0.0 mm/day; Range: -0.2 - +0.2 mm/day RCP8.5 Mean: : 0.0 mm/day; Range: -0.2 - +0.3 mm/day Data source: U.S. Geological Survey 2015

Guadalupe Fescue Species Status Assessment – Aug 2016

Figure 10. Projected changes in annual evaporative deficit (mm/month), Brewster County, Texas. 2050-2074 average compared to 1950-2005 average

25 24

20 19

15 10 10

5 3 2 1 1 Number of Climate = (N Models 30) Climate of Number 0 -12.5 to 0 0 to 12.5 12.5 to 25.0 25.0 to 37.5 37.5 to 50.0 50.0 to 62.5 Change in Annual Evaporative Deficit (mm/month)

RCP4.5 RCP8.5 RCP4.5 Mean: 13.7 mm/month; Range: 2.0 - 25.2 mm/month RCP8.5 Mean: : 21.0 mm/month; Range: 8.4 - 38.3 mm/month Data source: U.S. Geological Survey 2015

Pathogens.

The collected seeds of Guadalupe fescue sometimes contain an endogenous fungus (a fungus that lives within the host plant cells) (Poole 1989, p. 16). Investigators have not determined whether this fungal infection occurs naturally or is caused by handling and storage in seed banks, nor what effect this has on seed germination, viability, and vigor. This fungus may interfere with seed banking and propagation work conducted at Desert Botanical Gardens and elsewhere, and potentially threatens wild populations (Poole 2007). Nevertheless, we currently do not know if this endogenous fungus threatens Guadalupe fescue.

Summary.

The most important factors that may affect the continued survival of Guadalupe fescue include the demographic and genetic consequences of small population sizes and low population densities, the isolation of extant populations, changes in the wildfire cycle and vegetation structure, and competition from invasive species.

In Section II.8., we projected where potential habitats for Guadalupe fescue may be distributed. This includes 283 areas, totaling about 538,000 ha, in northern Mexico. Much of this land is privately owned or ejidos, where livestock grazing is a major land use. The only confirmed

Guadalupe Fescue Species Status Assessment – Aug 2016

population in Mexico is in the Maderas del Carmen mountains. We do not know if any other potential habitats in Mexico are occupied, but if they are, livestock grazing may be an important factor in their continued existence.

Trampling from humans and pack animals and erosion from trail runoff are potential factors at Big Bend National Park, but these factors have been alleviated through measures described in an approved Candidate Conservation Agreement. These factors arise from visitor use of a national park and are unlikely to affect the population at Maderas del Carmen.

It is likely that the projected climate changes, such as increased temperatures and changes in the amount and pattern of precipitation, will affect the survival of Guadalupe in complex ways; however, we do not currently know what the net result of beneficial and detrimental effects will be.

Fungal infection of seeds has been listed as a potential concern to Guadalupe fescue, but we have insufficient information to determine whether this is an actual factor affecting the species.

Guadalupe Fescue Species Status Assessment – Aug 2016

V. Conservation Efforts.

Texas.

In 1998, the Service and Big Bend National Park signed a Candidate Conservation Agreement for the conservation of Guadalupe fescue (Big Bend National Park and Service 1998), which expired in April 2005. The Service, Big Bend National Park, and Guadalupe Mountains National Park signed an updated 10-year Candidate Conservation Agreement on August 26, 2008 (Big Bend National Park and Service 2008). The 2008 agreement continues many of the provisions initiated under the 1998 agreement, including monitoring, seed banking, fire management, trail and visitor management, and establishment of an advisory team of species experts. The National Seed Storage Laboratory, Fort Collins, Colorado, has established a Guadalupe fescue seed bank to enable the restoration of populations lost to unanticipated catastrophic events. The 2008 agreement added new actions, including educating staff and visitors, monitoring and controlling invasive species, and cooperation with Mexican agencies and researchers to conserve the known populations and search for new ones. Scientific research objectives include the potential role of fire and other habitat management strategies, genetic structure and reproductive biology, continued surveys at Guadalupe Mountains National Park and in Coahuila, establishment of a germ plasm (live plant) bank, and techniques for reintroduction of the species. By May 2013, actions completed under the 2008 agreement included: Addition of Guadalupe fescue issues in National Environmental Protection Act (NEPA) reviews of Big Bend National Park management programs; annual briefings on Guadalupe fescue issues to trail crews and law enforcement rangers; suppression of human- ignited fires in the Guadalupe fescue habitat area; prevention of invasive species in the Guadalupe fescue habitat area; review of trail drainage plan; Big Bend National Park fire plan drafted that addresses Guadalupe fescue issues; annual monitoring of Guadalupe fescue population (discussed in section II.3); and a survey of the Maderas del Carmen populations (Sirotnak 2013, pp. 1–3; Valdes-Reyna 2009, pp. 1–23).

Guadalupe fescue is not legally protected by the State of Texas. The National Park Service manages all species on their lands in accordance with the National Park Service Organic Act of 1916, but there are no specific regulatory prohibitions protecting Guadalupe fescue.

Mexico.

Guadalupe fescue is not specifically protected under Mexican federal law (SEMARNAT 2010, p. 1; Valdes-Reyna 2010). However, three of the four Guadalupe fescue populations reported from Coahuila occur in two types of Protected Natural Areas (Areas Naturales Protegidas; ANP) designated by the Mexican federal government. One current and one historic population are in APFF Maderas del Carmen, and a second historic population is in APRN Sierra la Madera. These ANP designations give the Mexican federal government certain authorities over uses and development of the lands, but not ownership. Most ANP lands continue to be owned and used by private, corporate, or ejido entities. We know that as recently as 2009 the habitat and Guadalupe fescue population at Maderas del Carmen were intact, and the area is no longer grazed by livestock. We have no information on habitats, populations, or land uses at ANP Sierra la Madera.

Guadalupe Fescue Species Status Assessment – Aug 2016

VI. Current Status and Viability.

Texas.

Despite intensive searches, Guadalupe fescue has not been observed in McKittrick Canyon at Guadalupe Mountains National Park, Texas, where the type specimens were collected, since 1952 (Table 7). This population is presumed extirpated. The only known extant U.S. population is in the Chisos Mountains at Big Bend National Park, Texas. The site is protected and well managed. Nevertheless, the monitored portion of the population has declined from 127 individuals in 1993 to 47 individuals in 2014 (Table 7, Appendix B). Although we do not know what has caused this decline, we believe that it is likely to be due to insufficient recruitment. The data from these plots represent an estimated 25 to 50 percent of the total population at that site (Sirotnak 2014, p. 1). Therefore, we estimate that the current total population at the Chisos Mountains site is at least 100 to as many as 200 individuals.

Mexico.

In 2003, National Park Service personnel observed three colonies of Guadalupe fescue at the Sierra Maderas del Carmen that each had several hundred plants (Big Bend National Park and Service 2008, p. 3). Valdés-Reyna (2009) provided the most recent confirmation of an extant population there (Table 7), but he did not determine the population size or area occupied. He also looked for the Sierra El Jardín population, reported in 1973, but he did not find the species there. He was unable to access the population in Sierra la Madera and did not know about the population reported in 1941 northwest of Frailes, Coahuila.

We have no other information on the status of Guadalupe fescue in Mexico. Our estimate of potential habitats (section II.8) reveals that much larger amounts of potentially suitable habitat exist in northern Mexico than in Texas. A true understanding of the status of Guadalupe fescue should include a thorough survey of potential habitats in Mexico to determine whether or not extant populations occur there. However, due to prevailing security issues in northern Mexico, we do not know if or when these sites can be safely visited.

Guadalupe Fescue Species Status Assessment – Aug 2016

Table 7. Summary of current status of Guadalupe fescue populations.

Population Nation Most Recent Pop Size Occupied Recruitment Observation Habitat Guadalupe U.S. 1952 Extirpated Not Not Mountains applicable applicable Chisos U.S. 2015 Plots: 47–127 5 ha (12 ac) Low Mountains individuals; Estimated total: 100 – 200 individuals. Sierra Maderas Mexico 2009 3 colonies, each Unknown Unknown del Carmen with several hundred individuals Sierra El Mexico 1973 Unknown Unknown Unknown Jardin Sierra la Mexico 1977 Unknown Unknown Unknown Madera Ejido El Fraile Mexico 1941 Unknown Unknown Unknown

Guadalupe Fescue Species Status Assessment – Aug 2016

VI.1. Current Viability.

Resilience.

We estimate, provisionally, that viable populations should have at least 500 to 1,000 individuals. Of the two known extant populations, Maderas del Carmen may have had this many individuals as recently as 2003; the Chisos Mountains population is far below this level. Therefore, the resilience of both of these populations is considered low. Since the number of larger potential habitat patches is much greater in northern Mexico, it is possible that some of these support viably large populations; however, we do not know if populations survive in any of these un- surveyed potential habitats.

Redundancy.

Six populations of Guadalupe fescue have been documented since 1931. Currently, there are only two known extant populations, one is thought to have been extirpated, and three have not been seen in at least four decades. While it is difficult to say, definitively, how many populations are necessary for long-term viability, only two populations provide almost no redundancy.

Our geographic analysis of potential habitats (section II.8) indicates that there are 303 potential habitat patches greater than 200 ha (494 ac) in size, 20 in Texas and 283 in Mexico.

Representation.

Since only two extant populations of Guadalupe fescue are known, the species’ representation is presumed low. We know that the historic range of Guadalupe fescue extended at least as far south as southern Coahuila; we do not know whether it still occurs there or how much further south the range actually extends.

Summary

Our conclusion is twofold. Based strictly on the best available information, the species’ current resilience, redundancy, and representation are all very limited. However, the viability of Guadalupe fescue may be higher than we currently know based on the presence of extensive areas of potential habitat in Mexico. Although livestock grazing in many parts of Mexico is severe, the sheer remoteness and ruggedness of sky islands in the Chihuahuan Desert may have protected at least some sites. Also, wildfires have not been suppressed in Mexico as extensively as in the U.S., and in many areas fire is still used as a land management tool; thus, it is likely that Mexican forests have experienced more frequent, but less catastrophic wildfires, and this may be an important factor for maintaining Guadalupe fescue habitat.

Table 8 (below) summarizes the requirements, factors affecting survival, and current conditions of individuals and populations of Guadalupe fescue. The species’ overall viability is expressed in terms of its representation, redundancy, and resiliency.

Guadalupe Fescue Species Status Assessment – Aug 2016

Table 8. Summary of requirements, factors affecting survival, and current conditions of Guadalupe fescue individuals and populations, and the species’ viability (representation, redundancy, and resilience).

INDIVIDUALS POPULATIONS SPECIES I. Requirements of Guadalupe fescue. • Habitats: Understory of pine • Resiliency: Estimated MVP • Representation: Unknown and pine-oak woodlands in of 500 to 1,000. diversity; Known range Chihuahuan Desert at • Population size probably extends from Guadalupe elevations > 1,800 m. correlated to accumulated Mountains, Texas, to • Reproduction: depends on rainfall over multiple years. southern Coahuila, with wind pollination; needs • Larger habitat areas (>200 potential habitat further clustering of large numbers ha) probably able to support south. of individuals for effective larger populations with • Redundancy: Populations fertilization; seed dispersal greater genetic diversity, and should be distributed through deer. have greater potential for throughout range. • Breeding system: unknown future survival. (i.e. out-crosser or capable of • Role of fire is unknown, but self-fertilization). periodic fire may be • Rainfall: Reproduction and beneficial. survival rates are positively correlated with increasing rainfall over long terms.

Guadalupe Fescue Species Status Assessment – Aug 2016

II. Factors affecting the survival of Guadalupe fescue. • Livestock grazing probably • Small population size • Climate changes could reduces survival. increases risk of loss from reduce the species’ potential • Competition from invasive stochastic events and from range, both in latitude as species probably reduces reduced genetic fitness. well as elevation. If optimal survival. • Reduced fire frequency may elevations increase, the • Pathogens potentially reduce reduce population size, and amount of potential habitats survival. increases risk of catastrophic decreases. • Recreational activities could fire. lead to trampling of • Climate changes could individuals. reduce the suitability of habitats and reduce survival rates. • Isolated populations are subject to genetic drift; small populations prone to inbreeding and loss of genetic diversity. III. Current Conditions of Guadalupe fescue. • Survival rates of individuals • Population size at Big Bend • Representation is limited. at Big Bend National Park National Park has decreased The two populations known increased from 1993 to 2014. from 1993 to 2014; to be extant are less than 100 • Recruitment at Big Bend monitored population is km apart, and probably National Park may be too currently 47 individuals. represent a small fraction of low to sustain population. • Population at Maderas del the species’ genetic Carmen was probably of variation. sufficient size MVP in 2003; Redundancy is limited. Of 6 population observed in 2009, reported populations, only 2 but status unknown. are known to be extant, 1 is • Maderas del Carmen believed extirpated, and 3 population may be large have unknown status. enough to exceed MVP.

Guadalupe Fescue Species Status Assessment – Aug 2016

VII. Assessment of Future Viability.

We project what the viability of Guadalupe fescue could be in the future, in the range of 25 to 50 years from now. We chose this time frame as a period for which we can reasonably project potential habitat management and survey efforts to be conducted and to account for potential effects related to climate change. Given a life span of 3 to 4 years, this time frame would likely represent more than 10 generations and provides enough time to consider population-level responses to changing environmental conditions.

We considered future projections under three scenarios. The “better than expected” scenario represents substantial improvements over current conditions both in terms of management of current populations and the discovery of new populations. The “moderate” scenario represents some improvements over current conditions with some new or historical populations found. This moderate condition represents our estimate of the most likely future condition for the species. The “worse than expected” scenario represents deteriorating conditions and no new populations discovered. We describe, below, the relevant characteristics of these scenarios, and subsequently, their effects on populations. Table 9 summarizes our projections of the future species viability of Guadalupe fescue under each of these scenarios.

Better than Expected Scenario.

a. Conservation support: Government agencies, non-profit conservation organizations, and academic institutions, in both the U.S. and Mexico, collaborate and contribute sufficient human and financial resources to conserve Guadalupe fescue and its habitats. Private landowners and ejidos (in Mexico) are aware of the species and enthusiastically support its conservation. Development projects are evaluated and modified, if necessary, to avoid impacts to Guadalupe fescue and its habitats.

b. Surveys: Qualified botanists obtain access to survey a large number of the highest-potential habitats in the U.S. and Mexico. Both the presence and absence of Guadalupe fescue populations in these habitats contributes improved understanding of the species’ ecology, management, abundance, and true geographical range. c. Geographic range: Extant Guadalupe fescue populations are documented throughout the historic range (southern New Mexico to southern Coahuila) or (potentially) further south. d. Habitat management: Extant populations are managed appropriately. This may include prescribed burning, if studies of the species’ fire ecology confirm that periodic fire is necessary for long-term survival of populations. Habitat management may also include improved grazing management or excluding livestock from occupied habitats. e. Population management: Extant populations of Guadalupe fescue are monitored periodically to track demographic trends. Observed threats, such as invasive species or livestock encroachment, are prevented before populations are significantly impacted. Small, declining populations are recovered through facilitated augmentation of numbers and genetic diversity or other effective practices.

Guadalupe Fescue Species Status Assessment – Aug 2016

f. Climate changes: The effects of climate changes on Guadalupe fescue habitats are relatively moderate and are well tolerated by the species. g. Population responses:

Chisos Mountains Population. Population is well above minimum viable size and density and is increasing.

Maderas del Carmen Population. Periodic monitoring confirms that population is well above minimum viable size and is increasing.

El Jardín, Sierra la Madera, and Ejido el Fraile. Periodic monitoring confirms that populations are well above minimum viable size and are increasing.

Other potential, as yet undiscovered populations in the U.S. and Mexico. At least 5 additional viable populations well above minimum viable size are confirmed.

Moderate Scenario.

a. Conservation support: Government agencies, non-profit conservation organizations, and academic institutions, in both the U.S. and Mexico, collaborate and contribute some human and financial resources for conservation of Guadalupe fescue and its habitats. Public outreach has increased awareness of the species among private landowners and ejidos (in Mexico) and has generated increasing support for its conservation. Development projects are evaluated and modified, if necessary, to minimize or mitigate impacts to Guadalupe fescue and its habitats.

b. Surveys: Qualified botanists obtain access to survey a representative sample of the highest potential habitats in the U.S. and Mexico. Both the presence and absence of Guadalupe fescue populations in these habitats contributes improved understanding of the species’ ecology, management, and true geographical range.

c. Geographic range: New extant Guadalupe fescue populations are documented within the historic range (southern New Mexico to southern Coahuila) or further south.

d. Habitat management: At least some extant populations are managed appropriately. This may include prescribed burning, if studies of the species’ fire ecology confirm that periodic fire is necessary for long-term survival of populations. Habitat management may also include improved grazing management or excluding livestock from occupied habitats.

e. Population management: At least some extant populations of Guadalupe fescue are monitored periodically to track demographic trends. Observed threats, such as invasive species or livestock encroachment, are prevented before populations are significantly impacted. Small, declining populations are recovered through facilitated augmentation of numbers and genetic diversity or other effective practices.

Guadalupe Fescue Species Status Assessment – Aug 2016 f. Climate changes: Climate changes have significant impacts on Guadalupe fescue habitats, but with appropriate management the species’ overall status remains stable. g. Population responses:

Chisos Mountains Population. Population is stable or increasing but remains below the MVP size.

Maderas del Carmen Population. Population is stable or increasing but remains below the MVP size.

El Jardín, Sierra la Madera, and Ejido el Fraile. One or more of these populations are found to be extant, are stable or increasing, but remain below the MVP size.

Other potential, as yet undiscovered populations in the U.S. and Mexico. Less than 5 new populations are discovered and found to be stable or increasing but below the MVP size.

Worse than Expected Scenario. a. Conservation support: Government agencies, non-profit conservation organizations, and academic institutions, in both the U.S. and Mexico, fail to collaborate or contribute sufficient human and financial resources to conserve extant Guadalupe fescue populations and high potential habitats. Landowners and ejidos (in Mexico) remain largely unaware of the species and are unsupportive of its conservation. Development projects significantly impact Guadalupe fescue and its habitats. b. Surveys: Qualified botanists are unable to access representative samples of the highest potential habitats in the U.S. and Mexico. Nothing new is learned about the species’ ecology, management, and true geographical range. c. Geographic range: No new extant Guadalupe fescue populations are documented, or if additional populations are found, they cannot be protected or conserved. d. Habitat management: Known extant populations are not managed appropriately. e. Population management: Known extant populations of Guadalupe fescue are not monitored periodically. f. Climate changes: Climate changes have severe impacts on Guadalupe fescue habitats and the species’ overall status declines. g. Population responses:

Chisos Mountains Population. The population continues to decline.

Guadalupe Fescue Species Status Assessment – Aug 2016

Maderas del Carmen. Site cannot be monitored, and population size, extent, and viability remain unknown.

El Jardín, Sierra la Madera, and Ejido el Fraile. Populations continue to be presumed extirpated.

Other potential, as yet undiscovered populations in the U.S. and Mexico. No new populations are found and none are presumed to exist.

Guadalupe Fescue Species Status Assessment – Aug 2016

Table 9. Future species viability under a range of scenarios.

Viability Scenarios Elements Better than Expected Moderate Worse than Expected Population All populations (>10) All populations (5-10) show No new populations Resilience meet or exceed MVP of positive demographic are found. The two 500 to 1,000 individuals trends but remain below known extant and show stable or MVP levels. populations at Chisos positive demographic Mountains and trends. Maderas del Carmen decline or become extirpated. Species More than 10 5 to 10 populations would Species has virtually Representation populations across the provide some limited no representation with range would provide genetic and ecological only two relatively enhanced genetic and diversity but representation small populations. ecological diversity for is limited by low population the species. resiliency. Species More than 10 Some limited redundancy is Species has nearly no Redundancy populations across the provided by having 5-10 redundancy with only range would provide populations present across a two relatively small enhanced redundancy for larger geographic range but populations. the species to withstand is limited by low population catastrophic events. resiliency.

Guadalupe Fescue Species Status Assessment – Aug 2016

VIII. Literature Cited.

Aiken, S.G. and L.L. Consaul. 1995. Leaf cross sections and phytogeography: a potent combination for identifying members of Festuca subgg. Festuca and Leucopoa (Poaceae) occurring in North America. American Journal of Botany 82:1287–1299.

Aiken, S.G., M.J. Dallwitz, C.L. McJannet, and L.L. Consaul. 1996 onwards. Festuca of North America: Descriptions, illustrations, identification, and information retrieval. Version: 19th October 2005. http://delta-intkey.com/. Accessed March 6, 2009.

Appelhons, D.J. 1973. Grassland vegetation of a relict community in the Franklin Mountains, El Paso County, Texas. Master’s Thesis. University of Texas at El Paso, El Paso, Texas. 59 leaves.

Arnberger, C. 2015. Email from Claudia Arnberger, Big Bend National Park, to Chris Best, U.S. Fish and Wildlife Service. Subject: Question on Boot Canyon Trail Use/Camping Permits. December 8, 2015.

Barrett, S.C.H. and J.R. Kohn. 1991. Genetic and evolutionary consequences of small population size in plants: Implications for conservation. Pp. 3–30 in D. Falk and K.E. Holsinger, eds. Genetics and Conservation of Rare Plants. Oxford University Press, New York. 283 pp.

Big Bend National Park and U.S. Fish and Wildlife Service (Service). 1998. Conservation Agreement for Castilleja elongata (Tall Paintbrush) and Festuca ligulata (Guadalupe Fescue). U.S. Fish and Wildlife Service, Austin Ecological Services Field Office. 17 pp. + appendices.

Big Bend National Park and the U.S. Fish and Wildlife Service (Service). 2008. Candidate Conservation Agreement for Festuca ligulata (Guadalupe Fescue). U.S. Fish and Wildlife Service, Austin Ecological Services Field Office. 18 pp. + 5 pp. appendices.

Biology-online.org. 2011. http://www.biology-online.org/dictionary/. Accessed June 30, 2011.

Camp, A., H.M. Poulos, R. Gatewood, J. Sirotnak, and J. Karges. 2006. Assessment of top down and bottom up controls on fire regimes and vegetation abundance and distribution patterns in the Chihuahuan Desert borderlands: A hierarchical approach. Final Report to the Joint Fire Science Program # 03-3-3-13. Yale University School of Forestry and Environmental Studies. New Haven, Connecticut. 77 pp.

Clinton, President William J. 1999. Executive Order 13112 on invasive species. 64 FR:6183– 6186. February 3, 1999.

Comisión Nacional de Areas Naturales Protegidas. 2010. Areas Naturales Protegidas y áreas destinadas voluntariamente a la conservación. Actualización: 15 febrero, 2010. http://www.conanp.gob.mx/sig/información/info.htm. Accessed: March 18, 2010.

Guadalupe Fescue Species Status Assessment – Aug 2016

Correll, D.S. and M.C. Johnston. 1979. Manual of the Vascular Plants of Texas. The University of Texas at Dallas. 1,881 pp.

Darbyshire, S.J. and L.E. Pavlick. 2015. Grass Manual on the Web. 14.01. Festuca L. Utah State University Herbarium. http://herbarium.usu.edu/treatments/Festuca.htm. 41 pp. Accessed: May 6, 2015.

Edmands, S. 2007. Between a rock and a hard place: Evaluating the relative risks of inbreeding and outbreeding for conservation and management. Molecular Ecology 16:463–475.

Fowler, N.L. 2015a. MS Word document describing results of SAS analyses of Guadalupe fescue lifespan, survival rate, and correlations. 9 pp.

Fowler, N.L. 2015b. Email from Norma Fowler, University of Texas Department of Integrative Biology, to Chris Best, U.S. Fish and Wildlife Service. Subject: Correlations, sample sizes. 1 p. + attached Excel file: samples sizes & corrs 3-26-2015.xlsx. March 26, 2015.

Fowler, N.L. 2015c. MS Word document describing SAS GLIMMIX procedure. 6 pp.

Fowler, N.L. 2015d. Graphical output form SAS GLIMMIX procedure. 1 p.

Fowler, N.L. 2015e. Email from Norma Fowler, University of Texas Department of Integrative Biology, to Chris Best, U.S. Fish and Wildlife Service. Subject: Guadalupe Fescue Breeding System. April 12, 2015. 1 p.

Frankham, R., J.D. Ballou, M.D.B. Eldridge, R.C. Lacy, K. Ralls, M.R. Dudash, and C.B. Fenster. 2011. Predicting the probability of outbreeding depression. Conservation Biology 25:465–475.

Friedman, J. and S.C.H. Barrett. 2009. Wind of change: New insights in the ecology and evolution of pollination and mating in wind-pollinated plants. Annals of Botany 103:1515–1527.

Fryxell, P.A. 1957. Mode of Reproduction in Higher Plants. Bot. Rev. 23:135 – 233.

Ge, Y., F. Chunxiang, H. Bhandari, J. Bouton, E.C. Brummer, and Z. Wang. 2011. Pollen viability and longevity of switchgrass (Panicum virgatum L). Crop Science 51:1–8.

Gordon, W. and J.M. Poole. 2009. Texas Parks and Wildlife Department. Letter dated March 5, 2009. 25 pp.

Gould, F.W. 1975. The Grasses of Texas. Ill. V. Kapadia. Texas A&M Press. College Station, Texas. 653 pp.

Guadalupe Fescue Species Status Assessment – Aug 2016

Greer, M.J., G.W.T. Wilson, K.R. Hickman, and S.M. Wilson. 2014. Experimental evidence that invasive grasses use allelopathic biochemicals as a potential mechanism for invasion: Chemical warfare in nature. Plant Soil 385:165-179.

Integrated Taxonomic Information System. 2011. Festuca ligulata Swallen. http://www.itis.gov/servlet/SingleRpt/SingleRpt?search_topic=TSN&search_value=4080 3. Accessed: May 4, 2011. 3 pp.

Intergovernmental Panel on Climate Change (IPPC). 2013. Summary for Policymakers. In T.F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.), Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, New York, USA. 29 pp.

Janzen, D.H. 1984. Dispersal of small seeds by big herbivores: Foliage is the fruit. The American Naturalist 123:338–353.

Jefferson, L.V., M. Pennacchio, K. Havens, B. Forsberg, D. Sollenberger, and J. Ault. 2008. Ex Situ Germination Responses of Midwestern USA Prairie Species to Plant-Derived Smoke. The American Midland Naturalist 159:251–256.

Krausman, P.R. and E.D. Ables. 1981. Ecology of the Carmen Mountains White-Tailed Deer. Scientific Monograph Series No. 15. National Park Service, Washington, D.C. https://www.nps.gov/parkhistory/online_books/science/15/chap3.htm. Accessed: July 22, 2016.

Landis, T.D. 2000. Where there’s smoke...there’s germination? Native Plants Journal 1:25–29.

Loflin, B. and S. Loflin. 2006. Grasses of the Texas Hill Country. Texas A&M University Press, College Station, Texas. 195 pp.

Mace, G.M. and R. Lande. 1991. Assessing extinction threats: Toward re-evaluation of IUCN threatened species categories. Conservation Biology 5:148–157.

Meyer, R. 2009. Festuca thurberi. In: Fire Effects Information System [Online]. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory. http://www.fs.fed.us/database/feis/. Accessed: March 28, 2015.

Milly, P.C.D., K.A. Dunne, and A.V. Vecchia. 2005. Global pattern of trends in stream flow and water availability in a changing climate. Nature 438:347-350.

Moir, W. 1982. A fire history of the high Chisos, Big Bend National Park, Texas. The Southwestern Naturalist 27:87–98.

Guadalupe Fescue Species Status Assessment – Aug 2016

Moir, W.H. and J. Meents. 1981. Fire ecology study of the Chisos Mountains, Big Bend National Park, Texas: Effects of a prescribed burn in Boot Canyon and description of two new reference plots. CDRI Contribution No. 117. Final report of contract no. CX 70298005 for the National Park Service. Chihuahuan Desert Research Institute, Alpine, Texas. 25 pp. + appendix.

National Center for Biotechnology Information. 2015. Festuca ligulata tRNA-Leu (trnL) gene and trnL-trnF intergenic spacer, partial sequence; chloroplast. http://www/ncbi.nlm.nih.gov/nuccore/157266456?report=gbwithparts. Accessed: March 28, 2015.

National Climatic Data Center. 2015. Monthly Climatological Summaries 1991 – 2014, Panther Junction, Texas, US. http://www.ncdc.noaa.gov/cdo-web/search. Accessed: February 20, February 23, and March 6, 2015.

Natural Resources Conservation Service. 2011. The PLANTS Database. Festuca ligulata Swallen. (http://plants.usda.gov/java/nameSearch?keywordquery=Festuca+ligulata&mode=scinam e&submit.x=12&submit.y=4). Accessed: May 4, 2011. National Plant Data Center, Baton Rouge, LA 70874-4490 USA. 3 pp.

Pavlik, B.M. 1996. Defining and Measuring Success. Pp. 127–155 in D.A. Falk, C.I. Millar, and M. Olwell (eds.), Restoring Diversity: Strategies for reintroduction of endangered plants. Island Press, Washington, D.C. 505 pp.

Pedersen, J.F. and D.A. Sleper. 1993. Genetic manipulation of tall fescue. Agriculture, Ecosystems and Environment 44:187–193.

Poole, J.M. 1989. Status survey on Festuca ligulata. Texas Natural Heritage Program, Texas Parks and Wildlife Department. Unpublished report. U.S. Fish and Wildlife Service, Austin Ecological Services Field Office. 23 pp. + maps and figures.

Poole, J.M. 2007. Telephone conversation with Jackie Poole, Botanist, Texas Parks and Wildlife Department, Texas (2007).

Poole, J.M., W. R. Carr, D.M. Price, and J.R. Singhurst. 2007. Rare Plants of Texas. Texas A&M Press, College Station, Texas. 639 pp.

Poulos, H.M. 2014. Tree mortality from a short-duration freezing event and global-change-type drought in a Southwestern piñon-juniper woodland, U.S.A. PeerJ 2:e404; DOI 10.7717/peerj.404.

Reilley, J. 2016. Email from John Reilley, Natural Resources Conservation Service, to Chris Best, U.S. Fish and Wildlife Service. April 6, 2016. 1 p.

Guadalupe Fescue Species Status Assessment – Aug 2016

Rognli, O.A., N. Nilsson, and M. Nurminiemi. 2000. Effects of distance and pollen competition on gene flow in the wind-pollinated grass Festuca pratensis Huds. Heredity 85:550–560.

SAS Institute Inc. 2015. The GLIMMIX Procedure: Overview. SAS/STAT® 9.2 User's Guide, Second Edition. Http://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/statug glimmix. Accessed: April 21, 2015.

Secretaría del Medio Ambiente y Recursos Naturales (SEMARNAT). 2010. Normas Oficiales Vigentes Ordenadas por Materia. NOM-059-ECOL-2001. http://www.semarnat.gob.mx/leyesynormas/Pages/normasoficialesmexicanasvigentes.asp x. México, D.F. Accessed: March 16, 2010.

Shaffer, M.L. and B.A. Stein. 2000. Safeguarding our Precious Heritage. Pages 301–321 in B.A. Stein, L.S. Kutner, and J.S. Adams (eds.), Precious Heritage: The status of biodiversity in the United States. Oxford University Press. 399 pp.

Shaw, R.B. 2012. Guide to Texas Grasses. Texas A&M University Press, College Station, Texas. 1080 pp.

Sirotnak, J. 2013. Email from Joe Sirotnak, National Park Service, to Chris Best, U.S. Fish and Wildlife Service. Subject: FELI Conservation Agreement task table, May 2013 update. 1 p. + attached MS Word file: FELI Conservation Agreement Task Table Progress May 2013.docx. May 24, 2013.

Sirotnak, J. 2014. Email from Joe Sirotnak, National Park Service, to Chris Best, U.S. Fish and Wildlife Service. Subject: Bumming Guadalupe Fescue Updates.

Sirotnak, J. 2015. Email from Joe Sirotnak, National Park Service, to Chris Best, U.S. Fish and Wildlife Service. Subject: Re: 2014 Guadalupe Fescue Data. 1 p. + attached Excel file: FELI Raw data through 2014.xlx. January 23, 2015.

Sirotnak, J. 2016. Comment JS1 on draft Incremental Effects Memorandum for Guadalupe Fescue, received January 20, 2016.

Swallen, J.R. 1932. Five new grasses from Texas. American Journal of Botany 19:436–440.

Texas Natural Diversity Database. 2007. Element occurrence printouts for Festuca ligulata. Wildlife Diversity Program of Texas Parks and Wildlife Department. April 19, 2007. 9,079 pp.

Tropicos.org. 2011. Missouri Botanical Garden. http://www.tropicos.org/Specimen/1984542. Accessed: May 4, 2011. 1 pp.

Guadalupe Fescue Species Status Assessment – Aug 2016

U.S. Fish and Wildlife Service (Service). 2015. U.S. Fish and Wildlife Service species status assessment framework: an integrated analytical framework for conservation. Version 3.3. August 2015.

U.S. Geological Survey. 2014. National Climate Change Viewer Tutorial and Documentation. http://www.usgs.gov/climate_landuse/clu_rd/nccv/documentation_v1.pdf. Accessed: January 26, 2016.

U.S. Geological Survey. 2015. National Climate Change Viewer. http://www.usgs.gov/climate_landscape/clu_rd/nccv/viewer.asp. Accessed: January 26, 2016.

Valdés-Reyna, J. 2009. Gramíneas raras del desierto Chihuahuense: Estatus de Festuca ligulata en Coahuila, México. Reporte Final, Contrato OK44. Submitted to World Wildlife Fund and National Park Service. 23 pp.

Valdes-Reyna, J. 2010. Email from Jesús Valdes-Reyna, Universidad Autónoma Agronómica Antonio Narro, to Chris Best, Service. Re: Preguntas sobre Festuca ligulata. March 16, 2010. 1 pp.

Van Devender, T.R. and W.G. Spaulding. 1979. Development of vegetation and climate in the Southwestern United States. Science 204: 701–710.

Wang, Z., Y. Ge, M. Scott, and G. Spangenberg. 2004. Viability and longevity of pollen from transgenic and nontransgenic tall fescue (Festuca arundinacea) (Poaceae). American Journal of Botany 91:523–530.

Waring, E.F. and D.W. Schwilk. 2014. Plant dieback under exceptional drought driven by elevation, not by plant traits, in Big Bend National Park, Texas, USA. PeerJ 2:e477; DOI 10.7717/peerj.477.

Wikipedia. 2011-2015. Articles on the following glossary topics: Bonferroni correction, breeding system, chromosome, correlation, Digital Elevation Model, gene flow, genetic structure, genotype, habitat, non-parametric statistics, normal distribution, Pleistocene, populations, seral, shapefile, and significance. http://www.wikipedia.org. Accessed 2011, 2012, 2013, and 2015.

Worthington, R. D. 1991. Letter to the Texas Natural Heritage Program, March 13. Texas Parks and Wildlife Department files, Austin, Texas. 1 p.

Zimmerman, J. A. C. and W. H. Moir. 1998. Conservation status of Festuca ligulata in the Chisos Mountains, Big Bend National Park. RM-4251. Sustainability of Southwestern Forest and Woodland Ecosystems. USDA Forest Service. 15 pp.

Guadalupe Fescue Species Status Assessment – Aug 2016

IX. Photographic Credits.

All photographs may be used without permission with credit to “Chris Best, U.S. Fish and Wildlife Service.”

X. Acronyms Used.

ANP Area Natural Protegida NEPA National Environmental Policy Act APFF Area de Protección de Flora y NPS National Park Service Fauna BIBE Big Bend National Park PVA Population Viability Analysis CONANP Comisión Nacional de Areas SEMARNAT Secretaría de Medio Ambiente y Naturales Protegidas Recursos Naturales GUMO Guadalupe Mountains National TPWD Texas Parks and Wildlife Park Department INEGI Instituto Nacional de Estadística TXNDD Texas Natural Diversity Database y Geografía (Mexico) IPCC Intergovernmental Panel on Climate Change MVP Minimum Viable Population

XI. Scientific Units. ac acre in. inch C° Degrees Celsius km kilometer cm centimeter m meter F° Degrees mi mile Fahrenheit ft feet mm millimeter ha hectare

Guadalupe Fescue Species Status Assessment – Aug 2016

Appendix A. Glossary of Scientific and Technical Terms.

Term Definition Allele Alternate forms of a gene. Anther The pollen-bearing part of the stamen. (Correll and Johnston 1979). Area Natural (Spanish) Protected Natural Area. In Mexico, may be designated by state Protegida governments, or by Comisión Nacional de Areas Naturales Protegidas (CONANP), a federal government agency. Awn A bristle or hair-like prolongation of the nerves of a bract of a grass spikelet (Shaw 2012, p. 1045). Blade The (usually) flattened, expanded portion of a grass leaf above the sheath (Shaw 2012, p. 1045–1046). Bonferroni A method in inferential statistics that corrects for the increasing probability of correction for incorrectly rejecting the null hypothesis (Type I error) when testing multiple multiple tests hypotheses on a set of data (Wikipedia 2015). Bract A reduced leaf subtending a flower, usually associated with an inflorescence (Correll and Johnston 1979, p. 1747). Breeding System The ability of a plant species to reproduce via outcrossing, self-fertilization, apomixis, or a combination (Wikipedia 2015). Bunch-grass Grass that reproduces vegetatively through the proliferation of tillers from basal bud primordia. Chloroplast A double-membrane organelle found in higher plants in which photosynthesis takes place. Chromosome An organized structure consisting of DNA and protein containing a cell's genes, regulatory elements, and other nucleotide sequences (Wikipedia 2013). Cool-season A grass that grows primarily in late fall, winter, or early spring (Loflin and grass Loflin 2006). Correlation A statistically dependent relationship between two random variables or sets of data (Wikipedia 2015). Digital Elevation Digital model or 3D representation of a terrain's surface — commonly for a Model planet (including Earth), moon, or asteroid — created from terrain elevation data (Wikipedia 2015). Ecotype A genotype that is specifically adapted to a particular ecological area. Effective The size of an idealized population in which individuals contribute equally to population size the gamete pool and have the same variation in allele frequencies and levels of inbreeding as the observed population (Barrett and Kohn 1991). Ejido (Spanish) Collectively owned agricultural cooperative in Mexico. Endemic An organism restricted to a specific habitat or geographic range. Endogenous Contained within the tissues of an organism. Floret A single grass flower enclosed by a palea and lemma (Shaw 2012, p. 1047). Forb A broad-leafed herbaceous plant.

Guadalupe Fescue Species Status Assessment – Aug 2016

Gene flow The transfer of alleles or genes from one population to another (Wikipedia 2013). Gene sequence The sequence of nucleotide bases in a DNA molecule that constitute a gene. Generalized A type of statistical model fit to data with correlations or non-constant linear mixed variability and where the response is not necessarily normally distributed models (SAS 2015). Genetic drift A change in allele frequencies within a population over time. Genetic structure Any pattern in the genetic makeup of individuals within a population (Wikipedia 2011). Genotype The genetic composition of a cell, organism, or individual (Wikipedia 2012). Germ bank Genetic repository consisting of living tissues of organisms. Greenhouse gas Gases such as carbon dioxide, water vapor, and methane that contribute to the atmosphere's thermal insulation through absorption of light in the infra-red spectrum. Habitat Ecological or environmental area that is inhabited by a particular species of animal, plant or other type of organism (Wikipedia 2013). Herbarium A repository for long-term storage and study of preserved plant specimens. Historic A previously documented population that has been extirpated or can no longer population be found. Inbreeding Sexual reproduction between closely related individuals. Inbreeding The reduction of fitness caused by mating between relatives (Edmands 2007, depression p. 464). Inflorescence A plant structure bearing two or more flowers. Invasive Species that is non-native (or alien) to the ecosystem under consideration and whose introduction causes or is likely to cause economic or environmental harm or harm to human health (Clinton 1999; 64 FR:6183–6186, February 3, 1999). Lemma "The lowermost of the two bracts enclosing the flower in the grass floret…" (Shaw 2012, p. 1049). Ligule "A membranous or hairy appendage on the adaxial surface of a grass leaf at the junction of sheath and blade…" (Shaw 2012, p. 1049 Minimum viable The fewest individuals required for a specified probability of survival over a population specified period of time (Pavlik 1996; Mace and Lande 1991); see Population Viability Analysis. Mycorrhiza Mutualistic association of higher plant roots and specialized soil fungi. Non-parametric Statistics that are not based on parameterized families of probability statistics distributions…and that make no assumptions about the probability distributions of the variables being assessed (Wikipedia 2015). Normal In probability theory, a continuous probability distribution that is symmetric distribution about the mean and non-zero over the entire real line (Wikipedia 2015). Outbreeding The reduction in reproductive fitness in the first or later generations following depression attempted crossing of distinct populations (Frankham et al. 2011, p. 466).

Guadalupe Fescue Species Status Assessment – Aug 2016

Outcross In plants, sexual fertilization involving a different individual. Panicle As applied to grasses, any inflorescence where the spikelets are not sessile or individually pediceled on the main axis (Shaw 2012, p. 1050). Perennial A plant that lives for more than one full year. Pleistocene Geological epoch beginning about 2,588,000 years ago and ending about 11,700 years ago (Wikipedia 2013). Population Collection of inter-breeding organisms of a particular species (Wikipedia 2013). Population Statistical models used to predict the probability of extinction of a population Viability after a specified period of time. Analysis Ramet An individual, genetically identical plant reproduced as a clone of the parent plant. Recruitment Addition of new individuals to a population. Redundancy The number of populations or sites necessary to endure catastrophic losses (Shaffer and Stein 2000, pp. 308–310). Representation The genetic diversity necessary to conserve long-term adaptive capability (Shaffer and Stein 2000, pp. 307–308). Resilience The size of populations necessary to endure random environmental variation (Shaffer and Stein 2000, pp. 308–310). Rhizome Horizontal stems that grow under the surface of the ground. Ruderal Early stage of succession (colonization). Seed banking Storage of plant seeds, under conditions that extend viability, for species conservation and restoration. Seral An intermediate developmental stage in ecological succession (Wikipedia 2013). Shapefile A digital geospatial vector data storage format developed by Esri. (Wikipedia 2015). Sheath The tubular basal portion of a grass leaf that surrounds the stem (Shaw 2012, p. 1051). Significance Statistical assessment of whether observations reflect a pattern rather than just chance (Wikipedia 2013). Sky island Isolated areas of relatively cool, temperate habitat in mountains surrounded by expanses of warm desert at lower elevation. Soil seed bank Dormant and non-dormant seeds present in the soil that are able to germinate. Speciation The evolutionary process by which new biological species arise (Wikipedia 2015). Spikelet The basic unit of a grass inflorescence, consisting of bracts and one or more florets arranged along a short axis (Shaw 2012, p. 1051). Stochastic Random.

Guadalupe Fescue Species Status Assessment – Aug 2016

Subgenus A subdivision of a genus, comprising one or more species which differ from other species of the genus in some important character or characters (Biology- online.org 2011). Talus Accumulation of loose rock fragments at the base of a slope. Taxonomy Scientific classification of living organisms. Type specimen A specimen upon which the description of a new species is based. Understory Vegetation occurring beneath a canopy of trees or shrubs.

Guadalupe Fescue Species Status Assessment – Aug 2016

Appendix B. Analyses of Monitoring Plot Data at Big Bend National Park, Texas.

Introduction.

Status assessments of rare plant species take into account available information on population sizes, survival rates, lifespans, and reproductive output, as well as trends in these parameters. The only Guadalupe fescue population that has been quantitatively monitored is in the Chisos Mountains at Big Bend National Park in Texas. The population is scattered at low density over an area of about 5 ha (12.4 ac). Teams directed by Jackie Poole (TPWD) and Joe Sirotnak (NPS) collected population data during the months of September and October for 15 years over a 22 year period (1993, 1994, 1996, 1997, 1998, 2000, 2001, 2002, 2003, 2005, 2006, 2008, 2010, 2013, and 2014) (Sirotnak 2015). They established six 5 m- (16.4 ft-) radius circular monitoring plots positioned within this population where clusters of Guadalupe fescue plants were observed at the beginning of the study in 1993 (Figure B1). These plots represent an estimated 25 to 50 percent of the total population at that site (Sirotnak 2014, p. 1); it was assumed that the plots would adequately represent trends for the entire population. Within each of the six plots they recorded numbers of Guadalupe fescue individuals, their reproductive outputs, and the polar coordinates of their positions. Individual plants were identified by their polar coordinates (i.e. plot position), and were later marked with aluminum identification tags, which made it possible to track individual plants over time and to estimate lifespans and survival rates. Sirotnak (2015) provided a copy of the data set to the Service, and Dr. Norma Fowler (University of Texas) and Dr. Rebeca Quiñonez-Piñón (formerly of Universidad Autónoma Metropolitana, México) assisted with many of the analyses.

Guadalupe Fescue Species Status Assessment – Aug 2016

Data Analyses

As used here, “census” indicates a count of individuals, and “survey” indicates all data collected, including census, measures of reproductive output, and position in plot. The numbers of individuals observed in a single census ranged from 47 to 127; a total of 526 individuals were observed over the 15 censuses. Since surveys were conducted in 15 years over a 22-year period, intervals between surveys were not consistent, but ranged from 12 to 36 months. To compensate for variable interval length, some analyses required adjusted precipitation or adjusted survival rates, as described below. We used a non-parametric test of correlation for some tests, Spearman’s rank correlation (indicated as “SRC”), because the data are not normally distributed. The survey teams were unable to find Plot 5 in 2000 and 2001, and collected data from a substitute, Plot 5a, in those years. Although Plots 5 and 5a had similar numbers of individuals, they were different individuals, so we excluded data from these two plots for analyses of survival rates and lifespan, as described below.

Guadalupe Fescue Species Status Assessment – Aug 2016

Population trends on monitoring plots.

Correlation with rainfall:

It is particularly important to understand how population size, survival rates, lifespans, and reproductive output are influenced by rainfall, since climate change models indicate that the amount and distribution of rainfall in the Chihuahuan Desert region may change during the current century (Intergovernmental Panel on Climate Change (IPCC) 2013, p. 16); see discussion in Section V.

Table B1 compares the numbers of individuals, panicles, and annual rainfall totals at Panther Junction (National Climatic Data Center 2015), which is less than 15 km (9 mi) from the Chisos Mountains population (although 910 m (2,986 ft) lower elevation). Note that annual precipitation totals include 3 months (October through December) after censuses were conducted. Tests of correlations used only the precipitation that accumulated in specified periods that ended in the month that censuses were conducted; hence, these periods were 9 months (January through September), 21 months (previous year plus January through September), and 33 months (previous two years plus January through September). Since the interval between censuses ranged from 12 to 36 months, we also created an “adjusted precipitation” data set, consisting of the average 12-month precipitation between subsequent surveys, for some analyses (as indicated).

Population size was correlated (SRC) with precipitation during the 33 months preceding each census (ρ = 0.590, p = 0.021) (Fowler 2015a). In this context, ρ is a measure of the strength of correlation and p indicates probability that tests are incorrect. A ρ of 1 and -1 indicate perfect positive and negative correlations, respectively, and 0 indicates a complete lack of correlation. A p value for a single test of 0.05 indicates a 0.95 probability that the data support the conclusion. However, the Bonferroni correction for multiple tests indicates that p should be less than 0.0102 for significance at a probability of 0.95 that all tests are correct (Fowler 2015b). Population size had only weak, non-significant correlations with rainfall over the previous 9 and 21 months (ρ = 0.43, p = 0.11; ρ = 0.34, p = 0.22) (Fowler 2015a). Population size was not correlated with adjusted precipitation (ρ = -0.147, p = 0.602). The data suggests that rainfall over longer periods appears to have more influence on population size than rainfall over shorter periods.

Correlation with time:

Population size was negatively correlated (SRC) with survey year (ρ = -0.62377, p = 0.0130) (Fowler 2015a, pp. 9–10). In other words, the population has declined over the 22-year period of study.

Guadalupe Fescue Species Status Assessment – Aug 2016

Table B1. Guadalupe Fescue Population and Precipitation Data at Big Bend National Park. Shading indicates years when surveys were conducted.

- - - - Y e a r - - - - 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Individuals 125 127 112 117 103 78 60 69 Panicles 60 150 8 352 227 2 39 23 Spikelets 843 2058 146 6567 3147 54 648 498 Total Precipitation (in) 23.06 15.05 15.6 7.97 8.3 14.45 17.32 13.09 9.55 13.75 7.63 13.83 Total Precipitation (cm) 58.6 38.2 39.6 20.2 21.1 36.7 44.0 33.2 24.3 34.9 19.4 35.1

Figure B2. Guadalupe fescue at Big Bend National Park: population size and annual precipitation.

70.0 140

60.0 120

(cm) 50.0 100

40.0 80 30.0 60 20.0 40 Precipitation 10.0 20 Number of Individuals 0.0 0

Year

Total Precipitation (cm) Individuals

Guadalupe Fescue Species Status Assessment – Aug 2016

Survival rates.

We calculated survival rates over each census interval as the fraction of individuals present at the beginning of the interval that survived to the following census; thus, there were 14 intervals between the 15 censuses. To convert the raw survival rates into 1-year survival rates (i.e., to correct for gaps in survey years), we calculated 1-year survival as the square root of survival over 2 year intervals and the cube root of survival over 3 year intervals (table B2). Data from plot 5 in 2000 and plot 5a in 2002 were excluded from analyses, due to the location issue discussed above. Since we expect that adjusted survival would be influenced by precipitation over the entire survey interval, we tested the correlation (SRC) of adjusted survival with adjusted precipitation; however, these data were weakly and non-significantly correlated (ρ = 0.359, p = 0.208).

Table B2. Survival rates of Guadalupe fescue at Big Bend National Park.

Census Interval Adjusted Interval Length (yrs) Initial No. Surviving No. Mortality Raw Survival Survival 1 1 125 111 14 0.89 0.89 2 2 127 29 98 0.23 0.48 3 1 112 61 51 0.54 0.54 4 1 117 19 98 0.16 0.16 5 2 53 14 39 0.26 0.51 6 1 78 44 34 0.56 0.56 7 1 23 22 1 0.96 0.96 8 1 69 66 3 0.96 0.96 9 2 77 73 4 0.95 0.97 10 1 96 88 8 0.92 0.92 11 2 97 81 16 0.84 0.91 12 2 111 94 17 0.85 0.92 13 3 115 42 73 0.37 0.71 14 1 47 45 2 0.96 0.96 Average ------0.68 0.75

Guadalupe Fescue Species Status Assessment – Aug 2016

Figure B3. Guadalupe fescue at Big Bend National Park. Adjusted survival and adjusted precipitation.

50 1.00

40 0.80

30 0.60

20 0.40 Survival Rate 10 0.20

Adjusted Precipitation, cm 0 0.00 1994 1996 1997 1998 2000 2001 2002 2003 2005 2006 2008 2010 2013 2014 Year

AdjPpt-cm AdjSurv

Estimates of apparent lifespan (based on Fowler 2015a, pp. 1-3, Fowler 2015c).

A total of 526 individual Guadalupe fescue plants were identified during this study. However, using this data to determine their lifespans requires some interpretation, for several reasons: 1) Plants that were alive at the beginning or end of the survey period may have longer lifespans than observed; 2) observations were made no more often than once per year, so a plant observed in a single census may have lived anywhere from 1 to 23 months; and 3) the gaps in census years reduce the precision of lifespan measurements. Therefore, the term “apparent” lifespan is more accurate. Not all plants could be used for lifespan analysis. Sixty-two “phoenix” plants were observed again after an absence of one or more years. Since it was not possible to determine whether these were merely dormant during one or more censuses, or if a new plant grew where an earlier plant had died, these were excluded from lifespan analysis. Finally, due to the interruption in data collection from Plot 5 (described above), no plants from Plots 5 or 5a contributed to the lifespan analysis. For these reasons, the estimate of apparent lifespan includes data from 241 individuals. In addition to this pooled data set, we also estimated lifespan from 3 subset groups for comparison: • Group A – 171 plants that were first observed after 1993 and last observed before 2014; this group yields the most conservative estimate, since it only includes entire lifespans, but excludes long-lived plants. We also based estimates on Groups B and C to demonstrate how including those groups affects lifespan estimates. • Group B – 53 plants that were present in 1993 and died before 2014. • Group C – 17 plants that were not present in 1993 but present in 2014. Note that 1individual lived through the entire 22 years of the study. Average lifespans of these groups ranged from 3.1 to 7.4 years; the pooled group had an average lifespan of 3.6 years. Table B3 summarizes the numbers of individuals by years of lifespan, and Figure B4 charts the percent of individuals by apparent lifespan.

Lifespan can also be estimated from the survival rates. Based on the adjusted annual survival rate (0.75), plotted in Figure B4 as “lifespan based on survival”, the average lifespan is 3.9 years.

Guadalupe Fescue Species Status Assessment – Aug 2016

Table B3. Numbers of Guadalupe fescue plants and their apparent lifespans. See explanation of groups in text.

------Apparent Lifespans in Years ------

Average Group 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Total Lifespan

A 59 57 15 1 15 2 2 6 0 4 5 2 2 0 0 0 1 171 3.1 B 3 0 29 6 1 0 13 0 1 0 0 0 0 0 0 0 0 53 4.1 C 4 0 0 2 0 1 0 2 3 0 1 1 1 2 0 0 0 17 7.4 Pooled Ind. of No. 66 57 44 9 16 3 15 8 4 4 6 3 3 2 0 0 1 241 3.6

Figure B4. Guadalupe fescue at Big Bend National Park: estimates of apparent lifespan.

% of Individuals 20

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Apparent Lifespan (years) A B C Pooled Lifespan based on survival

GLIMMIX (generalized linear mixed models) of the effects of precipitation, year, and panicle number on survival rate.

Fowler (2015c, d) used the SAS GLIMMIX procedure to fit generalized linear mixed models to estimate the effects of precipitation, time, and panicle number on the probability of an individual surviving. Each observation was one plant in one survey interval. Phoenix plants were dropped from the data set before analysis due to the reasons described above. The 1998-2000 survey

Guadalupe Fescue Species Status Assessment – Aug 2016 interval data for plot 5 and the 2001-2002 survey interval data for plot 5A were also dropped prior to analysis, due to the plot relocation issue described above. This left 970 observations.

The overall survival rate of this data set was 62%. When plots are given equal weight, the model predicts an average survival rate of 64%.

The model reveals that, regardless of the precise variables selected, both time and precipitation have significant, positive relationships with survival rate. Figure B5 shows the relationships of survival to calendar year and adjusted precipitation. This is a more powerful and more accurate way to determine the relationships between survival, time, and precipitation than Spearman correlation coefficients (Fowler 2015d, p.1).

Guadalupe Fescue Species Status Assessment – Aug 2016

Figure B5. Relationship of survival rate to time and precipitation (Fowler 2015d).

Guadalupe Fescue Species Status Assessment – Aug 2016

Reproductive output.

The numbers of panicles and spikelets (Figures 3.3 and 3.3a) serve as surrogates for reproductive output. During the years of observation, an average of 34.9 plants per year reproduced at this site. The reproductive Guadalupe fescue plants produced an average of 18.1 spikelets per panicle (standard deviation 5.1), and the entire monitored population averaged 1,662 spikelets per year. Spikelets typically have 2 to 3 florets (Figure 3.3a), and florets may bear a single seed. Hence, on average, these plants could produce a theoretical maximum of 46 seeds per panicle, and the theoretical maximum average number of seeds is nearly 5,000 per year. The total annual reproductive output from all plots is summarized in Table B4. However, among grasses in general, the proportion of florets that produce viable seeds can be very low, particularly among species that are obligate out-crossers. Although the breeding system of Guadalupe fescue is currently unknown, some Festuca species are obligate out-crossers (Pedersen and Sleper 1993, p. 187; Fryxell 1957 p. 180). The numbers of viable seeds produced by this population has not been determined, as the collection of seeds might have interfered with reproduction at this site. Like all grasses, Guadalupe fescue is wind-pollinated. Considering the low population density and relatively small numbers of florets in this population that would be shedding or receptive to pollen at any one time, the probability of out-crossing is very low. Fortunately, seeds collected from this population have been successfully germinated at Desert Botanical Garden (Pheonix, AZ), so we know that it produces at least some viable seeds. If Guadalupe fescue is able to self- fertilize, it is likely that the remaining population is highly inbred and could be suffering from inbreeding depression (see Section II. 4).

Table B4. Annual reproductive output of Guadalupe fescue at Big Bend National Park, 1993 to 2014.

Measure Minimum Maximum Average No. of reproductive plants 0 99 34.9 % of reproductive plants 0 85.4 36.2 No. of panicles 0 352 97.5 No. of spikelets 0 6,567 1,662a Theoretical maximum no. of seeds 0 19,701 4,986 a. Average spikelets per year is less than average panicles x average spikelets per panicle because there were no spikelets in 2006.

Zimmerman and Moir (1998, p. 11–12) postulated that, since Guadalupe fescue is a cool-season grass, February through May rainfall should be critical for high panicle and seed production. Surprisingly, panicle production was not correlated with February – May rainfall months (ρ = 0.347); it is, however, positively correlated with precipitation over the previous 21 months (ρ = 0.686, p = 0.005).

Guadalupe Fescue Species Status Assessment – Aug 2016

Figure B6. Guadalupe fescue at Big Bend National Park: number of panicles and annual precipitation.

70.0 400 350

60.0

50.0 300 250 40.0 200 30.0 150

20.0 100 of No. Panicles Precipitation (cm) Precipitation 10.0 50 0.0 0

Year

Total Precipitation (cm) Panicles

Fowler (2015d, p. 2) also tested whether a plant's reproductive status at the beginning of a census interval affected its survival. Whether or not a plant reproduced had a statistically significant effect when this term was added to the GLIMMIX model described above. When all plants are pooled, the survival rates are 52% when reproductive and 67% when not reproductive. When plots are given equal weight, plants that reproduced had a 61% chance of surviving and plants that did not reproduce had a 71% change of surviving.

Considering the entire census of 526 plants, 59.3 percent reproduced one or more times, and 40.7 percent never reproduced. Simplifying lifespan to number of years observed alive during the study, plants that reproduced were alive an average of 3.33 years, and plants that did not reproduce were alive an average of 1.60 years. Figure B7 shows the percent of plants that reproduce by number of reproductive years.

Guadalupe Fescue Species Status Assessment – Aug 2016

Figure B7. Guadalupe fescue at Big Bend National Park: number of years that individuals reproduce.

45.0 100.0 40.0 90.0

35.0 80.0 30.0 70.0 60.0 25.0 50.0 20.0 40.0 15.0 30.0

10.0 Percent Cumulative Percent of Individuals 20.0 5.0 10.0 0.0 0.0 0 1 2 3 4 5 6 7 8 9 10 ReproductiveYears

Percent of Individuals Cumulative Percent

Summary of monitoring plot analyses at Big Bend National Park:

A total of 526 Guadalupe fescue plants were monitored for 15 years over a 22-year period on 6 permanent plots. The plots are distributed within a population that is scattered at low density over an area of about 5 ha. The size of the monitored population has declined over the period of study, and is likely to be positively correlated with rainfall over longer intervals (33 months). Estimated average annual survival rates ranged from 0.62 to 0.74, and average lifespans ranged from 3.1 years to 3.9 years, depending on the method used and assumptions about which plants to exclude from analysis. A generalized linear mixed model developed with the SAS GLIMMIX procedure estimated average survival ranged from 0.62 to 0.64, and found that average survival has a significant positive relationship with precipitation and with time; in other words, average survival rates have increased over the period of study, and also increase with increasing precipitation.

The total annual production of florets ranged from 0 to over 19,000 and averaged over 4,000 per year. However, the amount of seed set (the proportion of florets that contain a viable seed) has never been determined. Even if seed set is high, the potential amount of seeds produced by the population is relatively low compared to typical viable populations of perennial grasses. The breeding system of Guadalupe fescue is also unknown; if it is an obligate out-crosser, considering the low density of plants and wind pollination, the amount of seed set would likely be very low. Individual plants that reproduce represent 59.3 percent of the total observed, and were observed in twice as many years as plants that did not reproduce (3.33 versus 1.60 years). However, only 4 percent of plants reproduced in more than 3 years. Plants that reproduced in a given year had a lower survival rate (0.61) than plants that did not (0.71). This may be due to the plant consuming large amounts of its stored nutrients and water to produce reproductive structures and mature, viable seeds.

Guadalupe Fescue Species Status Assessment – Aug 2016

These data and analyses indicate that Guadalupe fescue is a short-lived perennial with relatively low fecundity. Survival rates and reproductive output (estimated through panicle production) are positively, significantly correlated with precipitation, and survival rates have increased significantly over time; conversely, the population size has decreased significantly over time. The inverse trends in increasing survival rates and decreasing population size over time may be explained by a recruitment rate that is too low to sustain the population. Low recruitment could be due to low fertilization and seed set and/or low establishment rates; the latter may result from poor germination and/or high seedling mortality. These in turn could be caused by lack of genetic fitness, habitat factors, such as the amount of light reaching the soil surface and depth of detritus layer, changes in temperature and rainfall, pathogens, or parasites, among other possible explanations.

Consequently, these results identify several priority research needs:

• Identify the breeding system; • determine the amount of seed set; • study effects of habitat parameters, temperature, and rainfall on germination and establishment; • search for evidence of pathogens and parasites.

Guadalupe Fescue Species Status Assessment – Aug 2016

Appendix C. Cause and Effects Evaluations for Guadalupe Fescue (Analysis Timeframe: 2015 to 2046). Table C1. Confidence terminology. This table explains our characterization of confidence levels in the following cause and effects tables.

Confidence Terminology Explanation We are more than 90% sure that this relationship or assumption accurately reflects the reality in the wild as Highly Confident supported by documented accounts or research and/or strongly consistent with accepted conservation biology principles.

We are 70 to 90% sure that this relationship or assumption accurately reflects the reality in the wild as Moderately Confident supported by some available information and/or consistent with accepted conservation biology principles.

We are 50 to 70% sure that this relationship or assumption accurately reflects the reality in the wild as Somewhat Confident supported by some available information and/or consistent with accepted conservation biology principles.

We are less than 50% sure that this relationship or assumption accurately reflects the reality in the wild, as Low Confidence there is little or no supporting available information and/or uncertainty consistency with accepted conservation biology principles. Indicates areas of high uncertainty.

Guadalupe Fescue Species Status Assessment – Aug 2016

Table C2. Demographic and genetic consequences of small population size.

Factor: Demographic and genetic consequences of small population size.

[ESA Factor(s): E] Analysis Confidence / Uncertainty Supporting Information SOURCE(S) of the actions The source of demographic and genetic risks is a result of Highly confident Barrett and Kohn 1991, causing the stressor small population sizes and a low density of individuals in pp. 3-30 populations.

- Activity(ies) occurring on the Historic: Populations were limited to "sky island" habitats Moderately confident that Barret and Kohn 1991, ground as a result of stressor and were historically isolated or had rare genetic exchange historically, isolated sky island pp. 3-30; Big Bend among the different "sky island" populations. However, the populations rarely experienced National Park and Service size of the populations and the density of individuals in "sky genetic exchange. 2008; Sirotnak 2015; island" populations was likely greater and there were likely Valdés-Reyna, J. 2009. metapopulations within the "sky island" populations with genetic exchange.

Current: There are only two known extant populations. The Highly confident that estimated size of the Boot Canyon population has varied populations are very isolated from 94 to 254 individuals between 1993 and 2014 and currently and are likely to probably has very low genetic diversity. The only other remain so. known extant population is in the Sierra Maderas del Carmen in Mexico, which in 2003 had 3 colonies with several hundred individuals; this population is likely to have greater genetic diversity than the Boot Canyon population. The 2 extant populations are isolated from each other and genetic exchange is unlikely between them.

Future: Without active management, the Boot Canyon population is likely to remain small or to decline, based on data collected from 1993 through 2014. The Sierra Maderas del Carmen population remains extant but it size and population trends are unknown. However, both populations are likely to remain isolated.

Guadalupe Fescue Species Status Assessment – Aug 2016

STRESSOR(S) (environmental A. Small populations are at greater risk of extirpation from A. Highly confident. Barrett and Kohn 1991, changes affecting the resources stochastic events; the Boot Canyon population is likely at pp. 3-30; Edmands 2007. needed by the species)(life history high risk, the Maderas del Carmen population at moderate and habitat needs) risk.

B. Low population density reduces out-crossing between B. Highly confident. unrelated individuals in wind-pollinated species, resulting in low rates of seed set and greater inbreeding. Out-crossing is likely to be insufficient at Boot Canyon, but may be sufficient in the larger, denser Maderas del Carmen population.

C. Small populations are more likely to have low genetic C. Highly confident. diversity, to incur inbreeding, and to incur genetic drift, than larger populations. The Boot Canyon population is likely to have low genetic diversity and to incur Inbreeding. Both populations are likely to have incurred genetic drift.

D. Inbreeding can cause inbreeding depression (reduced D. Low confidence that vigor and reproductive fitness), but we do not know if inbreeding depression affects Guadalupe fescue is susceptible to inbreeding depression. either population.

E. Low population genetic diversity reduces the resistance E. Moderately confident. to pathogens, adapt to changing environmental conditions, and ability to colonize new habitats. The Boot Canyon population is likely to be affected by low genetic diversity.

F. Populations that survive a genetic bottleneck may benefit F. High confidence that genetic through the elimination of harmful alleles. bottlenecks can benefit populations, but we do not know if this is true for Guadalupe fescue populations.

G. Genetic drift may be detrimental, neutral, or beneficial to G. High confidence; we do not a population. know what effect genetic drift has had on Guadalupe fescue populations.

Guadalupe Fescue Species Status Assessment – Aug 2016

- Affected Resource(s) needed by Genetic diversity of populations, viable population sizes, Highly confident the species (life history and and population density. habitat needs)

- Exposure of Stressor(s) to Occurs throughout the extant populations. The Boot Highly confident Big Bend National Park resources needed by the species Canyon population has ranged from 94 to 254 individuals and Service 2008; (When and where does the over the last two decades and is isolated from the only other Sirotnak 2015; Valdés- stressor overlap with the resource known extant population in Mexico; therefore, Boot Canyon Reyna, J. 2009. need of the species (life history is very exposed to demographic stressors. The Sierra and habitat needs)) Maderas del Carmen population in Mexico has 3 colonies with several hundred individuals as of 2003, and is somewhat less exposed to demographic stressors.

- Immediacy of Stressor(s) to the Historic: We assume that historically populations were Historic: Moderately confident resources needed by the species larger and less vulnerable to this stressor. (what is the timing and frequency of stressor) Current: Demographic stressors continually affect small Current and Future: Highly populations, but are most critical to the species confident that small populations reproduction. Guadalupe fescue usually flowers from late are currently exposed and will August to early October. continue to be exposed to demographic and genetic Future: Small populations will continue to be exposed to the stressors. genetic effects of small population sizes in the future.

Changes in Resource(s) needed by A. The Boot Canyon population is likely to be much A. Highly confident. the species smaller than an estimated minimum viable population size.

B. Low density of the Boot Canyon population means B. Highly confident. reproduction is likely to be largely via self-fertilization. This population likely has very little genetic diversity, with many inbred individuals.

C. Individuals at Boot Canyon may suffer from inbreeding C. Low confidence. depression.

Guadalupe Fescue Species Status Assessment – Aug 2016

Response to Stressors: A. Individuals that resulted from self-fertilization or A. Highly confident. - INDIVIDUALS through out-crossing between closely-related individuals are likely to be inbred.

B. Inbred individuals may suffer from inbreeding B. Low confidence. depression (reduced vigor and reproductive fitness).

C. Low population density reduces probability of C. Highly confident. reproduction through fertilization between unrelated individuals.

POPULATION & SPECIES RESPONSES

Effects of Stressors: A. Small populations are at greater risk from extirpation A. Highly confident. Barrett and Kohn 1991, - POPULATIONS from stochastic factors than are larger populations; the Boot pp. 3-30 [RESILIENCY] Canyon population has a high risk of extirpation from these stressors, and Maderas del Carmen has a moderate risk. What are the effects on population characteristics B. Small populations are also more likely to have lower B. Moderately confident that genetic diversity, which can reduce the ability to resist Boot Canyon is affected by low pathogens and parasites, adapt to environmental changes, genetic diversity. and colonize new habitats; the Boot Canyon population is likely to be affected by low genetic diversity.

C. Small, isolated populations can incur lower fitness C. Low confidence that either through the accumulation of deleterious mutations and population is affected by inbreeding depression; we do not know if either population inbreeding depression. is affected by inbreeding depression or the accumulation of deleterious alleles.

Guadalupe Fescue Species Status Assessment – Aug 2016

- SCOPE (what proportion of the Historic: Sky Island populations were separated by low- Historic: Moderately confident rangewide populations does this elevation Chihuahuan Desert, so genetic exchange from one stressor affect?) "sky island" population to another was possible but highly unlikely. However, the size of historic populations was likely larger and did not experience this stressor.

Current and Future: Based on data collected over a 21-year Current and Future: Highly span, the Boot Canyon population is currently small and is confident that the Boot Canyon decreasing over time and is affected by demographic population is affected by small stressors. This stressor is likely affecting the Maderas del effective population size. Carmen population, though to a lesser degree. Moderately confident that Maderas del Carmen is affected by small population size.

Effects of Stressors: A. The Boot Canyon population is declining and may A. Moderately confident. Fowler 2015a. - SPECIES (Rangwide) become extirpated due to this stressor during the analysis [REDUNDANCY] timeframe.

What are the expected future B. Since there are only two known extant populations, the B. Highly confident. changes to the number of loss of any population will significantly reduce overall populations and their distribution redundancy and will increase the risk of extinction. across the species' range?

Effects of Stressors: Any future loss of populations or the reduction in the Highly confident - SPECIES (Rangwide) resiliency of populations will continue to reduce overall [REPRESENTATION] genetic and ecological diversity of the species, further limiting the species' representation. What changes to the genetic or ecology diversity in the species might occur as a result of any lost populations?

Guadalupe Fescue Species Status Assessment – Aug 2016

RISK OF EXTIRPATION 2046 A. The Boot Canyon population has a moderate risk of A. Moderately confident. extirpation due to demographic and genetic stressors by 2046.

B. The Maderas del Carmen population has a low risk of B. Moderately confident. extirpation by 2046.

Table C3. Altered wildfire frequency.

Factor: Altered Wildlife Frequency

[ESA Factor(s): A, E] Analysis Confidence / Uncertainty Supporting Information SOURCE(S) of the actions Fire suppression at Big Bend National Park since 1944 has Highly confident Big Bend National Park causing the stressor altered the fire frequency; there have been no wildfires in and Service 2008; Camp Boot Canyon since 1944, leading to increased density of et al. 2006; Moir and trees and shrubs, increased canopy cover, and accumulation Meents 1981; of litter layer and fuel. Wildfire has not occurred at Maderas Zimmerman and Moir del Carmen since 1953. 1998.

- Activity(ies) occurring on the Increased density of trees and shrubs, increased canopy Highly confident ground as a result of stressor cover, accumulation of litter layer and fuel.

STRESSOR(S) (environmental A. Increased competition from shrubs and small trees. A. Highly confident. changes affecting the resources needed by the species)(life history B. Decreased light levels at soil surface. B. Highly confident. and habitat needs) C. Increase in surface litter depth may reduce vigor and C. Somewhat confident. reproduction of Guadalupe fescue.

D. Increased fuel loads and tree density contribute increase D. Highly confident. the risk of hotter, potentially catastrophic fires which could cause tree stand replacement (rather than thinning) and could sterilize the soil seed bank.

Guadalupe Fescue Species Status Assessment – Aug 2016

- Affected Resource(s) needed by Conifer-oak woodland habitat; appropriate light levels; Highly confident the species (life history and tolerable level of competition for soil, nutrients, water, and habitat needs) light; viable soil seed bank.

- Exposure of Stressor(s) to Exposure is greatest where there has been effective fire Highly confident resources needed by the species suppression (Big Bend National Park). Exposure increases (When and where does the (effects are more severe) with increasing time since last stressor overlap with the resource burn. The risk of catastrophic wildfire is greatest during the need of the species (life history summer and during hot, dry years. and habitat needs))

- Immediacy of Stressor(s) to the Past: Tree ring data confirms that periodic low-intensity Past: Highly confident. Big Bend National Park resources needed by the species wildfires occurred in the Chisos Mountains and Maderas del and Service 2008; Camp (what is the timing and frequency Carmen and maintained necessary habitat conditions. We et al. 2006. of stressor) assume that regular wildfire cycles occurred in most if not all montane “sky island” conifer-oak forests of the Chihuahuan Desert. Frequent low intensity fires reduced canopy shading and competition from woody plants, and prevented fuel loads from accumulating and creating conditions for an intense wildfire.

Current A: No wildfires have occurred at Boot Canyon Current A: Highly confident. since 1944 or at Maderas del Carment since 1953, leading to altered habitat conditions. The Big Bend National Park fire management plan allows a wildfire to move through the Boot Canyon site if the fire intensity is low; however, currently there are no plans to conduct prescribed burning in Boot Canyon. Hence, this stressor currently affects Guadalupe fescue and will increase over time.

Current B: In northern Mexico, other potential Guadalupe Current B: Moderately fescue habitats are remote, rugged, and difficult to access, confident. and federal, state, and local agencies have fewer resources and less training for wildfire suppressions; wildfires are likely to continue as they have in the past, so it is likely that many potential habitats are not affected by this stressor.

Guadalupe Fescue Species Status Assessment – Aug 2016

Future: The NPS management is willing to consider Future: moderately confident. conducting a prescribed burn in Boot Canyon if funds are available. In Mexico, wildfires are likely to continue as they have in the past.

Changes in Resource(s) needed by Past: Periodic wildfires 1) limited tree canopy cover, Past: Highly confident. the species allowing more light to reach the Guadalupe fescue plants; 2) reduced competition from woody plants; 3) prevented accumulation of leaf litter and reduced the fuel load throughout the species range in Texas and northern Mexico.

Current A: Fire suppression in the Chisos Mountains and Current A: Highly confident. lack of fire at Maderas del Carmen has increased competition from shrubs and small trees, decreased light levels at soil surface, and increased surface litter depth. Increased fuel loads and tree density contribute to a risk of hotter, potentially catastrophic fires which could cause tree stand replacement (rather than thinning) and could sterilize the soil seed bank.

Current B: At other potential habitats in Mexico, due to the Current B: Moderately continued frequency of low intensity fires, fuel loads are confident. likely reduced and the risk of catastrophic fire is lower.

Future: Habitat conditions are likely to decline in the Future: moderately confident. Chisos and Maderas del Carmen ranges unless low-intensity fires occur there. The risk of a catastrophic fire is likely to increase over time if there are no fires.

Response to Stressors: Low vigor and reduced reproductive output due to low light Moderately confident. - INDIVIDUALS levels and competition from woody plants; increased risk of mortality from catastrophic wildfire.

POPULATION & SPECIES RESPONSES

Guadalupe Fescue Species Status Assessment – Aug 2016

Effects of Stressors: A: Where wildfire cycle is disrupted, such as Boot Canyon A: Moderately confident. - POPULATIONS and Maderas del Carmen, populations are likely to decline [RESILIENCY] due to low light levels, competition from woody plants, and deep leaf litter accumulation; the risk of extirpation due to What are the effects on catastrophic wildfire increases as fuel load increases. population characteristics B: Other Mexican populations may have experienced more B. Somewhat confident. frequent wildfires and therefore would be less affected by this stressor.

- SCOPE A: Fire suppression has been conducted at Big Bend A: Highly confident. (what proportion of the National Park, and no fires have occurred at Boot Canyon rangewide populations does this since 1944. Wildfires have not occurred in Maderas del stressor affect?) Carmen since 1953.

B: Periodic, low intensity fires may occur in other areas of B. Somewhat confident. the species range in Mexico.

Effects of Stressors: This stressor could contribute to the extirpation of the Boot Highly confident. - SPECIES (Rangwide) Canyon population, and may contribute to the decline of the [REDUNDANCY] Maderas del Carmen population. Loss of either of the two known extant populations would severely reduce the species' What are the expected future redundancy. changes to the number of populations and their distribution across the species' range?

Effects of Stressors: Loss of either of the two known extant populations would Highly confident. - SPECIES (Rangwide) severely reduce the species' genetic and ecotypic diversity. [REPRESENTATION]

What changes to the genetic or ecology diversity in the species might occur as a result of any lost populations?

Guadalupe Fescue Species Status Assessment – Aug 2016

RISK OF EXTIRPATION 2046 A. Absence of low-intensity wildfire increases the risk of A. Highly confident. extirpation of the Boot Canyon population during the analysis timeframe, due to increased shade, competition from woody plants, and leaf litter accumulation, as well as the increased risk of a very hot fire with catastrophic impacts to vegetation.

Table C4. Livestock grazing.

Factor: Livestock Grazing

[ESA Factor(s): A, C] Analysis Confidence / Uncertainty Supporting Information SOURCE(S) of the actions Domestic and feral livestock. Highly confident Poole 1989, p. 13; Big causing the stressor Bend National Park and the Service 2008.

- Activity(ies) occurring on the Grazing (consumption of live herbaceous plant tissues by Highly confident ground as a result of stressor herbivores).

STRESSOR(S) (environmental Sudden removal of all or a portion of a plant's above-ground Highly confident changes affecting the resources tissues. needed by the species)(life history and habitat needs)

- Affected Resource(s) needed by Photosynthetic tissues that sustain growth and reproduction, Highly confident the species (life history and reproductive tissues (flower stalks and seeds), and minerals habitat needs) contained in these tissues.

- Exposure of Stressor(s) to A: Livestock grazing is prohibited at Big Bend National A. Highly confident. resources needed by the species Park, including the Chisos Mountains. Feral animals (When and where does the currently occur within Big Bend NP but have not been stressor overlap with the resource documented in the Chisos Mountains. Livestock have been need of the species (life history removed from the Mexican population at Maderas del and habitat needs)) Carmen.

Guadalupe Fescue Species Status Assessment – Aug 2016

B. If other Guadalupe fescue populations remain in northern B. Moderately confident. Mexico, it is likely that they are exposed to heavy livestock grazing.

C. Effects of grazing on Guadalupe fescue depend on timing: C. Highly confident. Winter grazing (after seed dispersal and before re-growth in spring) probably has little if any effect on survival and reproduction; spring and early summer grazing probably reduces growth, survival, and reproduction; late summer - early fall grazing is most severe, because it prevents reproduction.

- Immediacy of Stressor(s) to the Past: Livestock grazing occurred prior to 1944 in Big Bend Highly confident resources needed by the species National Park. Livestock were grazed until recently at (what is the timing and frequency Maderas del Carmen. of stressor) Current and Future: Livestock are no longer grazed in the Chisos Mountains and Maderas del Carmen sites. If other Guadalupe fescue populations remain in northern Mexico, it is likely that they are continuously or periodically exposed to heavy livestock grazing. Although feral animals could stray into Big Bend National Park, It is unlikely that they would remain at Boot Canyon long enough to effect Guadalupe fescue. The incidence and effects of feral animals at Maderas del Carmen are unknown.

Changes in Resource(s) needed by Livestock grazing removes tissues (leaves and flower stalks) Highly confident the species needed for growth and reproduction of Guadalupe fescue. Effects of grazing increase with increasing frequency and duration of grazing and density of grazing animals. Grazing during the late summer and early fall, when plants are flowering and before seeds mature, has the most severe effect, and grazing during the dormant season (winter) would have the least effect.

Guadalupe Fescue Species Status Assessment – Aug 2016

Response to Stressors: Depends on timing of impact in terms of the plants Highly confident - INDIVIDUALS phenology. Winter grazing (after seed dispersal and before re-growth in spring) probably has little if any effect on survival and reproduction; spring and early summer grazing probably reduces growth, survival, and reproduction; late summer - early fall grazing is most severe, because it prevents reproduction.

POPULATION & SPECIES RESPONSES

Effects of Stressors: The loss of even a few individuals from small populations Highly confident - POPULATIONS could reduce the population size and genetic diversity below [RESILIENCY] the levels necessary for long-term survival.

What are the effects on population characteristics

- SCOPE This could occur throughout the range. Highly confident (what proportion of the rangewide populations does this stressor affect?)

Effects of Stressors: - SPECIES (Rangwide) [REDUNDANCY]

What are the expected future changes to the number of populations and their distribution across the species' range?

Guadalupe Fescue Species Status Assessment – Aug 2016

Effects of Stressors: This stressor could contribute to the extirpation of either of Highly confident - SPECIES (Rangwide) the two known extant populations, and would severely reduce [REPRESENTATION] the species' redundancy.

What changes to the genetic or ecology diversity in the species might occur as a result of any lost populations?

RISK OF EXTIRPATION 2046 The risk of extirpation of the Boot Canyon and Maderas del Highly confident Carmen populations is low, since these sites are protected from livestock grazing.

Table C5. Recreation.

Factor: Recreation

[ESA Factor(s): A, C, E] Analysis Confidence / Uncertainty Supporting Information SOURCE(S) of the actions Hiking trail from Pinnacles to Boot Spring at Big Bend Highly confident. Big Bend National Park causing the stressor National Park bisects the Boot Canyon population. and Service 2008, p. 5

- Activity(ies) occurring on the Walking, trail maintenance, changes in surface flow of Highly confident. Big Bend National Park ground as a result of stressor rainwater. and Service 2008, p. 5

STRESSOR(S) (environmental Trampling by hikers straying from the trail, trampling or Highly confident. changes affecting the resources grazing by pack animals used by Big Bend National Park trail needed by the species)(life history maintenance personnel, runoff diverted by trail structures, and habitat needs) and erosion and debris flows caused by trail runoff during heavy rainfall. - Affected Resource(s) needed by Loss of individual plants due to trampling, grazing, soil Highly confident. the species (life history and erosion, or debris flows; decreased reproduction due to loss habitat needs) of seed production or fertilization.

Guadalupe Fescue Species Status Assessment – Aug 2016

- Exposure of Stressor(s) to A. About 600 m (1,968 ft) of the Boot Canyon Trail A. Highly confident. Arnberger 2015. resources needed by the species intersects the Boot Canyon population. Big Bend National (When and where does the Park maintenance personnel utilize this trail with pack stressor overlap with the resource animals (horses or mules) for maintenance and monitoring. need of the species (life history Big Bend National Park has issued 660 camping permits for and habitat needs)) 1,610 people in Boot Canyon in 2015; total estimated trail use is about 3,200 visitors per year. The incidence of damage off-trail currently appears to be minimal.

B. Exposure to this stressor at Maderas del Carmen is B. Moderately confident. unknown but is likely to be insignificant.

- Immediacy of Stressor(s) to the Historic: In 2005, debris flows below trail switchbacks Highly confident. Big Bend National Park resources needed by the species buried some of the monitored Guadalupe fescue plants up to and Service 2008, p. 5 (what is the timing and frequency 20 cm (8 in.) deep; trail runoff may or may not have caused of stressor) the debris flow.

Current and Future: Hikers are required to stay on the trail. Park personnel are educated about the Guadalupe fescue and are instructed to not trample or let pack animals graze or trample the species.

Changes in Resource(s) needed by Historically some individuals were lost due debris flow. Highly confident the species Currently, this threat is managed.

Response to Stressors: Injury, loss of vigor, reduced reproductive output, and Highly confident. - INDIVIDUALS mortality.

POPULATION & SPECIES RESPONSES

Effects of Stressors: If these actions are not managed then the resiliency of the Highly confident. - POPULATIONS population will decline but currently and in the future will be [RESILIENCY] managed.

What are the effects on population characteristics

Guadalupe Fescue Species Status Assessment – Aug 2016

- SCOPE The recreation stressor affects only the Boot Canyon Highly confident. (what proportion of the population. rangewide populations does this stressor affect?)

Effects of Stressors: Boot Canyon is one of only two known extant populations; Highly confident. - SPECIES (Rangwide) loss of this population would severely reduce redundancy of [REDUNDANCY] the species. However, this stressor is currently well managed. What are the expected future changes to the number of populations and their distribution across the species' range?

Effects of Stressors: Boot Canyon is one of only two known extant populations; Highly confident. - SPECIES (Rangwide) loss of this population would severely reduce representation [REPRESENTATION] of the species. However, this stressor is currently well managed. What changes to the genetic or ecology diversity in the species might occur as a result of any lost populations?

RISK OF EXTIRPATION 2046 The Boot Canyon population is at low risk of extirpation due Highly confident. to recreational uses because these are currently managed by Big Bend National Park and will continue to be managed in the future.

Guadalupe Fescue Species Status Assessment – Aug 2016

Table C6. Invasive species.

Factor: Invasive Species

[ESA Factor(s): A, C, E] Analysis Confidence / Uncertainty Supporting Information SOURCE(S) of the actions Introduced invasive plant species: King Ranch bluestem Highly confident. Presence in Big Bend National Park causing the stressor (Bothriochloa ischaemum) or other Old World bluestems, the region is well documented. and Service 2008, p. 5 Horehound (Marrubium vulgare), and potentially other introduced grasses such as Lehman lovegrass (Eragrostis lehmanniana).

- Activity(ies) occurring on the Historic: King Ranch bluestem is now abundant at site in Historic: Highly confident. Big Bend National Park ground as a result of stressor Guadalupe Mountains National Park where Guadalupe fescue and Service 2008, p. 5 formerly occurred and may have contributed to its extirpation. Horehound was present at Boot Canyon but was manually removed.

Current: King Ranch bluestem and Lehmann lovegrass are Current: Highly confident. abundant at some places in Big Bend National Park, but have not been found in Boot Canyon. Horehound is actively controlled in Boot Canyon and is not affecting Guadalupe fescue. We do not know if these or other invasive plants are present at the Maderas del Carmen site.

Future: King Ranch bluestem and Lehmann lovegrass could Future: Somewhat confident. colonize Boot Canyon; see Climate Change sheet.

Guadalupe Fescue Species Status Assessment – Aug 2016

STRESSOR(S) A. King Ranch bluestem and other Old World bluestems are A. Highly confident. Greer et al. 2014. (environmental changes affecting able to out-compete Guadalupe fescue if they invade habitats. the resources needed by the Competition is for water, nutrients, and light. species)(life history and habitat needs) B. King Ranch and other Old World bluestems may have B. Moderately confident. allelopathic affects.

C. Horehound and Lehmann lovegrass are currently not C. Low confidence. present in Guadalupe fescue habitats but could colonize sites and compete for resources.

- Affected Resource(s) needed by Survival of individual plants, reproductive output, and Highly confident the species (life history and recruitment. habitat needs)

- Exposure of Stressor(s) to Horehound was removed from Boot Canyon population but Highly confident Big Bend National Park resources needed by the species still occurs within the suitable habitat surrounding the and Service 2008, p. 5 (When and where does the population. King Ranch bluestem has not been found at Boot stressor overlap with the resource Canyon population but it does occur elsewhere in Big Bend need of the species (life history National Park. We do not know if invasive plants are present and habitat needs)) at Maderas del Carmen.

- Immediacy of Stressor(s) to the Current: King Ranch bluestem and Lehmann lovegrass are Highly confident. Big Bend National Park resources needed by the species highly invasive but have not been found at Boot Canyon. and Service 2008, p. 5 (what is the timing and frequency Horehound is naturalized in west Texas but is not highly of stressor) invasive; it has been removed from Boot Canyon.

Future: Big Bend National Park will continue to manage horehound, King Ranch bluestem, and Lehmann lovegrass to prevent encroachment into Boot Canyon.

Changes in Resource(s) needed by A. King Ranch bluestem and other Old World bluestems are A. Highly confident. Greer et al. 2014. the species able to out-compete Guadalupe fescue if they invade habitats. Competition is for water, nutrients, and light.

B. King Ranch and other Old World bluestems may have B. Moderately confident. allelopathic affects.

Guadalupe Fescue Species Status Assessment – Aug 2016

C. Horehound and Lehmann lovegrass are currently not C. Low confidence. present in Guadalupe fescue habitats but could colonize sites and compete for resources.

Response to Stressors: Individuals respond to competition from invasive plants Highly Confident - INDIVIDUALS through reduced vigor, reduced reproductive output, and death.

POPULATION & SPECIES RESPONSES

Effects of Stressors: Populations respond to this stressor by reduced recruitment Highly confident - POPULATIONS and population size (resilience), and extirpation. [RESILIENCY]

What are the effects on population characteristics

- SCOPE (what proportion of the A. King Ranch bluestem may have contributed to the A. Moderately confident. rangewide populations does this extirpation of Guadalupe fescue at Guadalupe Mountains stressor affect?) National Park.

B. The Boot Canyon population of Guadalupe fescue is not B. Highly confident. currently affected by these invasive plant species.

C. The Maderas del Carmen population may be affected by C. Low confidence. invasive plants, but we have no information in invasive plant presence there.

Guadalupe Fescue Species Status Assessment – Aug 2016

Effects of Stressors: A. Invasive plants may have contributed to the extirpation of A. Moderately confident. - SPECIES (Rangwide) the Guadalupe Mountains population, thus reducing the [REDUNDANCY] species redundancy.

What are the expected future B. The loss of either of the two remaining extant populations B. Highly confident. changes to the number of would further reduce redundancy of the species. populations and their distribution across the species' range?

What changes to the genetic or B. The loss of either of the two remaining extant populations B. Highly confident. ecology diversity in the species would further reduce representation of the species. might occur as a result of any lost populations?

RISK OF EXTIRPATION 2046 The Boot Canyon population is at low risk of extirpation due Highly confident to invasive plants because they are currently managed by Big Bend National Park and will continue to be managed in the future. We do not know if invasive plants are present at Maderas del Carmen or if personnel of this Protected Natural Area will be able to manage an invasive species infestation if it occurs.

Guadalupe Fescue Species Status Assessment – Aug 2016

Table C7. Climate change.

Factor: Invasive Species

[ESA Factor(s): A, E] Analysis Confidence / Uncertainty Supporting Information SOURCE(S) of the actions Combustion of fossil fuels; loss of carbon storage in Highly confident. IPCC 2013, entire. causing the stressor vegetation and soils; release of methane through petroleum development, livestock production, and melting permafrost.

- Activity(ies) occurring on the A. Increased ambient temperature. A. Highly confident. IPCC 2013, pp. 11, 16, 23; ground as a result of stressor Milly 2005, p. 347; USGS B. Increased incidence of heavy rainfall. B. Moderately confident. 2015.

C. Increased incidence of intense, prolonged drought. C. Moderately confident.

D. Decreased average annual precipitation. D. Low confidence.

STRESSOR(S) (environmental A. Increased evaporative deficit. A. Highly confident. IPCC 2013, entire; USGS changes affecting the resources 2015; Van Devender and needed by the species)(life history B. Increased erosion and debris flow from heavy rainfall. B. Moderately confident. Spaulding 1979. and habitat needs) C. Increased mortality and decreased recruitment from C. Moderately confident. drought. D. Increased risk of catastrophic wildfire during drought. D. Moderately confident.

E. Optimal habitat shifts to higher elevations in mountains, E. Moderately confident. thus decreased area and increased isolation.

- Affected Resource(s) needed by Ambient temperature, soil moisture, soil stability, suitable Highly confident. the species (life history and fire cycle. habitat needs)

Guadalupe Fescue Species Status Assessment – Aug 2016

- Exposure of Stressor(s) to These stressors will occur throughout the species range. Highly confident. resources needed by the species Evapotranspiration deficit will be greatest during the (When and where does the growing season. stressor overlap with the resource need of the species (life history and habitat needs))

- Immediacy of Stressor(s) to the A. Climate stressors may already affect the species (note A. Somewhat confident. resources needed by the species record high temperatures and low rainfall in 2011, and (what is the timing and frequency record high rainfall in 2015). of stressor) B. Climate change stressors will increase throughout the B. Highly confident. analysis timeframe.

C. Severity of stressors depends on climate scenario. C. Highly confident.

Changes in Resource(s) needed by A. Higher ambient temperatures. A. Highly confident. the species B. Increased evaporative deficit. B. Highly confident.

C. Increased soil erosion. C. Moderately confident.

D. Increased risk of catastrophic wildfire. D. Somewhat confident.

Response to Stressors: - A. Reduced vigor. A. Low confidence. INDIVIDUALS B. Shortened growing season. B. Low confidence.

C. Higher mortality. C. Low confidence.

D. Reduced reproductive output and recruitment. D. Low confidence.

POPULATION & SPECIES RESPONSES

Guadalupe Fescue Species Status Assessment – Aug 2016

Effects of Stressors: A. Increased mortality and/or reduced recruitment lead to A. Somewhat confident that B. Van Devender and - POPULATIONS lower population size and reduced resiliency. this climate changes reduces Spaulding 1979. [RESILIENCY] resiliency.

What are the effects on B. The species’ maximum optimal elevation range could B. Moderately confident that population characteristics increase but the potential habitat area diminishes and the optimal elevation range population isolation increases with increasing elevation in will increase due to climate mountain ranges. change.

C. Guadalupe fescue populations may not be able to migrate C. Moderately confident that as fast as the rate of climate change, and therefore lose habitat. Guadalupe fescue populations cannot migrate as fast as the rate of climate change.

D. Climate changes could extend the ranges of introduced D. Somewhat confident that invasive plant species, such as Lehmann lovegrass (Eragrostis climate changes could bring lehmanniana) and/or buffelgrass (Pennisetum ciliare,) into new invasive species into Guadalupe fescue habitats. Guadalupe fescue habitats.

E. The net interaction of multiple climate change effects their E. Highly confident in the synecological interactions on Guadalupe fescue populations is impossibility of predicting the extraordinarily complex and unpredictable. net results of synecological interactions induced by climate changes.

- SCOPE This stressor effects all extant populations. Highly confident. (what proportion of the rangewide populations does this stressor affect?)

Guadalupe Fescue Species Status Assessment – Aug 2016

Effects of Stressors: A. There are only two known extant populations. A. Highly confident. - SPECIES (Rangwide) [REDUNDANCY] B. The Boot Canyon population is small and declining. B. Highly confident.

What are the expected future C. Climate changes may contribute to or hasten the decline of C. Somewhat confident that changes to the number of this population within the analysis timeframe; its extirpation climate changes contribute to populations and their distribution would drastically reduce redundancy to a single known decline of the Boot Canyon across the species' range? population. population.

D. The Maderas del Carmen population, as recently as 2003, D. Highly confident that the was more resilient and therefore more likely to endure climate Maderas del Carmen changes longer. population is larger, more resilient, and likely to endure longer than Boot Canyon.

Effects of Stressors: If either of the two known extant populations is lost, a large Highly confident. - SPECIES (Rangwide) portion of the species ecotypic and genetic diversity would be [REPRESENTATION] lost.

What changes to the genetic or ecology diversity in the species might occur as a result of any lost populations?

RISK OF EXTIRPATION 2046 A. The Boot Canyon population has a moderate risk of A. Moderately confident. extirpation due to climate changes by 2046.

B. The Maderas del Carmen population has a low risk of B. Moderately confident. extirpation due to climate changes by 2046.