Spiranthes parksii Correll, the Navasota Ladies'-tresses, was listed on May 6, 1982, as an endangered species under the Endangered Species Act of 1973.  Critical habitat was not designated.  This species is listed as endangered by the State of Texas. The initial Recovery Plan was produced in 1984 (U. S. Fish and Wildlife Service, 1984).


    Spiranthes parksii was originally described as a distinct species by Donovan Correll (1947) from specimens collected by H. B. Parks in 1945.  Correll indicated that "This species....has no close allies in our flora.  Its affinity seems to be with several Mexican and Central American species.  The characteristically obovate petals and elliptic lip are distinctive and conveniently separate S. parksii from all other species."  Plants identified as S. parksii were not collected again until 1978 (Catling and McIntosh, 1979). This "rediscovery" of populations representing the species provided comparative material that was later used to gain a new perspective on both the distinction of this taxon as a species, as well as its relationship to other species of the genus from Texas.

    As indicated by Sheviak (1982), "Spiranthes are difficult taxonomic subjects."  Important characters, such as plant size, tend to show considerable variation depending of the vigor of the plant and variable associated with habitat.  Taxonomic difficulty is present at essentially all levels.  The genus has been circumscribed in the broad sense to include over 300 species with a world-wide distribution or, as is the case with most modern treatments, in the narrow sense to include "only a few dozen species, with a center of diversity in eastern North America" (Sheviak, 1982).  The eastern North American center of diversity includes a group of species, the "Spiranthes cernua" complex, that has a long history of taxonomic confusion and difficulty.  This group has received thorough and detailed study by Charles Sheviak, Curator of Botany, New York State Museum.  An overview of this group, including taxonomic review, character descriptions, cytology, breeding system, was published by Dr. Sheviak (1982).  This included a formal classification system, with line drawings and species distributions for all elements of the complex. This valuable work was accomplished prior to the 'rediscovery' of S. parksii.

    Dr. Sheviak traveled to Texas in the Fall of 1986, examined populations of S. parksii from both central and marginal areas of its range of distribution, and obtained samples for analysis in New York.  Examination of this material provided insight into the affinities of S. parksii (Sheviak, 1986):

I have found the chromosome number of S. parksii to be 2n = 60 and have verified peloria as the basis of the broad nature of the petals. Together with polyembryony (and hence adventitious embryony) and vegetative habit (tuberous roots and fugaceous leaves), these features clearly link S. parksii with S. cernua, populations of which, in the southern plains, exhibit this assortment of characteristics.
    Association of S. parksii with the S. cernua complex of species opened the question of its status as a distinct species, as opposed to placement within S. cernua as a subspecies or variety. Dr. Sheviak (1986) has indicated:
    The most distinctive features of S. parksii, (small flower size, lateral sepal shape, position, and proportions; and the open spiral of the inflorescence) must then be judged as to their significance. The small flower size and relative length, shape, and position of the lateral sepals are not predictable features of a segregate of S. cernua from the plains; these features do not seem to arise from the local magnicamporum-based S. cernua gene pool. Very likely a small-flowered species contributed genes to S. cernua and gave rise to this local entity. The question, then, is whether S. parksii is specifically distinct or merely an agamic race of repeatedly generated form of S. cernua.

    Meiosis in the collected material is quite regular, generally with strict bivalent formation or sometimes with a few quadrivalents. Hence, these plants may be capable of normal sexual reproduction. Yet there is no indication of the distinctive features of S. parksii being recombined in local populations of S. cernua. Most significantly, one plant was found that, on all morphological grounds, appeared to be perfectly intermediate between S. parksii and S. cernua. This apparent hybrid at once indicates that, 1) S. parksii is capable of sexual reproduction, 2) it is not merely a local chromosomally-based form of S. cernua (intermediates between such a form and others of S. cernua could not occur), and 3) despite the demonstrated possibility of hybridization, such hybridization is rare (one individual out of the six populations studied) and does not lead to a general incorporation of genes of S. parksii in the local S. cernua gene pool.

    Consequently, it appears that S. parksii is a product of the S. cernua complex but is not an interactive member. Its status is evidently similar to that of the northeastern S. casei, a species with a wide distribution that exhibits biogeographic details demonstrating a long history and supporting its recognition at the species level. Specific status for S. parksii is consequently entirely consistent with the evidence and with established precedent in the group.

    While this perspective contrasts with that provided by Dr. Correll in 1947 in terms of affinities to other species of the genus, it clearly places S. parksii as a biologically distinct species of the North American flora that is endemic to Texas.


    Orchid seeds, adapted for wind dispersal, are the smallest produced by any flowering plant. The seeds of Spiranthes parksii, and essentially all orchids, contain no nutritive tissue or endosperm. Germination and initial seedling growth under natural conditions is symbiotic in that this occurs only after the seed has been penetrated by fungal hyphae (Wells, 1981). The fungus provides nutrition for the growing embryo. The symbiotic association between orchid and fungus continues throughout the life of the plant.

    Following germination, the young Spiranthes plant remains underground as a small, colorless 'mycorhizome' that is fully dependent on the fungal symbiont for nutrition and growth. Time spent in the mycorhizome phase of growth varies among orchid species, but in 11 genera examined, it ranges from 1 to 15 years (Wells, 1981). The mycorhizome is eventually replaced in all orchids by a root tuber or similar perennating structure that is infected by the fungal associate and this, in turn, produces the first green leaf. While data for S. parksii are not available, the European S. spiralis mycorhizome exists for about 8 years prior to the root tuber stage. Another 3 years pass before the first green leaf is produced from the mycorhizome in S. spiralis. Thus, one can assume that observable plants within a given population of S. parksii represent only a fraction, possibly a small fraction, of plants that are actually growing at the site.

    Once photosynthesis begins, the annual cycle of growth and flowering in S. parksii is probably similar to that described for other members of the genus (Wells, 1981). Photosynthesis from 2 to 5 'rosette' leaves proceeds from late November to March or April. This produces a new root tuber which remains inactive underground until late October.  If photosynthate invested in the tuber is adequate, and environmental conditions supportive, a leafless flowering stalk is produced during the period between the fall rains and the first frost, usually from late October through November. The cycle of rosette leaf production is repeated during late fruit set in that new rosette leaves are often observed emerging at the base of the fruiting stalk in late November and December. These expand to produce the next group of rosette leaves which function to continue the cycle by producing a new root tuber during the photosynthetic period from November to late Spring.

    At flowering, S. parksii is an erect plant with the flowering stalk extending from 2 to 3 dm above the ground. The original description (Correll, 1947) notes that the inflorescence stalk is slender, glabrous below, glandular-pubescent above, with several tubular acuminate sheaths. The flowering spike is short, few-flowered, glandular-pubescent, from 3 to 7 cm long and ca. 1 cm wide with flowers in a single, spiral rank, ascending. Bracts subtending the flowers are ovate-lanceolate, acuminate, concave, 8-10 mm long and sometimes whitish at the tip. The sepals are pubescent on the outer surface, 3-nerved. The dorsal sepal is generally ovate-lanceolate, abruptly recurved at the acute-apiculate apex, deeply concave, about 6 mm. long and 2.8 mm wide below the middle. The lateral sepals are narrowly triangular-lanceolate, acuminate, oblique, with involute margins, about 7 mm. long and 2.2.5 mm. wide below the middle. The lateral petals are adherent to the dorsal sepal, oval to obovate or suborbicular, rounded and sometimes irregularly notched at the apex, with the anterior margin more or less erose, 5-nerved, scarcely oblique, about 5 mm. long and 2.5-3 mm. wide. The lowermost petal, or lip, is oval, broadly rounded or emarginate at the apex, minutely erose-laciniate on the upcurved margins, 5-5.5 mm. long, 3.8-4 mm. wide at about the middle; basal callosities stout, pubescent. The central reproductive structure, or column, is short, stout, and about 3.5 mm. long.


    Spiranthes parksii is closely associated with a narrow band of vegetation, the Post-Oak Savannah, that stands at the southwestern periphery of the deciduous forest formation of eastern North America. Positioned between prairies to the West and forest to the East this vegetation type, unique to Texas, includes a complex mixture of habitats and many rare, endemic plant species. Spiranthes parksii is adapted to life in the xeric, upland forest. It is rarely found in floodplain forests or openings or open areas dominated by tall grasses. Spiranthes parksii is not a 'colonizing' species of artificially disturbed sites. It is rarely found in old fields, power line right-of-ways, and other cleared sites and, when it is, it appears to be present as a remnant and not as a colonizer of new habitat. Natural disturbance is, however, a key ecological factor for this species.

    Within the context of its range of distribution, the Navasota River Valley, S. parksii is a high elevation species. With rare exceptions, the plant is found only above ca. 80 m. (250 feet) above sea level. Within the upland area, it is found, with few exceptions, along drainages that lead to the Navasota River. It is most abundant at the uppermost reaches of these drainages, at very distinctive, highly eroded sites that are now rare due artificial ponding. Populations extend along temporary stream margins downward from the eroded stream heads to the floodplain of permanent streams. Frequency of occurrence tends to diminish with elevation and increasing density of cover. Populations of S. parksii are rarely found beyond stream margins. They occupy stream banks and adjacent openings that lack cover due to natural disturbance, mostly slumping and tree fall resulting from meandering.

Associated Species

    Monitoring activity for this species has focused on a single "stream-bank" population along Sundew Creek at Lick Creek Park, just South of College Station and a series of populations in the vicinity of Carlos in Grimes County. Recent reports from both studies (Wilson, 1988; Parker, 1992) contain listings of associated plant species that are ranked by relative frequency.  Associated woody plants, comparable in both listings, reflect the Post Oak Savannah in terms of both species composition and relative frequency:

Ilex vomitoria Yaupon Holley
Quercus stellata Post Oak
Quercus marilandica Black Jack Oak
Vaccinium arboreum Tree Blue Berry
Callicarpa americana Beauty Berry
Ulmus alata Winged Elm
Forestiera ligustrina
Ilex decidua Deciduous Holley
Crataegus spathulata Hawthorne
    Ranked listings of the 20 most common herbaceous associates from both reports reflect differences associated with the scope of the studies.  The list from Sundew Creek ranks by frequency of direct associated (within 1 m.) with each monitored orchid plant at over 100 sites along a single stream, whereas the Grimes County list ranks species in terms of frequency of occurrence at a series of S. parksii populations in the county. This provides two perspectives of herbaceous associates.  The two most common species listed for Grimes County sites are present at the Sundew Creek site, but individual plants of these species were not noted as being present within 1 m. of a marked S. parksii plant.

Sundew Creek Grimes County

Schizachyrium scoparium              Polypremum procumbens
Helianthemum rosmarinifolium         Diodia teres
Ascyrum hypericoides                 Schizachyrium scoparium
Liatris elegans                      Boltonia diffusa
Hedyotis crassifolia                 Rudbeckia hirta
Croton capitatus                     Ascyrum hypericoides
Drosera annua                        Dichanthelium lanuginosum
Heterotheca graminifolia             Crotonopsus elliptica
Eryngium yuccifolium                 Croton capitatus
Oxalis violacea                      Eupatorium compositifolium
Andropogon ternarius                 Heterotheca pilosa
Croptilon divaricatum                Stylosanthes biflora
Chasmanthium sessiliflorum           Rubus trivalis
Aster ericoides                      Aster patens
Aster patens                         Bigelowia virgata
Heterotheca latifolium               Croton glandulosus
Hypoxis hirsuta                      Lechea tenuifolia
Paspalum floridanum                  Liatris spp.
Tridens flavus                       Ptilimnium nuttallii
Antennaria fallax                    Rhynchospora globularis

    The Grimes County listing also combines data from the two primary types of S. parksii habitat, i.e., wooded stream bank and open head of drainage. Thus, a fairly common inhabitant of eroded, exposed heads of drainage, Bigelowia virgata (images), is not present at the Sundew Creek site.  Its tendency to occupy this type of habitat places B. virgata (now placed as Bigelowia nuttallii) as a fairly good indicator species of S. parksii at head of drainage sites in both Grimes and Brazos County.  Indicator associates for the stream bank habitat include Liatris elegans (Pink Gayfeather - images), Heterotheca graminifolia (narrow leafed Heterotheca - now placed as Pityopsis graminifolia - images) and Drosera annua (Sundew - now placed as Drosera brevifolia - images).

Distribution and Abundance

    As indicated above, S. parksii is part of a 'complex' of species that are linked to a wide-ranging and polymorphic taxon, S. cernua. Structural differences between S. parksii and the 'typical' form of S. cernua that occupies Post Oak woodlands of the Navasota Valley are not conspicuous to the untrained eye.  However, past work with survey teams that include individuals with no background in plant systematics has demonstrated that - within the local context of Spiranthes variation - these two taxa can be accurately differentiated by the non-specialist.  Field identification of this species should be confirmed by a specialist.  While this specialist can be a trained plant systematist for many flowering plant groups, firm identification or 'determination' for specimens representing taxonomically difficult groups can only be provided by a trained systematist with research experience with the group in question.  Formal determination by a specialist is essential for those situations that involve specimens of a taxonomically difficult group that are found in unusual habitats or locations well removed from the documented range of the species.  Determination by specialists is a common, fundamental procedure of plant systematics and herbarium curation.  It either confirms or corrects plant identifications by the application of both expertise and objectivity.

    Working specialists for the genus Spiranthes in North America are Paul Catling and Charles Sheviak. They determined specimens associated with initial surveys and the S. parksii distribution depicted in the original Recovery Plan.  This distribution involved a series of documented populations along the Navasota Valley in Brazos, Grimes, and Robertson County and one population on the western wall of the Brazos River Valley in Burleson County. As indicated above, Dr. Sheviak also examined living plants from both central and marginal populations within this range in 1986.

    Subsequent survey work by teams from the Department of Biology at TAMU have produced no extensions beyond the Texas Counties depicted in the original recovery plan.  However, a map published by the Texas Parks and Wildlife Department (Poole and Riskind, 1987) extended the S. parksii range northward along the Navasota Valley into Madison County, westward into the Brazos and Yegua drainages in Lee and Washington Counties, and eastward beyond the Post Oak Savannah into Jasper County.  The most recent depiction of this species range has been provided by the Texas Municipal Power Agency (TMPA, 1991-Tab. 1). This extends the range further northward along the Navasota Valley to Leon and Freestone Counties.

    A summary of current information on populations within this range, modified from that provided by the TMPA (1991, Tab. 1) is presented below in Tab. 1. This includes, by county, the number of documented sites of occurrence (TMPA, 1991), number of plants counted at all sites (TMPA, 1991), confirmation of species identification by a specialist, number of records currently present in the Texas Natural Heritage Program database, and presence of a voucher herbarium specimen or preserved floral sample.

Table 1. Summary of County Distribution Records

Known Sites
Id Confirm
Record #

    Simple inspection of this expanded distribution of S. parksii, as depicted by state county maps produced by Poole and Riskind (1987) and the TMPA (1991) suggests a relatively broad range of distribution.  The number of new sites discovered since 1984 leads to the notion that, within this new range, "survey success is largely a function of level of effort expended" (TMPA, 1991).  This perspective on S. parksii distribution is also put forward by Bridges and Orzell (1989) in their report of plants that they identify as S. parksii from the Angelina National Forest.  They indicate that "The occurrence of Spiranthes parksii in Jasper Co. implies that the intervening counties westward to Grimes Co. which have outcrops of the Catahoula Formation are within the potential range of this species.  In addition, our cursory observations indicate potentially suitable habitat as far southwest as Colorado and Fayette counties."

    Section 7 consultations relating to S. parksii have, to date, involved two entities; the Texas Department of Transportation (Highway expansion, rt. 6) and the TMPA (lignite mining).  Data listed above relating the range extension from 4 (1984) to 10 (1991) Texas counties has been provided by employees of the State of Texas and the TMPA.  These workers have also promulgated a perspective, either in formal reports (TMPA) or the scientific literature (State of Texas), that suggests a broader distribution for this species.  As indicated by the result of the most recent Section 7 consultation (Short, 1991), this notion of a broad range of distribution can influence decisions on environmental impacts and their consequences.

    Given this circumstance, and the critical nature of distributional data in terms of listing and recovery of an endangered species, it is prudent to insure that distributional data documenting S. parksii populations from beyond the Navasota Valley are based on a firm foundation in terms of species identification and site documentation.

LEE COUNTY:  There appears to be no basis for this record, aside from the reference by Bridges and Orzell (1989); "Field work in 1986 and early 1987 has added Leon, Madison and Washington Counties to its known range, with rosettes possibly representing this species observed in Lee Co (Kathy Jordan, pers. comm.)."  It should be noted that differentiation between S. parksii and S. cernua on the basis of leaf characters is not possible. Updated corrections for Poole and Riskind (1987), issued by the Texas Parks and Wildlife Department in 1990, indicate that Lee County should be deleted from the S. parksii range and Leon County should be added.  Lee County is, however, included within the distribution of this species by the TMPA (1991).

JASPER COUNTY:   There evidently has been no confirmation of this species identification.  Given prior determination of Spiranthes from this site as Spiranthes X laciniata (Marietta and Nixon, 1983), taxonomic complexity of the S. cernua complex, and the unusual nature of the Jasper County habitat, firm identification for this record is essential.  Absence of S. parksii specimens in collections that represent the East Texas flora (10 herbaria cited in Bridges and Orzell, 1989) indicate that this record should have been determined by a specialist prior to publication.  The voucher specimen associated with the Jasper County site (Orzell and Bridges 4854) is not present in the cited herbarium of deposition, the University of Texas Herbarium.  The plant was not found during surveys of this area during the Fall of 1992.

FREESTONE COUNTY:   As indicated in the Texas Heritage Program database, this record was evidently established on the basis of a conversation involving representatives of the program and D. Severinson.  Reference to this population on a "seepage slope" in association with Eriocaulon, a genus of bogs and wetlands suggests that this record should be based on more than a conversation. However, a recent conversation with the collector, Mr. Standley Jones, indicates that a floral specimen from this site was collected and preserved.

WASHINGTON COUNTY:   This county record is documented by a single record in the Texas Natural Heritage Program database.  It was evidently established on the basis of a letter submitted to that program in 1987 from Kathie Jordan describing a population near Lake Sommerville.  A thorough survey of this area was conducted by TAMU survey teams in 1983 after discovery of the outlier population in Burleson County.  This was repeated in 1984.  The species showed a very strong flowering response over these two years and these surveys produced no new populations of S. parksii from the Lake Sommerville area.  Aside from the TNP reference, this county record is undocumented and the identification is not confirmed.  Kathie Jordan (= Kathie Parker = Tejas Ecological Services) was contracted by TMPA to transplant S. parksii plants from sites to be mined in Grimes County.  A source site for this transplantation was in the Lake Sommerville area.  Given the nature of available documentation for the Washington County record for this species presented by maps produced by the TNP (Poole and Riskind, 1987) and TMPA (1991), the possibility that this record is based on plants originally native to Grimes County cannot be excluded.

    Records for both Leon and Madison counties are confirmed and documented. Thus, if only documented county records are accepted, the distribution of S. parksii includes 6 counties. Inclusion of county records from counties discussed individually above, extends this to 10 counties.  As indicated by the numerical data presented in table 1, both interpretations produce the same conclusion.  In terms of genetic integrity and coherence, S. parksii is composed of two 'core' population systems.  While small, remote populations are of interest from both historical and genetic points of view, continued existence of this species will depend on the integrity and general survival of populations that occur within potential genetic contact in the core areas.  In addition to genetic continuity, the 'core' populations also tend to be well populated and stable, in that flowering plants are evident every year.  This is not the case with the peripheral populations. Flowering S. parksii plants were not found during revisits to the Burleson County population in 1991 and 1992.  Flowering plants have not been observed at the western Robertson County site, near Hearne, since 1983.  Searches of the Jasper County site produced only two flowering Spiranthes this year, neither could be identified as S. parksii.  Of seven S. parksii sites documented in 1985 for the eastern portion of the TMPA mining area (TMPA, 1991 - Site Compendium Map), only one contained flowering plants in 1991 (Parker, 1992).

    Known distribution of populations within the two core areas, southern Brazos County and central Grimes County, is depicted in Figure 1.  If table 1 contains accurate data, this includes 93% of the known population sites for this species and 95% of the plants censused in various surveys.

    Figure 1 contains a summary of mapped surveys conducted by both the Department of Biology, TAMU and the Texas Municipal Power Agency Environmental group and their contractors. The map contains primary and secondary roads, cities and towns, the 250' elevation line, and mapped survey sites. Areas explored with negative results are indicated by a triangle. Known locations of S. parksii populations are indicated by a circle. Large circles mark 'high density' sites, usually 'head of drainage' areas that carry more than 10 S. parksii individuals per 10 m. of stream margin.

Impacts and Threats

    Three factors that are moving this species toward extinction were listed in the original recovery plan: 1) human elimination or modification of the habitat; 2) collection by orchid fanciers; and 3) artificial maintenance of the habitat. Work with this plant over the past 9 years has demonstrated that factor #1, elimination or modification of the habitat, is the only significant threat to this species. There has been no evidence of collecting by fanciers and, if this has occurred, the impact is minimal compared to habitat destruction. Close monitoring of the population at Lick Creek Park has demonstrated quite clearly that natural disturbance is sufficient to maintain robust and viable populations over a 10 year period. Human intervention is not required to maintain habitat, at least over the short term (15-20 years).

    While the primary problem, human-mediated habitat destruction, stands as the only significant threat, events over the past 9 years indicate that the problem was not overstated in the original Recovery Plan. Despite an economic 'slow-down' during the 1980's, habitat destruction in southern Brazos and central Grimes counties has proceeded at rates comparable to those projected by the Recovery Plan.

Grimes County: The 30-year life-of-mine area of the Texas Municipal Power Agency's Gibbons Creek lignite mine in central Grimes County consists of 37,000 acres in the general area of the Gibbons Creek drainage extending from the Navasota Valley, through Carlos to Singleton (see Fig. 1). Much of this includes prime Post Oak woodland habitat occupied by S. parksii. Mining and associated activities of TMPA has consumed over 7,000 acres since 1985. Projected habitat destruction over the next 5 years will cover a comparable area. This will include 26 sites known to support S. parksii populations.

Brazos County: The City of College Station has expanded 7.6 square miles from 1983 to 1991. Much of this has been toward the center of S. parksii populations to the South. The southern City limit expanded to include 3,291 acres in 1991 that were not part of the City in 1983. Two, fairly large areas were annexed by College Station in 1992. One of these, Foxfire Development, contains S. parksii. Oak woodlands at the southern margin of this municipal expansion have been heavily impacted by habitat destruction over the past 9 years. The most extensive of these, 1400 acres of prime habitat between Greens Prairie Road and Texas World Speedway, is the Pebble Creek Golf Course and residential development. This massive project has extirpated numerous documented populations along the tributaries of Lick Creek. Other populations have been eliminated by recent construction of two medical facilities (Sandstone/Greenleaf) at the junction of State Rt. 6 and Greens Prairie Road, as well as operations of the College Station Land Fill which is positioned between Greens Prairie Road and Rock Prairie Road. Two hundred acres of prime woodland habitat along State Rt. 6, just north of Texas World Speedway has been set aside for development as the College Station Industrial Park. As indicated in mapped projections for development (College Station, 1990), oak woodlands North of Texas World Speedway, between State Hwy. 6 and Rock Prairie Road, will be developed to produce a mix of residential, commercial, and industrial areas in the near future.

    The unanticipated, and most disturbing, extirpations have occurred at two sites adjacent to Texas World Speedway. Both sites contained plants that were permanently marked for demographic studies. The single, high-density site in Brazos County was bulldozed in the spring of 1986. Many plants at this site, one of the two recommended for protection as a 'safe site' in the initial Recovery Plan, were selectively eliminated in that the stream bank was cleared and filled to produce a series of impoundments. Additional development to utilize the series of ponds produced, such as lots for homes, has not occurred to date. Inspection of this area in 1992 revealed that the species continues to occupy high banks and openings above the impoundments. The second site, a woodlot at the junction of Peach Creek Road and State Rt. 6, was selected as an 8-acre preserve to be set aside in mitigation for habitat destruction caused by expansion of State Rt. 6. The site was cleared and bulldozed prior to purchase. Despite the clearing, the site was purchased and fenced, evidently to fulfill the mitigation requirement. These episodes of habitat destruction, occurring in the heart of the Brazos County core area, do not appear to be associated with 'normal' impacts of municipal development and mining that were covered in the initial proposal. It is quite possible that elimination of these populations is related to activities associated with the Recovery Plan and section 7 consultations. If this is the case, then irrational actions by private land owners must be included as a threat to the long-term survival of this species.


    Work in 1983 with the population positioned North of Texas World Speedway included collection of fruit from six plants that had been enclosed in pollinator-exclusion devices prior to anthesis. All plants set fruit. Thus, S. parksii can either self-pollinate or, as is the case with other elements of the S. cernua complex, it can produce seeds without sexual reproduction (agamospermy). Examination of seeds by Dr. Sheviak indicates that agamospermy is probably the most common mode of seed production. The bumble-bee (Bombus sp.) is only potential pollinator observed on S. parksii. Foraging bumble-bees are rarely observed, but quite useful in locating plants when present. They are the common pollinator for this genus (Sheviak, 1982) and, since outcrossing appears to occur (Sheviak, 1986), a possible pollen vector for S. parksii.

Seeds produced by plants enclosed in pollinator exclusion devices germinated well in sterile tissue culture medium under artificial light. Seedling transfer to standard potting soil in closed containers presented no problem. As indicated in the initial Recovery Plan, efforts to transfer seedlings to 3" pots under greenhouse conditions were not successful.  However, plants retained in closed containers grew well under artificial lighting and growth chamber conditions for two years.  Plants transferred from these containers to natural habitat at Lick Creek Park survived to produce rosette leaves during the second year.  Plants did not survive to produce leaves the third year. The site was disturbed by feral pigs and, as a result, it is impossible to determine if future efforts with seedling establishment in nature will be successful.

    Mitigation for habitat destruction at the TMPA mining site has involved transplantation of mature plants.  These efforts, complicated by flooding at recipient sites and tracking problems (see Parker, 1992), have demonstrated that at least a portion of plants transplanted at the rosette, vegetative stage are able to mature to flowering.

    Work at 'safe sites' at Lick Creek Park in Brazos County and the TMPA mining area in Grimes County is defining annual population flux, at least as this is indicated by presence or absence of flowering plants.  The annual addition of 'new' flowering plants to the population at Lick Creek Park is present in Fig. 2.

This reflects the 'good' years of 1986 and 1987, the sharp decline in 1988 and 1989, and a positive trend in 1990 and 1991. A similar pattern is present in data taken from 4 'safe sites' in Grimes County as depicted in Fig. 3.

    While data presented in Figs. 2 and 3 reflect different aspects of flowering, congruence between high and low years suggest that flowering response is based on factors that impact the total range of the species. One conspicuous factor is available moisture during both the primary vegetative phase (April/May) and the flowering phase (August/September). Monthly rainfall from the weather station at Easterwood Airport (Brazos County) are presented in Fig. 4. This reflects combined rainfall for the period of rosette growth (April and May) and the period just prior to flowering (August September) as they relate to the 40 year average for these periods.

    General congruence between flowering responses and rainfall is indicated by a peak in 1986, depression in 1988 and 1989, and increases in 1990 and 1991. This supports other data (TMPA, 1991) to suggest a fairly strong linkage between available moisture during critical periods of vegetative growth and flowering and annual variation in flowering.

    Data presented above demonstrate that annual flux in flowering at S. parksii sites is extreme and, in addition, that this variation appears to be related to available moisture. The relationship between number of flowering plants and the actual number of individuals (flowering, vegetative, mycorhizomes) at a given site remains as an important question. Is it reasonable to assume that these sites carry large numbers of plants, with only a subset flowering during a given year?

Data relating to this question are presented in Figs. 5 and 6.  These charts depict the fate of the 1986 (Fig. 5) and 1987 (Fig. 6) cohorts of S. parksii that entered the Sundew Creek monitoring study as "new" flowering plants in those years. Subsequent flowering and rosette formation was tracked from the first year of flowering to the rosette formation period of 1992. Both cohorts show a similar pattern. Essentially all plants produced rosettes after flowering. However, very few produced inflorescences in subsequent years. Rosette production showed a general and relatively sharp decline.

    These data do not support the notion that perennialism in S. parksii is characterized by annual flowering. It appears that most plants in a given flowering cohort require much more than a single year of vegetative growth to accumulate enough photosynthate to support the production of a flowering stalk. The observed decline of rosette production in the tracked cohorts at Sundew Creek suggests the possibility that flowering within a given cohort is not cyclic, i.e., it is possible that most plants flower only once during their lifetime. If this is true, then the observed annual variation in frequency of flowering at S. parksii sites might reflect the presence of 'new' flowering plants as opposed to periodic turnover in the vegetative and flowering phase among plants in the population that have flowered in the past.

    Thus, the annual flux in observed flowering plants within a given S. parksii site could reflect variation in size of annual cohorts. The size of these cohorts, as indicated by number of flowering individuals, could mark annual differences in germination and development of S. parksii mycorhizomes. Since the process of germination and mycorhizome development in this species probably covers a period of many years, assessment of significant climatalogical variables would require examination of variation in weather data from, perhaps, a decade prior to peaks and depressions of flowering. The observed linkage between rainfall and flowering in monitored populations of this species (Figs. 2, 3, and 4) could reflect a response among past cohorts of mycorhizomes, developed underground and poised to respond, to a relatively superficial environmental trigger to either generate a pre-flowering rosette in the Spring, or move directly to flowering in the Fall. It is also quite possible that observed annual variation in floral expression in this species is based on a combination of responses from new waves of annual cohorts as well as plants that have flowered in the past.



    The initial Recovery Plan placed the creation of preserves, or 'safe sites' as a primary conservation objective. Movement in the direction of this objective has occurred in three areas; Lick Creek Park in Brazos County, the TMPA mining area in Grimes County, and the Peach Tree Preserve in Brazos County. Two of these areas, Lick Creek Park and the TMPA mining area, contain prime habitat and healthy populations of S. parksii. However, both areas have limitations with regard to long-term, continuous preservation of these populations. The Peach Tree Preserve, on the other hand, is well established as a long-term preserve. It, however, carries no populations of S. parksii.

Peach Tree Preserve: State Highway 6 transects the area of high population density South of College Station near Texas World Speedway. Construction associated with expansion of this highway extirpated S. parksii populations and large areas of prime habitat. Section 7 consultation involving the State Department of Highways and Public Transportation apparently established two areas know to carry populations of the species to be protected in perpetuity as mitigation for extirpations resulting from highway construction. While one of these areas, a highway right-of-way along State Highway 974 in northern Brazos County, was eventually rejected as a long-term safe site, another area at the northeast corner of the junction between Peach Tree road and State Highway 6, was formally designated as a long-term preserve. This was an ideal site in that survey teams from Texas A&M had marked many S. parksii plants at the site for annual revisits and monitoring. Unfortunately, the owner of this property cleared about 90% of this 8 acre site prior to finalization of the mitigation process. The process, however, proceeded. The land was evidently purchased by the State and, with little concern for biological reality, circumscribed by a tall, metal fence. A survey of this area in the Fall of 1990 (Wilson, 1991a) and 1991 indicates that, as one might expect, S. parksii is not part of the colonizing, mostly annual flora that occupies the cleared area at this site. The small section of forested land within the fenced area does not support S. parksii, although the species does occur along stream banks in the adjacent woodland. This well protected weed patch might progress through floristic succession to an oak-hickory forest over several human generations. However, given the obvious southerly developmental trend for the city of College Station, it is reasonable to assume that a seed source for S. parksii colonization of the site will not be present in the general area.

Lick Creek Park: Site visits associated with the Section 7 consultation for expansion of State Highway 6 also included meetings with representatives of the City of College Station. The City was involved at that time with land transfers associated with municipal and commercial development within the core area of S. parksii distribution in southern Brazos County, i.e., the College Station Industrial Park, the College Station Sanitary Land Fill, the Pebble Creek Golf Course/Residential Development, and other projects (see 'impacts' above). The City proposed creation of a 512 acre park that was to stand as a natural area. While not formally linked to the consultation process, State and Federal officials visited the proposed park site and viewed this development as a positive step that was fully consistent with goals of the recovery plan (Orzell and Poole, 1986). The area, eventually named Lick Creek Park, was subsequently development by the City to meet community needs that were formally stated. These included the establishment of 'protected habitat for rare and endangered species', a site for 'living outdoor education', an outdoor laboratory base for the study of natural science, and outdoor recreation opportunities.

    About 20% of the land within Lick Creek Park supports the upland Oak-Hickory habitat of S. parksii. The species occurs along four streams that drain the northern uplands of the park. In terms of genetic diversity, these populations represent an excellent sample from the southern Brazos County 'core' area. One of these populations, the Sundew Creek group, has been monitored since 1986 (see 'research' above). Thus, Lick Creek Park would appear to satisfy a major objective of the original Recovery Plan, i.e., a 'safe site' in the southern Brazos County center of distribution. There are, however, two significant problems associated with this area.

As a municipal park, Lick Creek Park is controlled by a Department of Parks and Recreation that operates under and changing directional control. The composition of City government and appointed positions in the City Parks Board changes through time. Developmental and management priorities for this type of facility are in continuous balance between recreational use and natural preservation. As indicated by a recent regional 'mountain bike' race at the park, promoted by local merchants and supported by the City, this balance can easily shift to diminish the 'safe-site' function of Lick Creek Park.

Lick Creek Park has not been formally designated as a long-term preserve. The area could be developed for other uses at any time. There is no long-term management plan and no formal commitment from municipal officials to either establish the area as a preserve in perpetuity, or develop long-term management plans for the park area. Thus, while Lick Creek Park stands as an ideal S. parksii preserve and monitoring area, as depicted in the Recovery Plan, it is not a stable safe site. This is indicated by a recent request to the City for permission to transfer plants of the endemic, Liatris cymosa, from an area to be impacted by local construction into Lick Creek Park. This was denied by City officials with an indication that the transfer would not be in the "City's best interest" because "the park has no formal master plan and the safe site designation may restrict future use" (Beachy, 1992).

Gibbons Creek Drainage: The U.S. Fish and Wildlife issued a "jeopardy" biological opinion relating to S. parksii to the Environmental Protection Agency in 1985. This concerned the impact of surface mining associated with the Texas Municipal Power Agency's Gibbons Creek Power Plant. This resulted in the designation of four 'protected' sites within the TMPA mining area in central Grimes County. A fifth site was established in association with a second, "nonjeopardy" biological opinion resulting from formal Section 7 consultation in 1991. The five TMPA preserves constitute just over 175 acres of prime habitat within the TMPA mining area near Carlos, Texas. The TMPA preserves are specially described and mapped in the amended conservation agreement of March, 1992. This agreement, based on the most recent Section 7 biological opinion (16 Oct 91), supersedes conservation agreements and management plans resulting from prior Section 7 consultation.

    Habitat with the five TMPA "safe" sites constitutes 5% of the S. parksii habitat that has been or will be destroyed by TMPA, assuming that 25% of the total land area designated for mining or construction either currently supports or, prior to past mining, supported S. parksii populations. Thus, the "nonjeopardy" biological opinion resulting from the most recent Section 7 consultation considered elimination, due to past or projected mining activities, of 95% of those populations of a federally listed species that occur within a 14,000 acre area that is positioned at a major center of distribution. While this would appear to represent a major impact on the species, plants inhabiting the five TMPA protected sites probably carry an adequate sample of genetic diversity contained within the Grimes County population system. Thus, if conservation of genetic diversity is a primary concern, then the small remnant in the five TMPA safe sites could function as a long-term reservoir representing the eastern extent of the current range of distribution. Unfortunately, the conservation agreement currently in affect includes no provision to insure that this will be the case.

    As currently defined, the TMPA "safe" sites are protected and managed for a defined period of time, i.e., the estimated 20 year period of TMPA mining operations. The current conservation agreement does not include the term "perpetuity" and there is no provision for transfer of land ownership during the period of mining. Land associated with the five protected areas is either owned directly by TMPA or leased by TMPA from private owners. Since provisions of the conservation agreement cease to be effective once mining is completed, it is quite possible that control of all five 'safe' sites now maintained within the TMPA mining area will revert to private ownership when mining operations are completed. This outcome, a likely consequence of land transfer conditions stated in the conservation agreement, would eliminate the temporary S. parksii preserves in the TMPA mining area and negate all management/monitoring work conducted in those areas during the period of mining activity.


    Efforts to cultivate S. parksii under artificial conditions, beyond those initiated at Texas A&M mentioned above, are currently centered at the Atlanta Botanical Garden. Work accomplished by Ron Determann, the Conservatory Superintendent at this facility, has resulted in 'hundreds' of S. parksii under cultivation. These plants flower each Fall, set fruit, and evidently disperse in the greenhouse environment to the extent that Mr. Determann characterizes the species as an 'invasive' weed. Mr. Determann also indicates that estimated time from seedling to flowering plant is less than two years under greenhouse conditions. Given current data and theory regarding growth and reproduction Spiranthes in general and this particular species under natural conditions (see Research above), this is an interesting result.

    The Atlanta population of S. parksii is based on seed that were collected at one of the 'protected' sites in the TMPA mining area in Grimes County, Texas by Kathy Parker of Tejas Ecological Services in September of 1987.  These were sent to the Eric Young Micropropagation Center at Marie Selby Botanical Garden in Sarasota, Florida.  Germination was accomplished on agar using a tissue culture medium. According to Mr. Harry Luther at Selby Garden, additional operations were performed which generated non-callus, vegetative grown from seedling tissue.  Plants resulting from this process were sent to the San Antonio Botanical Garden and distributed from that facility to the Mercer Arboretum (Houston), and the Atlanta Botanical Garden.  The last of six plants sent to the Mercer Arboretum died this year.  About 10 plants remain at the San Antonio Facility and Mr. Paul Cox of that facility indicates that, while these flower under cultivation in San Antonio, reproduction by seed has not bee observed.

    The narrow genetic base represented by plants currently under cultivation, combined with what appears to be clonal propagation of vegetative tissue at Selby Gardens, limits the value of this effort as a measure to conserve natural genetic variation represented by S. parksii. This effort does, however, demonstrate a potential to salvage genetic diversity from natural populations and 'bank' that diversity as either greenhouse-grown plants or tissue culture collections.

Part II