Ecological monitoring of the endangered Huachuca water umbel (Lilaeopsis schaffneriana ssp. recurva: Apiaceae).
At the time L. schaffneriana ssp. recurva was listed as endangered, the subspecies had been documented from 22 sites in Santa Cruz, Cochise, and Pima counties, Arizona, and adjacent Sonora, Mexico, and had been extirpated from six of the 22 sites. The 16 extant sites were located in four major watersheds: upper portion of the San Pedro River, upper portion of the Santa Cruz River, Rio Yaqui, and Rio Sonora; at elevations of 1,150-2,135 m (United States Fish and Wildlife Service, 1997). In spring 2001, Huachuca water umbel was discovered in two areas in a dense cattail-bulrush marsh at Bingham Cienega Preserve, Pima Co., Arizona, which is ca. 855 m in elevation in the floodplain of the lower portion of the San Pedro River (P. J. Titus, in litt.). This population was ca. 60 km north of, and 295 m lower in elevation than, known sites of the subspecies at the time.
Little is known about specific habitat requirements for this subspecies in the American Southwest, except that it occurs in springs and drainages in southeastern Arizona that were historically subject to natural-disturbance regimes, such as seasonal fires and flooding (United States Fish and Wildlife Service, 1997). Loss of habitat related to human alteration of hydrologic regimes is related to disappearance of [greater than or equal to] 4 historic locations of the subspecies (K. Johnson et al., in litt.). Lilaeopsis schaffneriana ssp. recurva is a clonal subspecies with an opportunistic strategy that ensures its survival in healthy riverine systems, cienegas, and springs. In upper watersheds that generally do not experience scouring floods, the subspecies occurs in microsites where interspecific competition in plants is low (Beacham et al., 2001). As a result of a transplant study that sought to improve population stability in an occurrence at San Bernardino National Wildlife Refuge, Cochise Co., Arizona, P. L. Warren (in litt.) concluded that success of transplants was related to competitive effects of surrounding vegetation and that L. schaffneriana ssp. recurva is susceptible to competition from wetland emergent species. Known populations typically are located relatively close to surface-water channels in areas with abundant sunlight and low competition with other species, although a few populations occur in shady wetland understories of Fremont cottonwood-Goodding willow (Populus fremontii-Salix gooddingii) forests. Anthropogenic changes in the landscape over the past century have resulted in altered disturbance regimes including fire suppression and intensified or controlled flooding. The subspecies was federally listed as endangered on 6 January 1997 and the listing was believed to be necessary because of threats posed by degradation and loss of wetlands in the region (United States Fish and Wildlife Service, 1997).
This diminutive and difficult-to-detect subspecies is perennial, and can grow vigorously through the mild winters of southeastern Arizona, if sufficient moisture is present (Titus and Titus, 2008x; K. Johnson et al., in litt.). The plant is clonal and produces weak, spreading, underground rhizomes that enable populations to expand significantly in size between years, seasons, or both; in response to local environmental conditions including temperature, light, and availability of water. The small hollow leaves often are <2 cm tall and the tiny multiple-flowered umbels arising from submerged rhizomes are pale green in color, shorter than the leaves, and consist of 3-10 individual flowers that are typically 1-2 mm wide. Flowering has been observed episodically March-October (Titus and Titus, 2008x). Fruits are relatively large (2 by 2.5 mm), corky, and buoyant (Affolter, 1985); thus, hydrochory is expected to be the main mechanism of dispersal.
This subspecies is of particular interest because it seems uniquely adapted to wetland conditions in southeastern Arizona and adjacent Sonora. Historically, drainages in this area consisted of broad, shallow watercourses in bottoms of valleys that collected precipitation and overland flow from large watersheds. During dry seasons, a primary central channel usually conveyed water, but during major precipitation events, water overflowed the main channel to fill marshy side channels and off-channel depressions (Betancourt and Turner, 1988). Small, weakly rooted clumps of the plant tear off as a result of scouring during flood events and float downstream to take root elsewhere. Some of these clumps survive, depending on specific conditions where the clump is deposited. Since the 1850s, many drainages in southeastern Arizona have been altered drastically by increased pressures on the landscape and currently most drainages consist of deeply incised channels that no longer are connected to the former broad floodplain (Johnson et al., 1989). Because many watercourses are channelized, scouring during flood events is much more intense and there are few off-channel habitats suitable for new colonization. Continued drought exacerbated by increased pumping of groundwater has been correlated with loss of perennial flow in many drainages throughout the range of L. schaffneriana ssp. recurva (Tellman, 2001). Threats posed to L. schaffneriana ssp. recurva related to watershed degradation also threaten other special-interest species that are uniquely adapted to southeastern Arizona drainages and wetlands, including the lowland leopard frog (Rana yavapaiensis), Mexican garter snake (Thamnophis eques megalops), Gila topminnow (Poeciliopsis occidentalis), Canelo ladies tresses (Spiranthes delitescens), and Sonora tiger salamander (Ambystoma tigrinum stebbinsi; United States Fish and Wildlife Service, 1997). For example, lowland leopard frogs were observed several times in our study area and known occurrences of Canelo ladies tresses and Gila topminnows are documented in the same locales as several populations of L. schaffneriana ssp. recurva. Consequently, declines in viability of populations of L. schaffneriana ssp. recurva could signal declines in other native species in the same habitats.
Lilaeopsis schaffneriana ssp. recurva is particularly vulnerable to extinction because its small and infrequent populations are highly affected by stochastic events (United States Fish and Wildlife Service, 1997), and because existing populations often are widely separated with a lack of suitable habitat continuity. This study was implemented to provide further information on ecology of L. schaffneriana ssp. recurva, including the role of competition in persistence of the subspecies in floodplain habitats.
MATERIALS AND METHODS--Study Area--Bingham Cienega Preserve (855 m elevation) is a 115-ha tract within the floodplain of the lower San Pedro River just north of the confluences of Redfield Canyon, which drains the Galiuro Mountains to the east, and Edgar and Buchman canyons, which drain the Santa Catalina Mountains to the west. A spring on the preserve provides supplementary water that contributes to support of riparian deciduous forest, mesquite bosque, extensive sacaton grasslands, and cienega wetlands. Regional climate is semiarid, with an annual precipitation of ca. 35 cm, most of which is bimodal and falls during winter and summer (http://cdo.ncdc.noaa. gov/climatenormals/clim60/states/Clim_AZ_01.pdf). The preserve is owned by Pima County Flood Control District and managed by The Nature Conservancy under a long-term contract. The site formerly was used for farming and cattle grazing for more than a century, during which the floodplain wetlands were ditched and bermed to partially drain portions of the land for agriculture. In 1989, Pima County Flood Control District purchased the property and began efforts at hydrologic restoration by breaching the berm and planting native riparian trees and big sacaton (Sporobolus wrightii) in portions of the former agricultural fields. Cattle were removed from the cienega in 1989 (The Nature Conservancy, in litt.). Soils at the preserve are characterized as Glendale silt-loam (http:// websoilsurvey.nrcs.usda.gov/app/WebSoilSurvey.aspx). The Glendale series consists of deep soils formed in mixed alluvium on stream terraces and alluvial fans. In May 2001, L. schaffneriana ssp. recurva was discovered in Bingham Cienega Preserve (P. J. Titus, in litt.), making this the northernmost extant population of the species. The population is somewhat different from other known populations in that it is located within a dense bulrush-cattail (Schoenoplectus-Typha) marsh and is relatively distant from the main channel of the San Pedro River.
The study site is located within the area containing the population of L. schaffneriana ssp. recurva, which occupies a shallow channel that conveys springflow along the western edge of former agricultural fields that now comprise an extensive bulrush-cattail marsh. A mesquite woodland borders the edge of the marsh and separates it from Redington Road. In September 2001, the marsh was dominated by bulrush (Schoenoplectus pungens) and cattail (Typha domingensis) and was inundated with 15-30 cm of water. During the course of this study, however, the entire marsh, including the study site, experienced a dramatic shift in aboveground vegetation, perhaps, due to prolonged drought exacerbated by other hydrological changes, such as withdrawals of groundwater and deposition of sediment following catastrophic fire in the upslope portions of the watershed. By 2005-2006, much of the cienega was devoid of vegetation and some areas were dominated by monospecific stands of the facultative upland sunflower Helianthus annuus, but by the end of 2006, springflow had returned and wetland vegetation was becoming reestablished in portions of the marsh.
Experimental Design--In autumn 2001, six 1-[m.sup.2] monitoring plots (three treatment and three control) were established in the study site ca. 6-8 m from the wetland-upland boundary. Plots were ca. 6 m apart and were assigned treatments randomly. Censuses of plots commenced in October 2001 and continued through 2006 on the following dates: 11 October 2001 (plant cover only), 25 October 2001, 1 and 8 November 2001, 4 and 25 April 2002, 9 May 2002, 12 September 2002, 10 and 24 October 2002, 14 November 2002, 30 January 2003, 27 February 2003, 28 April 2003, 11 September 2003, 23 October 2003, 28 March 2004, 7 September 2004, 15 March 2005, 5 October 2005, and 6 January 2006. These plots also were used for a seedbank study of Bingham Cienega (Titus and Titus, 2008b). Beginning on 25 October 2002, three of the census plots were clipped of potentially competing perennial species at the ground surface to examine whether reduction of interspecific competition for nutrients, water, or sunlight would affect the population of L. schaffneriana ssp. recurva. The only species in plots with the potential to compete with L. schaffneriana ssp. recurva was Schoenoplectus pungens due to its overwhelming dominance. Vegetation was clipped to ca. 5 cm beyond the boundary of the plot to create a small treatment buffer. After 24 October 2002, plots were not clipped due to lack of vegetation. Clippings were removed from the plot. Control plots were left undisturbed. Because L. schaffneriana ssp. recurva is a clonal subspecies and evaluation of individuals is impossible, leaves of L. schaffneriana ssp. recurva in each plot were counted and measured during each census to assess growth. At each census, plants also were examined for presence of flowers and fruits. At each census, percent cover of each taxon in each 1-[m.sup.2] plot was determined by visual estimation (Elzinga et al ., 1998). A 1-[m.sup.2] plastic frame was used to delimit boundaries of plots. Corners of plots were marked with stakes and wire flags.
Normality of number of leaves, length of leaves, and depth of water were tested by the Kolmogorov Smirnov test and homosedascity by the the Levene test. A natural-log transformation of number and length of leaves improved normality and homoscedascity. Depth of water did not require transformation. A repeated-measures ANOVA was performed to test if depth of water differed between the two treatments across the sampling period. To determine if depth of water was correlated with number and length of leaves, a Pearson correlation test was used. Depth of water in each plot at each census was tested for correlation with number and length of leaves for each of the two treatments for the sampling dates 25 October 2001-9 May 2002 (six censuses). The two treatments were tested separately because of the possibility of a confounding treatment effect. Number of leaves and mean length of leaves per plot were compared by repeated-measures ANCOVA with depth of water nested within treatment as the covariate in an auto-regression-covariance structure for sampling dates 25 October 2001-9 May 2002 (six censuses) using PROC MIXED (SAS Institute, Inc., 2003). Depth of water was the covariate to remove effects due to different depths of water in different plots. Pearson correlation test and ANCOVA did not consider results after the sixth census due to collapse of the population.
RESULTS--On the first sampling date in September 2001, vegetation at Bingham Cienega was dense and dominated by the obligate wetland species Schoenoplectus pungens, and there was standing water in many of the plots (Table 1). By September 2002, vegetation was similar but there was no standing water in the area, although soils were moist below the surface. By January 2003, soils were dry 30 cm below the surface and the ground surface was covered completely by thatch of Schoenoplectus, with a few small seedlings of Polygonum punctatum as the only live, aboveground vegetation. Lack of live, above-ground vegetation at this time might have been due to winter senescence; however, by September 2003, with continued drought, all aboveground vegetation was dead and a fire had burned the thatch of Schoenoplectus. Soils remained dry after that time, and plots became dominated by monospecific stands of Helianthus annuus, which colonized the area in autumn 2004. It is worth noting that although the cienega is fenced, tracks and related signs of cattle were observed in the study site and in the plots.
There were significantly more leaves of L. schaffneriana ssp. recurva in clipped plots than control plots, but number of leaves did not change significantly over time (Table 2). From October 2001 to May 2002, number of leaves of L. schaffneriana ssp. recurva increased in both control and clipped plots; however, the increase was much greater in clipped plots, where numbers of leaves had more than doubled by spring 2002 (Fig. la). By September 2002, number of leaves had declined precipitously in both control and clipped plots, and by 24 October 2002, no leaf of L. schaffneriana ssp. recurva was in any plot.
Depth of water did not differ between the two treatments; however, there was a significant interaction between treatment and time (Table 2) indicating that difference in depth of water between treatments was dependent on time of sampling. Based on results of Pearson correlation analysis, depth of water was significantly correlated with both number (r = 0.591, P = 0.010) and length (r = 0.581, P < 0.011) of leaves for control plots; and depth of water was correlated with number of leaves (r = 0.618, P = 0.006), but not length (r = 0.236, P = 0.346) for clipped plots. Thus, greater depth of water was correlated with more leaves in all plots, and longer leaves in control plots. Leaves were longer in plots where competing vegetation was clipped than in the control plots, but length of leaves did not change over time and there was not a significant interaction between treatment and time (Table 2, Fig. 1b).
Flowering was observed only during the census on 9 May 2002, and occurred in all three of the clipped plots, but in none of the control plots. In each clipped plot, there were hundreds of flowers.
DISCUSSION--The observed increase in number and length of leaves in clipped plots (autumn 2001 and spring 2002) and presence of flowers only in clipped plots suggests that interspecific competition for sunlight, water, or nutrients affect growth and reproductive potential of L. schaffneriana ssp. recurva, and that removal of competing vegetation is favorable for the subspecies under some conditions. In addition, in clipped plots, competing plants were clipped but not entirely eliminated as they continued to resprout; thus, availability of sunlight might be more important than nutrients in determining growth. Although autumn 2001 and spring 2002 were considered periods with drought, moist soils may have provided adequate moisture to allow an increase in leaf production by L. schaffneriana ssp. recurva. By autumn 2002, however, continued drought may have caused a level of desiccation of soil that was detrimental to the ability of L. schaffneriana ssp. recurva to produce leaves. Lack of leaves of L. schaffneriana ssp. recurva since autumn 2002 can be attributed to xeric conditions at Bingham Cienega. Plots subsequently were dominated by Helianthus annus, a species that prefers drier conditions than L. schaffneriana ssp. recurva.
In censuses on 9 September and 10 October 2002, number of leaves of L. schaffneriana ssp. recurva were low (Fig. 1a). In these censuses, more leaves of L. schaffneriana ssp. recurva were in control plots, a reversal of the previous pattern of many more leaves occurring in clipped plots. This suggests that when soil in the study site was dry, as it was in autumn 2002 (Table 1), removal of shading by other plants may have had a negative impact on growth of L. schaffneriana ssp. recurva. This may be because removal of overstory vegetation in clipped plots caused more sunlight to reach the soil surface and increase drying of the soil. Our treatment might have, at least partially, mimicked natural fire, grazing, flooding, or other low-intensity disturbances that historically influenced areas that supported L. schaffneriana ssp. recurva (Turner, 1987). Other rare plants, e.g., Plantago cordata, exhibit a similar dependency on low-level disturbance regimen in riparian systems (Bowles and Apfelbaum, 1989). As of October 2006, no aboveground L. schaffneriana ssp. recurva had been observed at Bingham Cienega Preserve. In 2006, southeastern Arizona received summer rains that enabled spring flow to resume and ponding was observed in the former marsh; however, flow did not reach the study site because of obstruction of the former flow channel by deposition of sediments, and soils within the study site remained dry.
Whether or not L. schaffneriana ssp. recurva uses a persistent seedbank will determine if it can persist in areas such as Bingham Cienega through long-term drought (Titus and Titus, 2008a). Presence of viable seeds in seedbank soils collected 2 years after drought commenced (Titus and Titus, 2008b) demonstrate potential of the subspecies to persist through catastrophic disturbance and prolonged drought. Thus, populations could re-emerge if hydrologies are restored. If the subspecies is present at Bingham Cienega in the future, this will support the importance of the seedbank, and continued monitoring could elucidate more information about L. schaffneriana ssp. recurva in regards to interspecific interactions. This information could assist in formulating potential management objectives for the preserve and other wetlands in southeastern Arizona, and in formulating a recovery strategy for the species.
Currently, the marsh at Bingham Cienega is subjected to periodic fire and drought, which might effectively maintain the plant communities in early seral-stage conditions. Portions of the cienega burned in 2000 and 2003, and periodic droughts have been recorded, including the drought at the conclusion of this study. Discovery of L. schaffneriana ssp. recurva at Bingham Cienega Preserve in 2001 indicates that the subspecies is able to persist, and perhaps even flourish, through these habitat alterations. Future monitoring of the population will enable an assessment of persistence through disturbances, and effects of disturbance on health and stability of the population.
Prior to discovery of this population, all extant populations were restricted to upper portions of four watersheds occupied by the subspecies. Discoveries of the population at Bingham Cienega, a record of an occurrence 55 km to the north (downstream), and a population within the Babocomari River, indicate that the range of this subspecies is larger than formerly assumed and extends beyond the upper reaches of the San Pedro watershed. Because the subspecies might re-emerge after prolonged drought, sites that were surveyed for this subspecies previously should not be eliminated from consideration as potential habitat simply because the subspecies was not detected at that time. This information coupled with the difficulty in detecting the subspecies during surveys underscores the importance of periodic reassessments in any attempt to conserve and recover the subspecies.
[FIGURE 1 OMITTED]
Thanks are extended to M. Falk, J. Fonseca, and The Nature Conservancy for access to the study site and logistical assistance. Biosphere 2 students A. Zuhlke, S. Scott, K. Sovenyhazy, M. Clinton, E. Robertson, H. Garrett, L. Kim, M. Evette, M. Hunter, and E. Stanford assisted in monitoring populations and provided budding enthusiasm that carried this project through to fruition. We thank L. Kennedy and N. Zierenberg for assistance in surveys of the Babocomari River and D. Ruppel for interest in the project and for facilitating access to survey locations on the Babocomari River. Thanks also are extended to I. Vassoler for the Spanish translation and T. Strakosh, C. Castaldo, P. Smilauer, and N. Boynton for statistical assistance. We also thank J. Poole and two anonymous reviewers for improvements to the manuscript.
AFFOLTER, J. M. 1985. A monograph of the genus Lilaeopsis (Umbelliferae). Systematic Botany Monographs 6:1-140.
BEACHAM, W., F. V. CASTRONOVA, AND S. SESSINE. 2001. Beacham's guide to the endangered species of North America. Beacham Publishing Corporation, Detroit, Michigan.
BETANCOURT, J. L., AND R. M. TURNER. 1988. Historic arroyo-cutting and subsequent channel changes at the Congress Street crossing, Santa Cruz River, Tucson, Arizona. Pages 1353-1371 in Arid lands, today and tomorrow: proceedings of an international research and development conference (E. E. Whitehead, C. F. Hutchinson, B. N. Timmerman, and R. G. Varady, editors). Westview Press, in cooperation with Tucson, Arizona, Office of Arid Lands Studies, University of Arizona.
BOWLES, M. L., AND S. I. APFELBAUM. 1989. Effects of land use and stochastic events on the heart-leaved plantain (Plantogo lanceolata Lam.) in an Illinois stream system. Natural Areas Journal 9:90-100.
ELZINGA, C. L., D. W. SALAZAR, AND J. W. WILLOUGHBY. 1998. Measuring and monitoring plant populations. United States Department of the Interior, Bureau of Land Management, Denver, BLM/RS/ST-98/ 005+1730:1-476.
JOHNSON, R., P. S. BENNETT, AND L. T. HAIGHT. 1989. Southwestern woody riparian vegetation and succession: an evolutionary approach. Pages 135-139 in Proceedings of the California riparian systems conference: protection, management, and restoration for the 1990s (D. L. Abell, editor). United States Department of Agriculture, Forest Service, General Technical Report PSW 110:1-544.
SAS INSTITUTE, INC. 2003. SAS software, version 9.1 of the SAS system for Windows. SAS Institute, Inc., Cary, North Carolina.
TELLMAN, B. 2001. Water Resources in Pima County: Sonoran Desert Conservation Plan. Water Resources Research Center, University of Arizona, Tucson.
TITUS, J. H., AND P. J. TITUS. 2008a. Introduction of Huachuca water umbel, an endangered wetland species, into a spring-fed wetland in southeastern Arizona. Ecological Restoration 26:311-321.
TITUS, J. H., AND P. J. TITUS. 2008b. Seedbank of Bingham Cienega, a spring-fed marsh in southeastern Arizona. Southwestern Naturalist 53:393-399.
TURNER, M. G. 1987. Effects of grazing by feral horses, clipping, trampling, and burning on a Georgia salt marsh. Estuaries 10:54-60.
UNITED STATES FISH AND WILDLIFE SERVICE. 1997. Endangered and threatened wildlife and plants; determination of endangered status for three wetland species found in southern Arizona and northern Sonora, Mexico. Final rule, 50 CFR Part 17, January 6, 1997. Federal Register 62(3):665-689.
Submitted 4 September 2007. Accepted 8, June 2008. Associate Editor was Jonis K. Bush.
PRISCILLA J. TITUS AND JONATHAN H. TITUS *
Department of Biology, SUNY-Fredonia, Fredonia, NY 14063
* Correspondent. firstname.lastname@example.org
TABLE 1--Mean percentage of plant cover in three 1-[m.sup.2] control plots and three 1-[m.sup.2] treatment plots in which vegetation was clipped at Bingham Cienega Preserve, Pima Co., Arizona. Data represent means of plots averaged across the census ([+ or -] 1 SD). Depth of water during each season is provided. In autumn 2006, plots were dry and exclusively dominated by the facultative upland sunflower Helianthus annuus and its litter. Dates of seasonal censuses were: autumn 2001, 11 and 24 October, 1 and 8 November; spring 2002, 4 and 25 April, 5 May; autumn 2002, 12 September, 10 and 24 October. Season Autumn 2001 Species Control Treatment Anemopsis californica <1 0 Aster subulatusvar. ligulatus <1 <1 Berula erecta <1 1 [+ or -] 1 Chara and other algal species 2 [+ or -] 2 23 [+ or -] 16 Cirsium neomexicanum 1 [+ or -] 1 0 Eleocharis macrostachya 0 <1 Juncos interior <1 <1 Lemna minor <1 <1 Lilaeopsis schaffnenana ssp. recurva 1 [+ or -] 1 1 [+ or -] 1 Medicago sativa (non-native) 0 0 Nasturtium officinale 0 <1 Polygonum punctatum <1 1 [+ or -] 1 Rumex crispus (non-native) 0 0 Schoenoplectus pungens 100 [+ or -] 0 26 [+ or -] 8 Sisymbrium Trio (non-native) <1 0 Sonchus asper (non-native) 0 0 Typha domingensis 0 0 Unidentified seedlings <1 <1 Total percentage of cover by vegetation 104 [+ or -] 3 53 [+ or -] 15 Thatch of Schoenoplectus 30 [+ or -] 9 7 [+ or -] 3 Bare ground 0 0 Depth of water (mm) 38 [+ or -] 6 47 [+ or -] 4 Season Spring 2002 Species Control Treatment Anemopsis californica 0 0 Aster subulatusvar. ligulatus 0 0 Berula erecta <1 6 [+ or -] 2 Chara and other algal species <1 67 [+ or -] 18 Cirsium neomexicanum 0 0 Eleocharis macrostachya 0 0 Juncos interior <1 <1 Lemna minor 0 <1 Lilaeopsis schaffnenana ssp. recurva <1 5 [+ or -] 4 Medicago sativa (non-native) 0 0 Nasturtium officinale 0 <1 Polygonum punctatum 0 <1 Rumex crispus (non-native) <1 0 Schoenoplectus pungens 93 [+ or -] 3 38 [+ or -] 10 Sisymbrium Trio (non-native) 0 0 Sonchus asper (non-native) <1 0 Typha domingensis 0 <1 Unidentified seedlings <1 <1 Total percentage of cover by vegetation 95 [+ or -] 2 117 [+ or -] 8 Thatch of Schoenoplectus 57 [+ or -] 13 8 [+ or -] 5 Bare ground 0 7 [+ or -] 4 Depth of water (mm) 30 [+ or -] 14 34 [+ or -] 10 (dry) Season Autumn 2002 Species Control Treatment Anemopsis californica 0 0 Aster subulatusvar. ligulatus 0 0 Berula erecta 0 0 Chara and other algal species 0 0 Cirsium neomexicanum 0 0 Eleocharis macrostachya <1 <1 Juncos interior 0 <1 Lemna minor 0 0 Lilaeopsis schaffnenana ssp. recurva <1 <1 Medicago sativa (non-native) <1 0 Nasturtium officinale 0 0 Polygonum punctatum 0 <1 Rumex crispus (non-native) <1 <1 Schoenoplectus pungens 98 [+ or -] 5 39 [+ or -] 5 Sisymbrium Trio (non-native) 0 0 Sonchus asper (non-native) <1 <1 Typha domingensis 0 0 Unidentified seedlings <1 <1 Total percentage of cover by vegetation 98 [+ or -] 5 39 [+ or -] 5 Thatch of Schoenoplectus 53 [+ or -] 18 10 [+ or -] 4 Bare ground 0 42 [+ or -] 15 Depth of water (mm) 0 [+ or -] 0 0 [+ or -] 0 (dry) TABLE 2--Results of repeated-measures ANCOVAs contrasting number and length of leaves of Huachuca water umbel (Lilaeopsis schaffneriana ssp. recurva) in control and clipped plots and a repeated-measures ANOVA contrasting depth of water in control and clipped plots at Bingham Cienega Preserve, Pima Co., Arizona. Number and length of leaves were natural-log transformed for analysis. Number of leaves and mean length of leaf per plot were compared by repeated-measures ANCOVA with depth of water nested within treatment as the covariate in an auto-regression-covariance structure for 25 October 2001-9 May 2002 (6 censuses) using PROC MIXED (SAS Institute, Inc., 2003). Number and length of leaves and depth of water were not compared after the sixth census due to collapse of the population. Variable Factor df F P Number of leaves Treatment 1 13.50 0.001 Time 5 1.41 0.255 Treatment x time 5 0.55 0.738 Length of leaf Treatment 1 4.99 0.035 Time 5 0.79 0.565 Treatment x time 5 0.06 0.997 Depth of water Treatment 1 3.59 0.080 Time 5 5.01 0.057 Treatment x time 5 8.37 0.025
|Printer friendly Cite/link Email Feedback|
|Author:||Titus, Priscilla J.; Titus, Jonathan H.|
|Date:||Dec 1, 2008|
|Previous Article:||Evaluation of predator attendance at pitfall traps in Texas.|
|Next Article:||Population trends of breeding birds on the Edwards Plateau, Texas: local versus regional patterns.|
|The upper San Pedro partnership.|
|Seedbank of Bingham cienega, a spring-fed marsh in southeastern Arizona.|