We measured population size and density of Pyrgulopsis morrisoni and Heterelmis sp. within a single spring in central Arizona over four sampling periods in 2001 to evaluate the effect of sampling without replacement. Our analysis detected significant differences in total population size across sampling periods. Sampling without replacement caused a transitory decline in total population size of each organism, though P. morrisoni was again locally abundant the following year. Spring ecosystems are affected by several anthropogenic stressors and many endemic aquatic invertebrates have been afforded Federal protection. Studies should not contribute stress. Until more is known about fecundity, recruitment, and population fluctuations, employ sampling methods that do not remove significant numbers of individuals.
Introduction
Resource professionals in the American Southwest have begun focusing significant attention on spring-endemic organisms such as springsnails (Hydrobiidae) and riffle beetles (Elmidae). These aquatic organisms are often unique and exhibit a high rate of endemism because isolation in spring ecosystems provides ideal situations for rapid evolution, causing significant character drift and speciation (Shepard 1992, 1993).
A variety of sampling techniques have been employed, but little is known about effects, especially on isolated populations of endemic species. In 2001, a study was initiated to evaluate habitat use of the Page springsnail (Pyrgulopsis morrisoni; Taylor 1987, Martinez and Thome 2006). During that study, sufficient data was gathered for the Page springsnail and a riffle beetle (Heterelmis sp.; Shepard, pers. comm.) from a single spring to evaluate the effect of sampling without replacement. These are small aquatic organisms, <3 mm in size (Brown 1983, Taylor 1987).
Cave Spring is located in central Arizona near Oak Creek and the town of Page Springs on a State fish hatchery (Fig. 1). Water emanates below a west-facing slope and drains into a culvert that transports it into an underground collection gallery where it is used in hatchery operations. Water depths range from 4-11 cm and substrate is dominated by gravel, pebble, and woody debris; silt, sand, cobble, and aquatic macrophytes are also present.
Since little is known about the effect of sampling without replacement on populations of spring-snails and riffle beetles, the objective of our study was to assess differences in total population sizes between sampling periods for P. morrisoni and the Heterelmis sp.
Methods
We measured population size and density of P. morrisoni and Heterelmis sp. in Cave Spring over four sampling periods in 2001. Eight modified Hester-Dendy artificial substrate samplers (Martinez and Thome 2006) were placed randomly within the aquatic environment of the spring to quantify invertebrate densities. Sampling periods were: March 23 to May 10, May 11 to June 21, June 22 to August 2, and August 3 to September 25. At the end of each period, aquatic invertebrates were placed into Whirl-Paks, preserved with alcohol, and transported to the lab for identification and enumeration under a Stereozoom 7 Microscope.
Population and density estimates were calculated based on the total counts on sample plots methodology (Cochran 1977, Seber 1982, Lancia et al. 1996). Total population size and variance was calculated as:
and where N=total population size, s=number of randomly selected sample plots on which counts are made, A=total area occupied by the population, a=area of each sample plot, S=A/a=total number of potential sample plots in A from which the s plots are selected, x=number of animals counted per sample plot. Density and variance was estimated as: and The total area occupied by the population was calculated with the line intercept method (Lehman 1991) based on the wetted surface area of the spring. The approximate area of the plot, A, was calculated as: where i=the lengths of the lines intercepted by the area of the plot, L=the perimeter of a rectangle completely enclosing the plot, and r=the area of the rectangle.
We tested the null hypothesis that there were no differences in population sizes between sampling periods. Testing was conducted with CONTRAST, which uses a general Chi-square statistic to test differences among abundance estimates using contrasts (Hines and Sauer 1989, Sauer and Williams 1989).
Results
We collected 1,776 P. morrisoni and 473 Heterelmis sp. from Cave Spring. The total area occupied by the populations (S) was 2.055 m2 and the area of each sample plot (a) was 330.857 cm2. Population size, density, standard error, 95% confidence interval, and sample size for each sampling period were calculated for P. morrisoni (Table 1) and Heterelmis sp. (Table 2) and are presented by month data collected.
Table 1.
P. morrisoni population size and density statistics within Cave Springs, central Arizona, 2001.
Table 2.
Heterelmis sp. population size and density statistics within Cave Springs, central Arizona, 2001.
The overall test for homogeneity detected significant differences in total population sizes for P. morrisoni (χ2=7.8608, df=3, P=0.0490; Fig. 2) and Heterelmis sp. (χ2=8.9584, df=3, P=0.0298; Fig.3), between sampling periods.
Discussion
Our analysis showed a stark decline in the total population size of both P. morrisoni and Heterelmis sp. within a single spring in response to sampling without replacement during 2001. We believe this decline was a direct consequence of removing a significant number of individuals from a relatively closed system. Isolated populations of Hydrobiids and Elmids are most likely sustained primarily by births rather than immigration since their dispersal capability is minimal. Aquatic snails can disperse by becoming attached to waterfowl and shorebirds (Dundee et al. 1967), but this mechanism is stochastic. Adult riffle beetles commonly fly after pupation, but such dispersal flights are usually short due to the beetle's small size (Brown 1983, 1987). Also, our study period was short, further limiting the potential effect of immigration.
Monitoring surveys conducted in 2002 detected remarkably high density of P. morrisoni, showing the species was again locally abundant the following year (Sorensen et al. 2002). For instance, using tile samplers in June 2002 they estimated a density of 0.1231 springsnails per cm2. Those authors concluded that collections in 2001 had only a temporary effect on the springsnail and the species appeared resilient to disturbance and reductions in abundance. We suspect this resiliency may be related to fecundity and recruitment rates, for which there is scant information available for springsnails and riffle beetles. The abundance of springsnails the year following our sampling effort may indicate high rates of fecundity and recruitment. It is also possible these organisms naturally exhibit substantial fluctuations in population numbers. Though sampling without replacement caused a transitory reduction in populations of P. morrisoni and Heterelmis sp., we cannot determine at what point it would preclude repopulation, if at all. More information regarding fecundity, recruitment rates, and natural population fluctuations would be useful.
As of November 2004, 32 species of snails (aquatic and terrestrial), 70 species of clams, and 21 species of crustaceans receive Federal protection under the Endangered Species Act within the United States (US Fish and Wildlife Service 2004). Many are endemic organisms limited in distribution to environments such as springs. Spring ecosystems are affected by several anthropogenic stressors including livestock grazing, water consumption, water diversion, contaminants, recreation, exotic species, spring manipulation, and wildland fire (Williams et al. 1985, Shepard 1993, Myers and Resh 1999, Lang 2002, Sada et al. 2005). Ecological studies should not contribute additional stress.
Isolated populations of endemic aquatic invertebrates may be resilient to transitory population declines. However, until more is known about fecundity, recruitment rates, and natural population fluctuations, it may be prudent to employ sampling methods that do not remove significant numbers of individuals. This is consistent with a study of San Bernardino springsnail (P. bernardina; Hershler and Landye 1988) whose authors recommended sampling with live release during reproductive seasons (Malcom et al. 2003).
Acknowledgments
Views expressed in this manuscript are the authors and do not necessarily reflect those of the US Fish and Wildlife Service. We thank the Arizona Game and Fish Department for access to the study area. We thank the following individuals for their contribution: L. Allison, W. Chen, T. Gatz, K. Goodhart, J. Graves, S. Hatten, C. Marr, W. Shepard, A. Tavizon, and D. Thome.
