Conservation Physiology of the Plethodontid Salamanders Eurycea nana and E. sosorum: Response to Declining Dissolved Oxygen

Eurycea sosorum and E. nana are plethodontid salamanders endemic to several karst springs in central Texas. Landscapes around these habitats are increasingly urbanized. At the Barton Springs complex, where E. sosorum occurs, average dissolved oxygen (DO) in the main flow is approximately 6.5 mg L⁻¹. However, DO is quite variable, ranging between 2.4 and 10 mg O₂ L⁻¹, and recent data suggest a positive relationship between DO and spring discharge in Barton Springs Pool, though this relationship may not be as strong under extreme low-flow conditions. Here we examine sensitivity of a surrogate species, E. nana, to experimental variation in oxygen availability (DO); due to limited availability of E. sosorum, they were examined in only a subset of experiments. A suite of traits was measured on adults: spontaneous activity, metabolic rate, and mortality during 28 days of exposure. A separate experiment examined growth of juveniles across levels of DO during 60 days of exposure. Levels of DO below 3.4 mg O₂ L⁻¹ appeared to pose a grave threat to salamander survival over a 28-day study, whereas DO above 4.5 mg O₂ L⁻¹ gave no observable effects in any experiment. Between these values is a critical range in which salamanders became progressively compromised. An ambient water quality criterion for DO in lentic systems (5 mg O₂ L⁻¹, 24 hour minimum) appears adequate to protect Eurycea.

Fig. 1

Experimental set-up for the 28-day oxygen toxicity experiments. Each of 20 controlled atmosphere boxes held three aquaria (one salamander per aquarium); only two aquaria are shown in the figure.

Fig. 2

Example of the log survivorship analysis of activity for one of the salamanders showing (A) the location of the breakpoint at 0.82 min between activity bouts and (B) raw voltage trace from infrared activity meter with activity bouts drawn above according to the breakpoint identified in (A).

Fig. 3

Spontaneous activity of Eurycea nana in response to ramped dissolved oxygen. (A) Raw voltage traces and fitted bouts for each of eight salamanders and a blank chamber superimposed on the trace of dissolved oxygen. (B) Dots are total number of salamanders active (out of eight), and the line is a fitted loess curve (local regression, with smoothing, smoothing parameter = 0.3).

Fig. 4

Metabolic rates of Eurycea nana (A) and E. sosorum (B) across ramped levels of dissolved oxygen (DO). Lines represent best fits of the Biological Oxygen Demand model (Eq. 3). See Table 3 for summaries of parameter values and statistical significance.

Fig. 5

Percent mortality of Eurycea nana exposed to varying dissolved oxygen content.

Fig. 6

Percentage rank and log-transformed plot for a distribution of discrete dissolved oxygen measurements for Barton Springs •, Eliza Spring ○, and Old Mill ▾ locations in central Texas. The corresponding correlation coefficients for the regression lines fitted to each sampling site are 0.97, 0.96, and 0.65, respectively. Vertical reference lines represent the LC₅₀ (3.4 mg L⁻¹), LC₂₅ (3.7 mg L⁻¹), LC₁₀ (4.2 mg L⁻¹), LC₅ (4.5 mg L⁻¹), and NOAEL (4.4 mg L⁻¹), respectively, for 28 d adult mortality and 60 d juvenile specific growth rates of Eurycea nana exposed to varying dissolved oxygen concentrations.

Contributor Notes

¹ Division of Biological Sciences, University of Montana, Missoula, Montana 59812; E-mail: art.woods@mso.umt.edu. Send reprint requests to this address.

² Section of Integrative Biology, The University of Texas at Austin, Austin, Texas 78712; E-mail: (MFP) mpoteet@mail.utexas.edu; and (PDH) phitchings@hotmail.com

³ Department of Environmental Science, Center for Reservoir and Aquatic Systems Research, Baylor University, Waco, Texas 76798; E-mail: (RAB) richard_brain@baylor.edu; and (BWB) bryan_brooks@baylor.edu

Associate Editor: E. Schultz.

Received: 02 Feb 2009

Accepted: 19 May 2010