Salamander Tolerance of Acidity 
39 
controls. All water was filtered through cheesecloth before eggs were 
added. All tests were carried out in a controlled temperature (15°C) 
cabinet at a photoperiod of 12L:12D. Oxygen levels in the aquaria 
remained high (i.e., near saturation) through the tests and were monitored 
periodically with an electronic oxygen meter. 
Two egg masses were placed in each aquarium. Although we 
measured masses having a volume greater than 200 cm 3 and containing 
more than 200 eggs, the volume of masses in our tests was 65.2 ± 10.2 
cm 3 (mean ± one standard error). These contained 57.9 ±0.8 eggs /mass. 
Experimental groups included tests of survival of eggs in pond 
water with acid added (either H 2 S0 4 or HNO3), or with aluminum 
added. Aluminum used in all tests was in the form of A1 (S0 4 ) 2 . Acid 
from a concentrated stock solution was added in tiny amounts until a 
pH of 3.0 to 4.0 was obtained. The buffering capacity of individual egg 
masses then caused acidity to change. This resulted in pH values of 3.1 
to 4.0 (see Table 2). Aluminum sulfate was added in amounts sufficient 
to produce 0.05, 0.10, or 0.23 ppm aluminum in the solution. Controls 
were performed in exactly the same manner as experimental groups. In 
addition, we tested survival of eggs in local spring water having a 
slightly higher pH and little or no dissolved organic materials (pH = 
4.65 ± 0.30; N = 5). 
Egg masses were inspected daily, and embryos were classified 
according to Harrison stage (Pough 1976; modification of Rugh 1962). 
Mortality of salamander eggs may be a function of shrinkage of the 
gelatinous matrix surrounding the eggs. Accordingly, the length and 
width of each egg mass was measured with a plastic ruler at the 
beginning of each test, and at 3, 6, 14, and 25 days. Volume of 
individual masses was then computed from the prolate spheroid formula: 
V = 0.523 AB 2 , where A is the length of the mass (cm), B is the width, 
and V is the volume in cm . All free-swimming larvae were removed 
within 48 hours of hatching, examined for morphological defects, and 
released into a central holding tank. No long-term tests of mortality of 
larvae were performed; references to larval survival in the present paper 
are valid only for the 24 hours after hatching. Most surviving larvae 
were returned to the breeding pond. 
RESULTS 
All temporary ponds (64) whose pH was known were acidic, and 
the majority (63/64 = 98.4%) had pH values during the spring Ambystoma 
breeding season that were below levels known to cause mortality (i.e., 
pH < 6.0; see Pough 1976, Pough and Wilson 1977, Cook 1983). A 
small percentage (5/64 = 7.8%) had pH’s of less than 4.0 (Table 1). Of 
the total of 168 ponds that were surveyed, only a few (12/ 168 = 7.1%) 
