Cottus Life History in Delaware 79 



sandstone (both usually flat and concave beneath). Both occur naturally 

 in this part of the Delmarva Peninsula, but the large local concentration 

 was probably dumped there for old construction. Only two clusters 

 (5.1%) were found on naturally-occurring material (log and branch). The 

 general absence of rock or other suitable spawning sites in Delmarva 

 Peninsula streams may limit C. bairdi distribution. 



Four substrate samples from these two gravel riffles yielded particles 

 that ranged in size from greater than 2.5 cm to smaller than 0.05 cm; 

 about half the particles (by weight) were in the size category 0.8 to 2.5 cm 

 (Table 2). By contrast, about half the particles (by weight) in three 

 samples from three gravel patches in Horse Pen Branch were 0.05 cm or 

 smaller (Table 2). The largest pieces of gravel collected in the sampling 

 area riffles measured (maximum length and width, cm) 7.5 X 3.9, 6.1 X 

 3.5, and 5.9 X 4.2, while the largest from Horse Pen Branch were 2.2 X 

 1 .7, 2.3 X 1 .4, and 2.0 X 1 .4. Cottus bairdi was much more common over 

 the riffles of larger gravel, perhaps because under natural conditions 

 these facilitate spawning. 



We consider 32 of the 39 clusters to represent the spawn of 1 female, 

 3 the spawn of 2 females, and 4 the spawn of 3 females. Ludwig and 

 Norden (1969) noted a mean of 3.3 spawns per cluster. Koster (1936) 

 observed up to six clusters in a nest, but noted that most nests contained 

 from one to four egg masses. Clusters resulting from multiple spawnings 

 can be identified by their larger size, varied egg mass colors, and irregular 

 configuration. There were four instances of two single spawns found 

 under one piece of substrate (each cluster was counted as the spawn of 

 one female), and several cases of single- and multiple-spawn clusters on 

 substrate with multiple-spawn clusters. Thus, the result of a total of 50 

 spawnings was found, deposited on a total of 31 pieces of substrate. 



Color of live eggs was usually pink-orange, but the range in color of 

 clusters was from orange to yellow to tan, and one cluster was almost 

 white. About half of the clusters contained one to several white eggs, 

 presumed to be unfertilized. Most clusters were approximately circular in 

 outline, with eggs arranged in irregular rows, usually to four or five layers 

 deep. They adhered strongly to the substrate and to each other. Egg 

 cluster mean length, width, and height, respectively, and the range of 

 each (in mm), for the total sample sizes given previously were: single 

 spawn, 24.8 (16-37), 20.6 (13-29), 8.1 (6-12); two spawns, 40.0 (28-45), 

 24.3 (22-26), 8.6 (7-10); and three spawns, 48.2 (40-56), 30.0 (28-32), 1 1.7 

 (8-19). 



Egg number in 1 1 single-spawn clusters ranged from 64 to 201, with 

 an egg diameter range (measured to nearest 0.1 mm) of 2.2 to 3.5 mm, 

 and a mean egg diameter per cluster range of 2.32 to 3.09 mm; egg 

 number for one triple-spawn cluster was 623, with a range in egg diameter 



