Previous studies reported skin tumors diagnosed as pigmented subcutaneous spindle‐cell neoplasms in 22% of gizzard shad (Dorosoma cepedianum) from Lake of the Arbuckles, Oklahoma, USA. Those studies could not confirm chemical carcinogens or retroviruses as etiological agents. The present study reports the neoplasms in 20% of shad from two additional lakes, Lake Murray and Lake Texoma in south‐central Oklahoma, extending the range of the lesion. No neoplasms were found in shad from a reference site, Lake Carl Blackwell, Oklahoma. Further investigations into the etiology of the lesions were conducted. Significant levels of potentially carcinogenic trace elements in the water, sediment, or tissues were not identified by inductively coupled plasma mass spectrometry. Radioactivity, analyzed by liquid scintillation counting of radon and gross alpha/beta radiation, was not above background levels. Genetic marker and band‐sharing analysis by random amplified polymorphic DNA and double‐stringency polymerase chain reaction could not separate tumor‐bearing shad from nontumor‐bearing ones. Of 2,128 shad examined, 387 exhibited lesions, with a significantly higher number occurring dorsally (79.5%) than ventrally (20.5%). Overall, this study showed the epizootic is not limited to a single lake and tended to rule out some known carcinogens and radioactivity as proximate causes of the epizootic.

Abstract-Previous studies reported skin tumors diagnosed as pigmented subcutaneous spindle-cell neoplasms in 22% of gizzard shad (Dorosoma cepedianum) from Lake of the Arbuckles, Oklahoma, USA. Those studies could not confirm chemical carcinogens or retroviruses as etiological agents. The present study reports the neoplasms in 20% of shad from two additional lakes, Lake Murray and Lake Texoma in south-central Oklahoma, extending the range of the lesion. No neoplasms were found in shad from a reference site, Lake Carl Blackwell, Oklahoma. Further investigations into the etiology of the lesions were conducted. Significant levels of potentially carcinogenic trace elements in the water, sediment, or tissues were not identified by inductively coupled plasma mass spectrometry. Radioactivity, analyzed by liquid scintillation counting of radon and gross alpha/beta radiation, was not above background levels. Genetic marker and band-sharing analysis by random amplified polymorphic DNA and double-stringency polymerase chain reaction could not separate tumor-bearing shad from nontumor-bearing ones. Of 2,128 shad examined, 387 exhibited lesions, with a significantly higher number occurring dorsally (79.5%) than ventrally (20.5%). Overall, this study showed the epizootic is not limited to a single lake and tended to rule out some known carcinogens and radioactivity as proximate causes of the epizootic.