Effect of threadfin shad density on production of threadfin shad and gizzard shad larvae in a Tennessee reservoir

Production of larvae by threadfin shad, Dorosoma petenense, and gizzard shad, D. cepedianum, varied over two orders of magnitude and was regulated by adult threadfin shad abundance over five years in Normandy Reservoir, Tennessee. Significantly more larvae of both species were produced in years following winterkills of threadfin shad (repeated-measures ANOVA, df=4, 75; F > 21.44, p=0.0001). Peak geometric mean catch of threadfin shad larvae in neuston samples was inversely related to biomass (kg ha^–1; r = – 0.91; p=0.031) and density (no. ha^–1; r=– 0.89; p = 0.043) of adult (> 70 mm total length) threadfin shad in mid-summer cove samples. Peak geometric mean catch of gizzard shad larvae was also inversely related to adult threadfin shad biomass (r = – 0.93; p=0.022) and density (r=– 0.88; p=0.046) in cove samples. Winterkills of threadfin shad were size selective, killing all fish under 60 mm total length but allowing some unknown percentage of larger fish to survive. When threadfin shad stocks were reduced by winterkills, surviving threadfin shad and gizzard shad may have taken advantage of less competition for food resources in early spring and increased condition enough to spawn successfully.