“…The results of these studies have been mixed. Many have shown that character correlations are not stable across environments (Adams, 1967;Primack and Antonovics, 1981;Antonovics and Primack, 1982;Singh and Chowdhury, 1983;Schlichting 1986Schlichting , 1989aAriyo et al, 1987;Lechowicz and Blais, 1988;Andersson, 1989;Pigliucci et al, 1991Pigliucci and Schlichting, 1998;Callahan and Waller, 2000;Pigliucci and Marlow, 2001). Other studies support the hypothesis of character integration: they have shown stability of correlations (Winn and Evans, 1991;Meerts, 1992;Kudoh et al, 1996), that putatively functionally or developmentally related traits are more highly correlated across environments than other traits (Waitt and Levin, 1993); and that within populations, those individuals with the greatest integration of characters across environments have the highest fitness (Schlichting, 1989b).…”