SHOOT DEMOGRAPHIC RESPONSES TO MANIPULATION OF REPRODUCTIVE EFFORT BY BUD REMOVAL IN A WILLOW

Willow reproductive allocation was manipulated by removing shoot or inflorescence buds in winter, thus altering the ratio of inflorescences to developing shoots while removing the same number of buds in each treatment. When responses were measured 3 years later, shoot and bud populations had been increased by the shoot-bud-removal treatment; shoot lengths did not respond. The long term effects on crown shoot population and fecundity can be explained by an increase of shoot numbers in the summer after bud removal. The increase was owed either to spring recruitment of buds that would normally be dormant or to autumn retention of new shoots that would normally be shed. Shoot-bud removal elicited over compensatory shoot population growth the same year, which increased shoot population and reproductive effort 3 years later. Inflorescence bud removal increased shoot population growth after a year's delay, possibly by a delayed effect on shoot bud numbers in the second year. The relation between size versus number of shoots and that between shoot length and number of inflorescence buds produced by a shoot were unaffected in later years. The experiment did not reveal any substantial growth-cost or module-substitution-cost of reproduction, but did demonstrate an ability to compensate shoot bud losses by increased recruitment and survival of new shoots. The two treatments should have responded in opposite directions if manipulations altered the growth:reproduction allocation ratio; their similarity of response suggests that bud removal is not a valid method of testing the cost hypothesis in plants with arborescent or frutescent architecture.