Costs of Reproduction in a Willow: Experimental Responses Vs. Natural Variation

Abstract

We report on the interaction between reproduction and subsequent reproductive and vegetative growth in female plants of a dioecious, early successional, tall shrub, the feltleaf willow (Salix alaxensis). We measured natural variation among individuals in flowering, and its correlation with subsequent shoot growth and flowering. We also experimentally curtailed reproductive allocation in two ways: (1) by removing inflorescence buds in April when plants were dormant, and (2) by covering inflorescences to prevent pollination and fruit development. We hypothesized that these treatments would cause manipulated plants to retain more resources and to deploy them subsequently in shoot growth or reproduction. However, treatments had no significant effect upon shoot elongation, which decreased slightly after floral buds were removed. Treatments did significantly increase the fraction of buds differentiating as inflorescences instead of shoots. We compared the experimental outcome with the natural, phenotypic variation among plants. Among plants of either the treated or the unmanipulated groups, there was a significant negative correlation between current reproduction and subsequent growth; the corresponding regressions did not differ between the two groups. Thus, natural phenotypic variation suggested a tradeoff between reproduction and growth, but a direct test by experiment revealed only a tradeoff between current and future reproduction. Furthermore, the negative correlation seen in unmanipulaed plants was unaffected by treatments curtailing reproductive investment. We conclude that this correlation is not directly related to allocation contraints, and does not provide reliable evidence about costs of reproduction. The increased commitment to flowering that occurred after experimental reduction of current reproductive investment may be evidence for a cost of reproduction, although the response need not have been caused directly by resource allocation constraints.