Neighbour effects on gender variation in Ambrosia artemisiifolia J.T. LUNDHOLM AND L.W. AARSSEN' Department of Biology, Queen's University, Kingston, ON K7L 3N6, Canada Received October 13. 1993 LUNDHOLM, J.T., and AARSSEN, L.W. 1994. Neighbour effects on gender variation in Ambrosia artemisiifolia. Can. J. Bot. 72: 794-800. In this study we tested the prediction that male gender allocation in natural populations of common ragweed (Ambrosia artemisiifolia L.) depends more on relative plant height within the neighbouring canopy than on absolute plant height. This is consistent with the assumption that success as a pollen donor for an anemophilous plant within a crowded population will be greater when neighbours are shorter than when neighbours are taller. Data on height and proximity of neighbours, irradi- ance, and target plant height, biomass, and gender allocation were collected from two natural populations. In one population, these data were also recorded for a group of target plants that had local neighbours artificially removed when the target plants were seedlings. Allocation to male flowers was most strongly positively correlated with height relative to that of close neigh- bours and with percent irradiance in natural populations. Numerous all-female plants were recorded among the smallest individuals that were suppressed by a dense overhead canopy of neighbours. No relationships were found between plant size and gender when plants were taller than their neighbours or when neighbours were artificially removed. Hence, previously reported size-dependent gender variation in this species may depend on the presence of neighbours. It is postulated that rag- weed individuals may sense the presence of neighbours through the phytochrome system, and that effects of neighbours on light quantity and quality cues a shift to increased female function. However, plants that grew from the seedling stage without neighbours were heavier and more female but were not taller than plants with neighbours left intact. The interpretation of this effect is unclear but may reflect a change in plant architecture corresponding with the removal of neighbours. Key words: gender, plant height, plasticity, pollen dispersal, phytochrome, shading. LUNDHOLM, J.T., et AARSSEN, L.W. 1994. Neighbour effects on gender variation in Ambrosia artemisiifolia. Can. J. Bot. 72 : 794-800. Dans cette Ctude, les auteurs vCrifient l'hypothese que l'allocation au gendre mile dans les populations naturelles d'herbe i poux (Ambrosia artemisiifolia L.) serait plus reliee i la hauteur relative des plants dans la canopCe avoisinante, qu'i la hauteur absolue des plants. Ceci concorde avec l'hypothese que, chez une plante ankmophile, le succks comme donneur de pollen i llintCrieur d'une population nombreuse sera meilleur lorsque les voisins sont plus courts que lorsqu'ils sont plus hauts. Les auteurs ont rCcoltC des donnCes sur la hauteur et sur la proximitC des voisins, sur I'irradiance, sur la hauteur des plantes cibles, sur l'allocation en termes de biomasse et de gendres, dans deux populations naturelles. Dans une population, ils ont Cgalement enregistrk ces donnCes pour un groupe de plantes cibles autour desquelles ils ont artificiellement ClirninC les voisins, au moment oh les plantes cibles Ctaient i 1'Ctat de plantules. L'allocation aux fleurs mdles montre une trks forte corrClation avec la hauteur relative par rapport aux voisins immCdiats ainsi qu'avec le pourcentage d'irradiance dans les popu- lations naturelles. On observe de nombreuses plantes entikrement femelles chez les individus les plus petits qui ont CtC sup- primCs par une dense canopCe formCe par les plantes avoisinantes. On ne trouve pas de relations entre la dimension de la plante et le genre, lorsque les plantes sont plus hautes que les voisines, ou quand les voisines sont artificiellement supprimkes. ConsCquemment, chez cette esptce, les variations dCji rapportCes dans les genres en fonction de la dimension pourrait dependre de la presence des voisins. Les auteurs tmettent le postulat que les individus de l'herbe i poux peuvent percevoir la prCsence des voisins via le systkme phytochrome et que les effets des voisins sur la quantitC et la qualit6 de la lumikre servent de signal pour favoriser la fonction femelle. Cependant, les plantes qui poussent sans voisins i partir du stade plantule ont des poids plus ClevCs et ont plus de fleurs femelles, mais ne sont pas plus hautes que les plantes autour desquelles les voisins sont demeurCs intacts. L'interprCtation de cet effet n'est pas clair mais reflkte un changement de l'architecture de la plante correspondant i 1'Climination des voisins. Mots elks : genre, hauteur des plants, plasticitC, dispersion du pollen, phytochrome, ombrage. [Traduit par la rCdaction] Introduction Labile sex allocation in plants is viewed as an adaptive strategy to maximize fitness of an individual when the relative fitness returns from male and female functions vary with envi- ronmental conditions (Charnov 1982; Lloyd and Bawa 1984). The factors that cause variation in the relative opportunities for female contributions to reproduction as opposed to male ones may differ between species. One sex may be favoured over the other if there is local competition for mates; if one sex predominates in the surrounding area, an individual might maximize its fitness by allocating to the other sex (Charnov and Bull 1977). The risk of predation may also be patchy and could affect fitness if structures of either sex are favoured by 'Author to whom correspondence should be addressed Prinlcd in Canada 1 lmprimd au Canada predators (Charnov and Bull 1977). Other environmental parameters that could influence sex allocation include temper- ature, soil moisture, light and nutrient availability, and popu- lation density (Charnov and Bull 1977; Traveset 1992). These may all influence plant size and are likely to result in the reported sex allocation patterns that seem to vary primarily with various measures of plant size (Abul-Fatih et al. 1979; Burd and Allen 1988). In several species, conditions favouring growth promote femaleness, while stress promotes maleness (Charnov 1982; Lloyd and Bawa 1984; Solomon 1989). Studies show that this stress can be in the form of conditions that reduce growth, individual chances for survival, or the plant's ability to allo- cate resources to reproduction (Charriov 1982). In numerous species, moisture stress is accompanied by a corresponding shift in sex expression from female to male (Freeman et al.