Abstract. Scientists were concerned with the division of labor research early, and much current research in clonal plant ecology has proved environmentally induced or a developmentally programmed division of labor. However, the sexually dependent division of labor in clonal plants has received little attention, and few experimental tests have been conducted to date. Here, using reproductive ramets and non-reproductive ramets in the clonal plant Iris laevigata, we test this hypothesis on morphology and physiology through field and manipulative experiments, including an isotope ( 15 N) labeling method. With regard to morphology, under natural conditions, non-reproductive ramets of I. laevigata were more abundant, taller, and consisted exclusively of leaves, while the relatively rare, smaller reproductive ramets consisted of leaves, flowers, or seeds. However, the mean biomass of non-reproductive ramets was nearly half that of reproductive ramets. With regard to physiology, non-reproductive ramets had a significantly higher net photosynthetic rate over a 24-h period than reproductive ramets. Furthermore, the reproductive ramets stopped leaf growth when flower buds appeared in late May and lost photosynthetic capacity before late June, while its growing fruits still required a constant supply of many nutrients. Manipulative experiments showed that the non-reproductive ramets supplied the nutrient demands for reproduction of the reproductive ramets through clonal integration as high as d 15 N of the value (196.1 AE 73.39) of the seed. Thereby providing the first clear evidence of sexually dependent division of labor in clonal plants: Certain individuals constitute a very small amount of reproductive castes (reproductive ramets), whereas a large amount of other individuals (non-reproductive ramets) are "workers" as in social insect societies whose role is primarily as producers. The rapid withering of reproductive ramets that failed to be pollinated highlights their role as irreversibly focused on reproduction, and those unable to reproduce do not contribute to the colony. Additionally, "producer" ramets positively influenced reproductive castes growth, and the division of labor was assumed to underlie the resource allocation pattern among ramets.