Costs of reproduction in flowering plants

Abstract

Costs of reproduction arise when investments into current reproduction reduce future reproductive fitness. Studies on reproductive costs use diverse approaches, including the analysis of gene expression, physiology, trade-offs between reproduction and growth/survival, and the impact of reproductive investments on population growth. These studies demonstrate that reproductive trade-offs have far-reaching effects on plants, affect their fitness, and are therefore important for shaping the evolution of life histories. However, not all studies have detected costs of reproduction, and c. 90% of these were conducted in natural populations, where controlling for variation in plant resource status is challenging. For dioecious plants, there is a common perception that fruit production should result in greater costs of reproduction for females than males, but divergent reproductive costs between the sexes are not supported by studies of reproductive trade-offs in dioecious plants. Other aspects of reproductive costs remain poorly understood, including ecological costs of reproduction, the fitness effects of reproductive trade-offs involving growth or physiological processes, and how the male sex role influences reproductive costs. Progress will be enabled by the use of measurements that allow for easier comparisons across studies and by more clearly distinguishing between the processes that contribute to current vs future reproductive fitness.

Keywords: cost function; demographic costs of reproduction; life‐history trade‐offs; reproductive costs; sex‐allocation theory; somatic costs of reproduction.

Fig. 1

Costs of reproduction are understudied in the most diverse plant families. The dotted line indicates the null expectation if the number of species studied for reproductive costs per family was proportional to the number of species per family. The dashed line indicates the null expectation if the number of species studied was proportional to the number of angiosperm species.

Fig. 2

Results of a logistic regression used to examine whether the ability to detect costs of reproduction in plants varies across contexts and approaches. Studies of the costs of reproduction in natural populations, and studies of entire plants, more commonly reported that costs were not detected than did studies that were conducted in controlled environments or studies of smaller subunits within plants, for example branches. The frequency with which costs of reproduction were reported from observational (vs experimental) studies, woody (vs herbaceous) plants, or from studies of plants or within (vs between) flowering seasons, did not vary significantly. Significant parameter estimates are indicated by red (filled) symbols. Values are the logistic regression parameter estimates ± SE. Parameter estimates and other details of the logistic regression are reported in Supporting Information Table S1.

Fig. B1

(a) Association between reproductive allocations and fertility can have various shapes. (b, c) Costs of reproduction depend on the association between the allocation of resources to reproduction and the effect of that allocation on expected future reproductive fitness (i.e. how allocations affect residual reproductive values). Because the two sex roles are thought to be characterized by different gain curves, we expect a different cost–benefit association between the sexes. The net benefit depends on the strengths of the trade‐offs arising from reproductive allocations and can be linear (blue), nonlinear concave‐down (orange), or nonlinear convex‐down (not shown; Sletvold & Ågren, 2015). If allocations to the female and male sex roles arise from a common resource pool, the shapes of the trade‐off functions are expected to be similar for the two sex roles. However, if the resources used for the female sex role differ from those used for the male role, the cost curves might also differ between the sex roles, increasing the scope for divergence in reproductive allocations between the sex roles. Note the difference between reproductive allocations and costs, at least for hermaphroditic plants. For example, if allocations to the male sex role are associated with a linear cost function, only small allocations to reproduction are justified (under the portion of the male gain curve with a steeper slope than that depicted for the linear trade‐off function). By contrast, if allocations to the female sex role are associated with a nonlinear cost function, a broad range of allocations yield greater fitness benefits than costs.

Fig. B2

Costs of reproduction have been assessed across hierarchical levels of biological organization. Studies of processes associated with lower levels of biological organization are typically associated with estimates of the somatic costs of reproduction. By contrast, processes associated with higher levels of biological organization are typically associated with estimates of the demographic costs of reproduction.