The results suggest that pollinator discrimination negatively affects females' relative fitness when they are rare. Thus, the initial spread of females in a population, the first step in the evolution of gynodioecy, may be made more difficult due to pollinator discrimination.

Although fitness costs associated with plant defensive traits are widely expected, they are not universally detected, calling into question their generality. Here, we examine the potential for life-history trade-offs associated with herbicide resistance by examining seed germination, root growth, and above-ground growth across 43 naturally occurring populations of Ipomoea purpurea that vary in their resistance to RoundUp®, the most commonly used herbicide worldwide. We find evidence for life-history trade-offs associated with all three traits; highly resistant populations had lower germination, shorter roots, and smaller above-ground size. A visual exploration of the data indicated that the type of trade-off may differ among populations. Our results demonstrate that costs of adaptation may be present at stages other than simply the production of progeny in this agricultural weed. Additionally, the cumulative effect of costs at multiple life cycle stages can result in severe consequences to fitness when adapting to novel environments.

Plant species with separate genders often exhibit gender differences in traits related to reproductive allocation. In gynodioecious species, females often produce more seeds than do hermaphrodites, leading to a higher reproductive cost. The mechanisms that allow females to meet the high costs of reproduction are currently under debate. In this study, we test the hypothesis that there are genetically based gender differences in physiological traits that enable females to finance these costs through higher photosynthetic carbon gain in the gynodioecious perennial Geranium maculatum. Females and hermaphrodites were compared in a greenhouse study that minimized environmental and selfing rate differences between the genders. We found that females produced smaller flowers but more of them and more fruits than did hermaphrodites. However, genders did not differ in their seed number, seed mass, fruit set, and reproductive allocation. In addition, genders did not differ in photosynthetic rate (A), leaf N, and water use efficiency inferred from leaf carbon isotope ratio (d 13 C). Overall, G. maculatum shows no genetically based gender differences for most of the reproductive traits or any of the physiological traits measured. Our results suggest that for G. maculatum, the gender fitness differences previously identified in natural populations may be caused by gender differences in microhabitat and/or selfing rate.

13 14 Total word count: 5690 15 Introduction word count: 1242 16 Materials and Methods word count: 1163 17 Results word count: 912 18 Discussion word count: 2325 19 Acknowledgements word count: 48 20 Number of figures: 4 21 Number of tables: 0 22 Supporting information: 3 tables 23 24 This article contains supporting information online at: xxx. 25 2 Summary 26• Although fitness costs of adaptation to herbicide are widely expected, they are not 27 universally detected, calling into question their generality. This lack of identified 28 costs could be due to experimental constraints such as a limited focus on 29 measures of plant fecundity and low replication. 30• Here we examine the potential for a cost of herbicide resistance on seed 31 germination, root growth, and above-ground growth using 43 naturally occurring 32 populations of Ipomoea purpurea that vary in their resistance to RoundUp®, the 33 most commonly used herbicide worldwide. 34• We show substantial costs in all three traits. Highly resistant populations had 35 lower germination rates, slower root growth and slower above-ground growth. A 36 visual exploration of the data indicated that the type of cost may differ and even 37 trade off among populations. 38• We place our findings into a broad context using a simple model to show that the 39 strength of the germination cost could act to slow the rate of resistance evolution 40 in this species. Our results demonstrate that costs of adaptation may be present at 41 stages other than the production of progeny. Additionally, the cumulative effect of 42 costs at multiple life cycle stages can result in severe consequences to fitness 43 when adapting to novel environments. 44 45

No abstract