A book titled Sex, Size and Gender Roles certainly grabs the attention of potential readers—even a fellow passenger on an airline flight seemed interested—but the volume's subtitle, Evolutionary Studies of Sexual Size Dimorphism, captures the book's contents much better. The book focuses on morphological differences between males and females, with a major emphasis on overall body size as the dimorphic trait of interest.

The 20 chapters of this edited work are divided into three sections. The first six chapters on “macro-patterns” contain very useful reviews and analyses of dimorphism in mammals, birds, reptiles, amphibians, insects, and spiders. Although some may cry, “Where are the fishes!” (and likewise for some other neglected taxa), as a whole, this section serves as a greatly needed, updated reference work. It also introduces much of the adaptive reasoning that is the subject of closer scrutiny in the second section, this one on “micro-patterns.” This section contains eight case studies covering one or a few species each, ranging from the dimorphic plant Silene latifolia to eight hartebeest subspecies. Finally, the book closes with five chapters on some of the proximate mechanisms that underlie sexual size dimorphism.

Edited volumes are typically tough to compile. Researchers tend to be an opinionated bunch of people. Even after a workshop—and this book is based on one—it can be hard to ensure that they speak a common language. It is a merit of Sex, Size and Gender Roles that the boundaries between the sections are rather fluid. For example, adaptive explanations for observed patterns can be found in all of the sections, despite an initial (incorrect) warning that they will be absent in the last section of the book. Even so, it took me too long to figure out the common framework that underlies all the studies. In hindsight, I wish I had read the chapters in a different order—the book eventually clarifies some of the issues on selection that I struggled with initially, but it took time.

Why did I have trouble piecing some of the arguments together? To answer the question, “Why are males smaller [or bigger] than females in species X?” many chapters explain that a smaller size makes males more agile (which is beneficial when mate searching), or that a larger size confers an advantage during male combat, or that fecundity selection acts on females to make them larger. But whether these advantages should be reflected in selection currently favoring small or large males or females is a different question. Why? In chapter 9, by Daphne Fairbairn, readers are reminded that if both sexes have reached their optimum size, one expects stabilizing selection, not directional selection, around the mean size of each sex. If, on the other hand, size does not evolve independently in the two sexes because of genetic correlations, we expect genetic conflict to prevent the two sexes from reaching their optima (a point that is further clarified in chapter 18 by Stéphanie Bedhomme and Adam Chippindale). If so, we can indeed expect to find directional selection, as the average individual of neither sex reaches its optimal size.

But why did I have to read until chapters 9 and 18 (and other chapters toward the end of the book) before this issue became clear? Admittedly, the introduction before the first section mentions this point briefly, but I would have understood it sooner if the chapters had been offered in a different order. Readers would do well to start with the later chapters. It is there that we learn about adaptive theories together with proximate mechanisms, which are intimately linked issues in a field where genetic constraints can have a major impact on evolutionary outcomes. Only after one is armed with this knowledge can the single-species studies (section 2) —and finally the wider taxonomic patterns (section 1)—be understood.

To mention another example, Wolf Blanckenhorn's statement that “data do not support the differential-equilibrium model of SSD” (chapter 10) means, in the simpler language of Fairbairn, that selection is directional rather than stabilizing. Only after figuring this out could I accept Blanckenhorn's claim that since “net sexual selection on males is stronger than fecundity selection on females, current selection can explain why males are larger” (pp. 110–111). At first, this statement appeared to conflict with Fairbairn's prediction that at equilibrium, one does not necessarily expect directional selection at all.

Many of the taxonomic overviews near the beginning of the book take it as more or less granted that current directional selection pressures—for example, estimates of sexual selection or fecundity selection—can be used to explain current differences in the mean body sizes of males and females. This might indeed be a sensible null expectation, because females and males share the same genome (give or take one or a few sex chromosomes). As pointed out by Turk Rhen (chapter 16), a shared genome creates a massive constraint that prevents us from seeing wholly independent evolutionary trajectories that would lead to optimal sex-specific body sizes and the disappearance of directional selection. This is a clearly expressed point, one that might shed light on a question that receives relatively little attention in this book: Why don't males and females more often evolve separate niches, freed of their initial similar-size constraints? Yet the next chapter, by Russell Bonduriansky, appears to offer a somewhat different explanation for why there can be directional selection at equilibrium: condition-dependence of sexual traits.

Bonduriansky's style of argumentation will be familiar to researchers interested in traits subject to female choice. In that context, the lek paradox asks why there is still female choice for traits that should no longer show heritable variation after generations of directional selection resulting from female choice. The same question is relevant for sex-specific body size, or indeed any trait that has important fitness consequences and is subject to directional selection. Why are some males still small if larger ones always win fights? Bonduriansky tackles this question with clarity, but I wish that the authors of other chapters had paid more attention. The book does not describe how condition-dependence is related to the other approaches taken to explain how current selection for larger or smaller body size can persist. Instead, the chapter immediately following Bonduriansky's simply notes, “apparently, even with every opportunity to adapt, the average fitness of individuals often remains low relative to the fittest.”

Of course, a somewhat suboptimal ordering of chapters doesn't mean that one cannot eventually assimilate everything on offer. Perhaps my quibbles should be interpreted as a sign that there are still unresolved issues in this field, which in turn makes it an exciting one. Take, for example, Rensch's rule. This is an allometric rule formulated in several different ways throughout the book, but Lukás Kratochvíl and Danile Frynta expressed it simply in chapter 15: in small species, females tend to be larger; in large species, males do. The pattern applies in some taxa more consistently than in others. If we understood why, we would probably have a beautiful sequence of ideas from basic theory to explanations within and across species, taxon-specific where required, but with an understanding of why the exceptions exist. My impression is that we are not quite there yet, but the journey toward that goal is already offering some very rewarding views.

The best and most influential books might be those that do not solve all the problems they describe. If you need food for thought that requires some chewing, I can absolutely recommend this book. It might be best enjoyed, however, if, like an eager young child, you eat the dessert first and then work your way toward the inspiring taxonomic variety of the starters.