Ecology, 88(9), 2007, pp. 2229–2239 Ó 2007 by the Ecological Society of America REPRODUCTION ON THE EDGE: LARGE-SCALE PATTERNS OF INDIVIDUAL PERFORMANCE IN A MARINE INVERTEBRATE SARAH E. LESTER, 1,3 STEVEN D. GAINES, 1,2 AND BRIAN P. KINLAN 1,2 1 Department of Ecology, Evolution, and Marine Biology, University of California, Santa Barbara, California 93106 USA 2 Marine Science Institute, University of California, Santa Barbara, California 93106 USA Abstract. Reproductive output is a central attribute of life history, providing a measure of individual and population performance. The fields of ecology, biogeography, and evolutionary biology take disparate approaches in addressing spatial variation in reproduction, and thus we lack clear predictions for how reproductive output should vary geographically. We empirically investigate these contrasting theoretical approaches by determining geographic patterns in reproductive output for intertidal populations of the purple sea urchin, Strongylocentrotus purpuratus, at 15 sites spanning a large geographic distance (98 span of latitude) from central California, USA, to Baja California, Mexico. Contrary to predictions from biogeography, some of the highest values of reproductive output are at sites near the species’ southern range boundary. Additionally, sea urchins do not exhibit a latitudinal gradient in reproduction, but rather show considerable mesoscale variation in reproductive output. Spatial analyses reveal that this variation is correlated with coastal topographic features that are known to influence the pattern of nearshore nutrient flux and circulation. We hypothesize that urchins’ reproductive output may be driven by the spatial distribution of their food supply, drift macroalgae, the abundance of which is influenced by both nutrient supply and alongshore transport processes that are coupled to topographic features. Large-scale studies such as this provide valuable insight into the causes of species’ range limits, population connectivity, habitat reserve design, and forecasting the effects of climate change on species’ distributions. Key words: biogeography; coastal topography; drift algae; food supply; gonad index; individual performance; latitudinal variation; reproductive output; sea urchin; Strongylocentrotus purpuratus. INTRODUCTION Understanding the causes and consequences of variation in life history characteristics is a central aim of evolutionary ecology. Geographic variation in species’ traits can yield particular insights for biogeo- graphic questions, such as the factors determining species’ distributional limits (Caughley et al. 1988, Gaston 2003). Reproductive output, an important component of individual performance, is one such trait. Reproductive output influences patterns of population persistence and often has implications for the causes of species’ range boundaries (Brown and Lomolino 1998, Gaston 2003). However, different disciplines, such as biogeography and evolutionary biology, take distinct conceptual approaches to spatial variation in reproduc- tive output. Furthermore, although there are notable exceptions (e.g., Dugan et al. 1991, Defeo and Cardoso 2002, Jump and Woodward 2003), few studies have explicitly or adequately examined how reproductive output varies over large geographic scales. More often, reproduction is studied over time or at only a handful of sites. As a result, we lack clear predictions for how reproductive output varies spatially, especially at the scale of species’ ranges. Biogeographic thinking about range-wide patterns of individual performance metrics, such as reproductive output, is shaped by ecological niche theory (Hutch- inson 1957). The concept of the niche is applied to the study of species’ ranges, with particular emphasis on the environmental factors impacting species’ geographic distributions (Brown and Lomolino 1998, Gaston 2003). Thus, it is commonly assumed that species’ ranges are influenced by an environmental gradient (reviewed in Sagarin and Gaines [2002a]) or a gradient in spatially autocorrelated niche requirements (Brown 1984). Based on this supposition, it is further assumed that the centers of species’ ranges are physiologically optimal, supporting populations with high individual performance and abundance, with declining perfor- mance and abundance further from the range center as populations become increasingly physiologically stressed (Brown 1984, Brown et al. 1995, Gilman 2005). These assumptions are subsequently incorporated into numer- ous ecological and evolutionary theories (Sagarin and Gaines 2002a, Guo et al. 2005, Sagarin et al. 2006). An abundant center distribution (species are most abundant near the center of their range) is widely assumed (Whittaker 1967, Hengeveld and Haeck 1982, Manuscript received 23 October 2006; revised 7 February 2007; accepted 20 February 2007. Corresponding Editor: S. Morgan. 3 Present address: COMPASS and Institute of Marine Sciences, University of California, Santa Cruz, California 95064 USA. E-mail: slester@ucsc.edu 2229