Aquatic Toxicology 87 (2008) 115–126 Variability in measures of reproductive success in laboratory-kept colonies of zebrafish and implications for studies addressing population-level effects of environmental chemicals Gregory C. Paull a,∗ , Katrien J.W. Van Look b , Eduarda M. Santos a , Amy L. Filby a , D. Melati Gray a , John P. Nash a , Charles R. Tyler a a Environmental and Molecular Fish Biology Group, School of Biosciences, Hatherly Laboratories, University of Exeter, Prince of Wales Road, Exeter EX4 4PS, United Kingdom b Institute of Zoology, Zoological Society of London, Regent’s Park, London NW1 4RY, United Kingdom Received 14 November 2007; received in revised form 14 January 2008; accepted 15 January 2008 Abstract Laboratory tests that quantify reproductive success using model fish species are used to investigate for population-level effects of endocrine disrupting chemicals (EDCs) and other chemicals discharged into the environment. Even for the zebrafish (Danio rerio), however, one of the most widely used laboratory models, surprisingly little is known about the normal variability in measures of reproductive success and this information is crucial for robust test design. In this study, the dynamics of breeding and inherent variability in egg output/viability and sperm quality were characterized among individuals/colonies and over time in 34 colonies of laboratory-kept zebrafish over a 20-day study period. For this work, a ‘6 × 6’ (six males and six females) colony size was adopted, as this is both environmentally relevant and optimal when considering egg output and animal welfare combined: an initial experiment showed egg output per female increased with decreasing colony size however, there was also a parallel increase in aggressive behavior. Both egg output and viability in ‘6 × 6’ colonies were highly variable among colonies (with co-efficients of variation (CVs) of 30 and 11%, respectively) and over the 20-day study duration (considering egg output and viability of all the colonies combined, the CVs were 20 and 12%, respectively). The patterns of egg production also differed among the ‘6 × 6’ colonies, and they included a cyclical output, a consistent daily output, an infrequent egg output with intermittent days of very high egg output, and an output with no obvious pattern. Sperm quality, measured as percentage motility and curvilinear velocity (VCL), was variable both among individuals within ‘6 × 6’ colonies and across colonies, with percentage motility being the most variable parameter (mean CVs of 82% inter-individual within colonies and 49% inter-colony). Sperm quality did not, however, vary over a 24h period. A minimum number of six replicate ‘6 × 6’ colonies, assessed daily for a period of 4 days, was required per treatment to detect a 40% change in egg output. The minimum numbers of individual males required per treatment to detect a 40% change in sperm quality using the breeding system adopted were 32 males for percentage motility and 12 males for VCL, equivalent to six and two ‘6 × 6’ colonies, respectively. These data demonstrate the need for high levels of replication when testing for effects of EDCs on reproductive output in the zebrafish model in an environmentally relevant (‘6 × 6’) breeding matrix. © 2008 Elsevier B.V. All rights reserved. Keywords: Zebrafish; Danio rerio; Reproduction; Egg output; Sperm quality; Test system; Endocrine disruption; Populations 1. Introduction For the past two decades, there has been worldwide concern about the impacts of endocrine disrupting chemicals (EDCs) discharged into the environment that have been associated with ∗ Corresponding author. Tel.: +44 1392 263792; fax: +44 1392 263700. E-mail address: g.c.paull@exeter.ac.uk (G.C. Paull). deleterious impacts on sexual development and reproduction in a wide range of wildlife species, in particular fish (reviewed in Tyler et al., 1998). The driving force behind the heightened concern about endocrine disruption is based on the hypothesis that these reproductive abnormalities may have serious impli- cations for normal population dynamics, community structure and ecosystem processes, questions that have yet to be fully addressed. Population-level effects that might result from repro- ductive abnormalities such as intersex, as a consequence of 0166-445X/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.aquatox.2008.01.008