Contents lists available at ScienceDirect Applied Animal Behaviour Science journal homepage: www.elsevier.com/locate/applanim Improving canine welfare in commercial breeding (CB) operations: Evaluating rehoming candidates Judith Stella a, ⁎ , Traci Shreyer b , James Ha c , Candace Croney b a USDA-APHIS Center for Animal Welfare, 625 Harrison St, West Lafayette, IN, 47907, USA b Purdue University, Dept of Comparative Pathobiology, 625 Harrison St, West Lafayette, IN, 47907, USA c University of Washington, Dept. of Psychology, PO Box 351525, Seattle, WA, 98195, USA ARTICLE INFO Keywords: Commercial dog breeding Rehoming Welfare Management ABSTRACT Disposal of dogs retired from commercial breeding facilities presents a scientific and ethical problem. This pilot study aimed to develop criteria for identifying commercial breeding (CB) dogs likely to be at risk for problems during transitioning to rehoming, and to identify breeder practices associated with higher and lower risk. Dogs (n = 283) over 18 months of age from 17 CB kennels located in Indiana and Illinois, USA were assessed. The Field Instantaneous Dog Observation (FIDO) tool was used to assess behavior during a 4-step stranger approach test. At each step, the dog was given a RYG score; R, red (fearful), Y, yellow (ambivalent), or G, green (affiliative or neutral). After behavior assessment, 50 mg of hair was shaved from each dog’s lower back for analysis of hair cortisol concentration (HCC). A FitBark monitor collected 24 -hs of activity data. Additionally, a questionnaire was developed to interview each facility owner about their management practices. Overall mean HCC (n = 266) was 8.92 pg/mg (2.01–108.99 +/-14.24 pg/mg) while mean HCC at each facility ranged from 5.65 to 38.48 pg/mg. Mean activity score (n = 250) was 9.04 ± 4.3 and ranged from 2.8 to 32.8. Eight components from a principal component analysis (PCA) were retained as predictors in two mixed effects linear regression models. In Model 1, the component Socialization was significantly associated with HCC (p = 0.01). In Model 2 the components Late RYG (p < 0.001) and Activity (p = 0.002) were positively associated with Early RYG while Dog:Care (dog to caretaker ratio) (p = 0.03) and PD (periodontal disease) (p < 0.001) were negatively asso- ciated. Over half the dogs assessed using the stranger approach test exhibited fearful responses, suggesting that they might be at greater risk for problems if selected to transition to a new home. Additionally, the breeders’ management practices varied widely, with the number of socialization practices employed inversely associated with HCC. Further research is needed to assess the predictive value of the risk assessment and the general- izability of these results in a larger population of dogs and kennels. 1. Introduction Dogs have become increasingly popular companion animals in the USA with approximately 60 million households reportedly keeping at least one companion dog (American Pet Products Association, 2017). Commercial dog breeders help to meet high demand for purebred dogs. However, the welfare of dogs maintained in high volume breeding kennels is socially contentious (Bauer et al., 2017), especially given that many people now view dogs as family members (Bir et al., 2016). Commonly raised concerns include maintenance of the dogs in squalid conditions without appropriate attention to their behavioral and phy- sical health needs resulting in irreversible, adverse welfare con- sequences. For example, McMillan et al. (2011) reported that dogs placed in pet homes that were previously used as breeding stock in commercial breeding (CB) establishments displayed severe, persistent behavioral and psychological issues. Because of the purported deficits in the selection, rearing, and management practices used in high vo- lume CB kennels, societal concerns about such operations have been heightened (Osborne, 2008). A major limitation in drawing conclusions about the conditions in CB kennels is the paucity of published research on the topic. Until re- cently, most have relied primarily on owner reports of behavior of former breeding dogs and indirect inferences about CB dog welfare based on related questionnaires (McMillan et al., 2011, 2013; McMillan, 2017) rather than on direct observations of dogs in com- mercial kennels. Thus, significant deficits in knowledge of the actual conditions and welfare states experienced by dogs and puppies from CB facilities exist. This in turn constrains understanding of causal factors https://doi.org/10.1016/j.applanim.2019.104861 Received 21 December 2018; Received in revised form 19 July 2019; Accepted 28 August 2019 ⁎ Corresponding author. E-mail addresses: judi@gooddog.com (J. Stella), tshreyer@gmail.com (T. Shreyer), jcha@uw.edu (J. Ha), ccroney@purdue.edu (C. Croney). Applied Animal Behaviour Science 220 (2019) 104861 Available online 28 September 2019 0168-1591/ © 2019 Elsevier B.V. All rights reserved. T