Effects of non-aversive versus tail-lift handling on breeding productivity in a C57BL/6J mouse colony

Abstract

Non-aversive handling is a well-documented refinement measure for improving rodent welfare. Because maternal stress is related to reduced productivity, we hypothesized that welfare benefits associated with non-aversive handling would translate to higher production and fewer litters lost in a laboratory mouse breeding colony. We performed a randomized controlled trial to examine the effects of a standard method of handling (tail-lift with forceps) versus non-aversive handling with transfer tunnels ('tunnel-handled') on breeding performance in 59 C57BL/6J mouse pairs. Intervention assignments could not be concealed from technicians, but were concealed from assessors and data analyst. An operationally significant effect of tunnel-handling (large enough differences to warrant programmatic change) was defined before study initiation as a 5% increase in productivity, or one extra pup over the reproductive lifetime of each pair. Pairs were randomly allocated to handling intervention and cage rack location, and monitored over an entire 6-month breeding cycle. For each group, we measured number of pups born and weaned, and number of entire litters lost prior to weaning. Differences between transfer methods were estimated by two-level hierarchical mixed models adjusted for parental effects and parity. Compared to tail-lift mice, tunnel-handled mice averaged one extra pup per pair born (+1.0; 95% CI 0.9, 1.1; P = 0.41) and weaned (+1.1, 95% CI 0.9, 1.2; P = 0.33). More tunnel-handled pairs successfully weaned all litters produced (13/29 pairs, 45% vs 4/30 pairs, 13%; P = 0.015), averaged fewer litter losses prior to weaning (11/29 pairs [38%] vs 26/30 pairs [87%]; P <0.001), and had a 20% lower risk of recurrent litter loss. The increase in numbers of pups produced and weaned with tunnel handling met threshold requirement for operational significance. These data and projected cost savings persuaded management to incorporate tunnel handling as standard of care across the institution. These data also suggest that overlooked husbandry practices such as cage transfer may be major confounders in studies of mouse models.

Fig 1. Distributions of pups born and weaned per pair with tunnel handling (n = 29 pairs) vs tail-lift with forceps (n = 30 pairs).

Frequency distribution and boxplots. (A) Total number of pups born per pair. Median difference (solid black line) was 2 pups more per pair with tunnel handling. (B) Total pups weaned per pair. Median difference (solid black line) was one more pup weaned per pair with tunnel handling.

Fig 2. Frequency of complete litter loss and pair non-productivity.

Complete litter loss events, as defined as disappearance of the entire litter before weaning due to pup death or disappearance, and pair non-productivity defined as no litter produced for >60 days past the previous litter weaned. Orange group = tunnel-handled pairs (n = 29); blue group = tail-lift with forceps handled pairs (n = 30).