Egg weight loss and chick weight loss are well-documented for chickens and Pekin ducks. However, the Mule duck, a hybrid between the Pekin duck (Anas platyrhynchos domesticus) and the Muscovy duck (Cairina moschata), is less represented in scientific literature. This specificity generates particular embryonic characteristics that may influence moisture loss dynamics.
In this article, we focus on so‑called Mule eggs, which are used for producing Mule ducklings. These eggs are, in fact, laid by Pekin ducks, a distinction that is essential to avoid confusion with Pekin eggs intended for producing Pekin ducklings.
Understanding weight-loss patterns in Mule eggs is therefore essential to ensure that incubation conditions meet the physiological requirements of the embryo at each developmental stage. This study aims to characterise weight-loss evolution throughout Mule egg incubation.
Data collection
Data were collected throughout the entire year 2025 in the two largest Mule duck hatcheries operated by our company in France. In both facilities, all eggs entering the hatchery were systematically weighed upon arrival to establish a precise baseline mass before incubation. Additional intermediate measurements were performed at key physiological stages, specifically at 14 and 28 days of incubation, using calibrated industrial scales integrated into routine process workflows. At hatch, the weight of saleable ducklings was recorded from several selected batches each day of the hatching season.
This monitoring strategy enabled the construction of a comprehensive database containing more than 2,000 correlated data points. Each entry included egg weight, water-loss percentage, hatch performance, and multiple parameters associated with breeder flock characteristics and management. Collecting such detailed measurements within an industrial environment presented substantial logistical challenges compared with controlled R&D facilities, due to production constraints, workflow optimisation needs, and limited interruption tolerance.
The resulting dataset corresponds to approximately ten precise daily measurements for every hatch day over the full year, providing a robust foundation for analysing weight-loss dynamics in Mule duck eggs under commercial incubation conditions.
Weight loss
For many decades, water-loss dynamics during avian incubation have traditionally been described as a predominantly linear process, with a relatively constant rate of evaporative loss throughout embryonic development. More recently, however, several incubation specialists have proposed a two-phase model in which the early incubation period follows a slow, quasi-linear trajectory, while the second half exhibits a steeper decline in egg mass.
This shift is biologically plausible: during the first half of incubation, metabolic activity is driven mainly by biochemical reactions that produce little heat, limited CO2, and consequently minimal additional moisture release. In contrast, the second half of development is dominated by intense physiological processes associated with rapid embryonic growth, increased metabolic heat production, rising CO2 output, and greater evaporative water loss.
Although this pattern has been described and widely disseminated for chicken eggs, comparable data for duck species, and particularly for Mule ducks, remain scarce. In our study, we investigated water-loss evolution in Mule duck eggs under commercial incubation conditions. After aggregating data by incubation day, monotone PCHIP interpolation produced a smooth curve closely following observed daily means and revealing a slight convexity towards the end of incubation. This curvature is consistent with the acceleration reported in chicken studies and suggests that Mule embryos may follow a similar non-linear pattern. Building on this observation, we fitted a non-linear exponential model to the aggregated data (mean ~ a * exp(b * setting_day)).
The model converged rapidly (5 iterations; residual standard error = 0.00075) and accurately reproduced the expected cumulative loss by late incubation, in agreement with estimates derived from robust linear regression (~11–14%). Overall, our findings support the notion that water loss in Mule duck eggs is not strictly linear but better represented by a gently accelerating, exponential-type trajectory (Fig. 1).

We next evaluated whether water-loss patterns varied with parental age. Although a Welch ANOVA indicated a modest overall effect, post-hoc comparisons did not identify clear differences between specific age groups. From a practical perspective, these statistical results suggest that parental age alone is not a strong or consistent driver of water-loss dynamics in Mule duck eggs. However, the field observations tell a more nuanced story. Eggs originating from young and old breeder flocks showed markedly higher dispersion and greater variability compared with eggs from breeders in peak lay.Â
This pattern is commonly observed in commercial duck production: early in the laying cycle, flocks are still stabilising physiologically, and reproductive maturity fluctuates. It can cause variations in egg weight, which increases the dispersion of egg weight loss. Likewise, at the end of the laying cycle, declining shell quality is frequently reported, driven by cumulative physiological fatigue, reduced absorption efficiency, and increased sensitivity to management conditions.
Such variation directly affects eggshell conductance and weight homogeneity, two key determinants of water loss during incubation. Consequently, even when mean water-loss values appear similar across age groups, the reliability and predictability of shell performance differ substantially. In our data, peak-lay flocks produced eggs with tighter, more homogeneous conductance values and egg sizes, resulting in more stable and uniform water-loss trajectories. In contrast, young and old flocks generated eggs with uneven quality, which translated into a wide range of weight loss rates. These findings suggest that hatchery managers should pay particular attention to breeder-flock condition at the beginning and end of lay, where day-to-day variability may be higher despite similar average performance. Rather than parental age per se, it is the management-driven physiological stability of the flock that appears to shape water-loss consistency throughout incubation (Fig. 2).

Duckling weight loss
We further examined the relationship between initial egg mass and the final weight of saleable ducklings in order to characterise total water-loss dynamics across incubation. This analysis incorporated a wide range of potentially influential factors, including parental age, incubated egg weight, and ambient weather conditions (temperature, rainfall, and associated humidity). Among the direct effects, flock age emerged as a significant negative predictor of duckling weight loss (p = 7.457 Ă— 10-6 ***).
This result contrasts with trends commonly reported in the literature for egg weight loss, where egg weight loss is typically positively correlated with breeder age. In our case, however, the response variable concerns duckling weight loss, not egg weight loss, which introduces an important influence of developmental stage at hatch. For a given level of egg weight loss, a fresher (later hatching) duckling will be heavier at hatch, thereby exhibiting a lower apparent duckling weight loss. We therefore hypothesise that eggs from older breeders (typically bigger and requiring more time to reach core incubation temperature) experience a slight developmental delay. This delay may persist until hatch, producing slightly heavier and therefore “fresher” ducklings, which in turn show reduced duckling weight loss. However, this mechanism is limited by the overall quality of the egg, particularly shell integrity, which may decline with breeder age. These observations underline the importance of carefully assessing egg quality to anticipate responses under varying environmental and incubation conditions.
Ambient outside temperature emerged as another major driver: higher outside temperatures generally increased water loss, likely because warm periods in France are typically accompanied by drier air. A larger hygrometric gradient between the egg interior and the environment promotes passive diffusion and accelerates evaporation. Rainfall, in contrast, tended to slightly reduce water loss, as increased ambient humidity narrows this gradient and diminishes the efficiency of passive moisture transfer.
Beyond these main effects, two interactions proved particularly informative. The most robust was the interaction between parental age and outdoor temperature. In younger flocks, high temperatures gently increased weight loss, which is consistent with a drier atmosphere when the temperature increases in a temperate climate. In older flocks, however, the temperature effect was amplified by creating a steeper line. This suggests that shell characteristics and egg size in late lay may respond differently to environmental stress (Fig. 3).

A second, though less stable, interaction was observed between outside temperature and initial egg weight: heavy eggs exposed to high temperatures exhibited disproportionately large losses, whereas heavy eggs in cooler conditions lost much less. These results emphasise that temperature acts as a major modulator of evaporative dynamics, but only in systems where ventilation is static and thus sensitive to external climate fluctuations. Indeed, when the temperature of the incubation room decreases, the setters must adapt to this change. This can lead to variations in environmental conditions within the setter and impact the duckling’s weight loss. In contrast, dynamic-ventilation systems can maintain stable incubation conditions year-round, buffering eggs from environmental variability. Overall, our findings highlight the complex interplay between breeder physiology, egg characteristics, and climatic conditions in shaping water-loss outcomes in commercial Mule incubation.
Our final analysis examined the relationship between hatch performance (expressed in Mule ducks as the percentage of male ducklings obtained) and total weight loss from setting to the moment of saleable hatching duckling. It is important to note that all data were collected within the operational limits of our commercial process, which is optimised to deliver consistent results to provide our customers. Consequently, we do not intentionally push eggs towards extreme water-loss values that might compromise hatchability.
Under experimental conditions, previous work in avian species suggests that increasing water loss enhances hatch results only up to an optimal threshold, beyond which excessive evaporation becomes detrimental, generating a classic inverted-U (bell-shaped) response curve. In our production context, however, weight-loss values remain within a relatively narrow range, preventing the detection of this non-linear optimum. Instead, we observed a clear and robust positive linear relationship: within the operational envelope of our hatcheries, higher water-loss percentages were consistently associated with higher male-hatch rates.
Interestingly, the total weight loss measured in our Mule systems appears slightly higher than the classical recommendations for chickens and Pekin ducks, typically cited around 33%. The observed values may therefore reflect species-specific metabolic patterns or shell-conductance characteristics unique to Mule eggs under commercial conditions. This could also be a reflection of a particularly hot and dry year in France (Fig. 4).

These findings highlight the value (and the limitation) of working exclusively within routine hatchery settings: while operational data provide realistic and directly applicable insights, they inherently restrict the observable range of physiological responses.To fully characterise the nonlinear relationship expected between duckling water loss and hatch performance, controlled experimental studies would be required. Such trials could deliberately modulate humidity, shell conductance, and ventilation conditions to generate a wider spectrum of weight-loss outcomes. This would allow the identification of the true physiological optimum for Mule embryos and help refine the recommended target range for total weight loss. Ultimately, defining this optimal interval could improve both hatch performance and uniformity, offering a valuable opportunity to further optimise Mule incubation protocols.

