Litter size is essential for an efficient production of rabbit meat. A diallel cross between four maternal lines was carried out and the analysis of the components of litter size has been already done. This paper presents the analysis of litter size traits themselves (total born (TB), number born alive (NBA), number weaned (NW)) and kindling interval (KI), that complete the analysis of the reproductive performance. The 16 genetic groups were distributed in four Spanish farms. The V line was present in all farms in order to be used as reference group. A total of 34 546 parities from 7111 does, were analysed. The crossbreeding parameters were estimated according to Dickerson model. The differences between lines performance were of low magnitude and not significant for litter size traits. The LP line showed the shortest KI followed by H respect to lines A and V. These differences reflected the differences between direct and maternal genetic effects. The differences between the average of all crosses and line V were found to be significant and seemed to be important, being 0.46 for TB, 0.56 for NBA, 0.75 for NW and −2.21 days for KI. The differences between reciprocal crosses for litter size were of low magnitude and non-significant, which indicate that the maternal effects are not important between these lines. In general, the lines did not show significant differences in direct and maternal genetic effects for TB, NBA and NW but there were some significant differences for KI, which ranged from 1.54 to 6.85 days in direct effects and from 0.63 to 3.38 days for maternal effects. A positive and, in some cases, relevant heterosis was found. The largest heterosis was for TB in the HV cross (1.05 rabbits), followed by the AH (0.74 rabbits), AV (0.57 rabbits) and LH (0.55 rabbits) crosses. For NBA, significant heterosis was found in HV (1.11 rabbits) and AV (0.49 rabbits) and for NW in AV (0.90 rabbits), LH (0.70 rabbits) and LV (0.58 rabbits). Favourable and significant heterosis for KI was found in AV and LV crosses, whereas it was unfavourable in AL and in LH. The more recommended crosses were AV and LH, showing the greatest performances on NW, while the cross HV, which shows the greatest performances an NBA, could be highly recommended when cross-fostering is practised to equalise litter sizes after birth, in this situation the lower NW performance observed could be improved.

Improving feed efficiency ( $${\text{FE}}$$ FE ) is a key factor for any pig breeding company. Although this can be achieved by selection on an index of multi-trait best linear unbiased prediction of breeding values with optimal economic weights, considering deviations of feed intake from actual needs ( $${\text{RFI}}$$ RFI ) should be of value for further research on biological aspects of $${\text{FE}}$$ FE . Here, we present a random regression model that extends the classical definition of $${\text{RFI}}$$ RFI by including animal-specific needs in the model. Using this model, we explore the genetic determinism of several $${\text{FE}}$$ FE components: use of feed for growth ( $${\text{WG}}$$ WG ), use of feed for backfat deposition ( $${\text{FG}}$$ FG ), use of feed for maintenance ( $${\text{MW}}$$ MW ), and unspecific efficiency in the use of feed ( $${\text{RFI}}$$ RFI ). Expected response to alternative selection indexes involving different components is also studied. Based on goodness-of-fit to the available feed intake ( $${\text{FI}}$$ FI ) data, the model that assumes individual (genetic and permanent) variation in the use of feed for maintenance, $${\text{WG}}$$ WG and $${\text{FG}}$$ FG showed the best performance. Joint individual variation in feed allocation to maintenance, growth and backfat deposition comprised 37% of the individual variation of $${\text{FI}}$$ FI . The estimated heritabilities of $${\text{RFI}}$$ RFI using the model that accounts for animal-specific needs and the traditional $${\text{RFI}}$$ RFI model were 0.12 and 0.18, respectively. The estimated heritabilities for the regression coefficients were 0.44, 0.39 and 0.55 for $${\text{MW}}$$ MW , $${\text{WG}}$$ WG and $${\text{FG}}$$ FG , respectively. Estimates of genetic correlations of $${\text{RFI}}$$ RFI were positive with amount of feed used for $${\text{WG}}$$ WG and $${\text{FG}}$$ FG but negative for $${\text{MW}}$$ MW . Expected response in overall efficiency, reducing $${\text{FI}}$$ FI without altering performance, was 2.5% higher when the model assumed animal-specific needs than when the traditional definition of $${\text{RFI}}$$ RFI was considered. Expected response in overall efficiency, by reducing $${\text{FI}}$$ FI without altering performance, is slightly better with a model that assumes animal-specific needs instead of batch-specific needs to correct $${\text{FI}}$$ FI . The relatively small difference between the traditional $${\text{RFI}}$$ RFI model and our model is due to random intercepts (unspecific use of feed) accounting for the majority of variability in $${\text{FI}}$$ FI . Overall, a model that accounts for animal-specific needs for $${\text{MW}}$$ MW , $${\text{WG}}$$ WG and $${\text{FG}}$$ FG is statistically superior and allows for the possibility to act differentially on $${\text{FE}}$$ FE components.

An alternative implementation of the animal model including indirect genetic effect (IGE) is presented considering pair-mate-specific interaction degrees to improve the performance of the model. Data consisted of average daily gain (ADG) records from 663 pigs kept in groups of 10 to 14 mates during the fattening period. Three types of models were used to fit ADG data: (i) animal model (AM); (ii) AM with classical IGE (AM-IGE); and (iii) AM fitting IGE with a specific degree of interaction between each pair of mates (AM-IGEi). Several feeding behavior phenotypes were used to define the pair-mate-specific degree of interaction in AM-IGEi: feeding rate (g/min), feeding frequency (min/day), the time between consecutive visits to the feeder (min/day), occupation time (min/day) and an index considering all these variables. All models included systematic effects batch, initial age (covariate), final age (covariate), number of pigs per pen (covariate), plus the random effect of the pen. Estimated posterior mean (posterior SD) of heritability was 0.47 (0.15) using AM. Including social genetic effects in the model, total heritable variance expressed as a proportion of total phenotypic variance (T 2) was 0.54 (0.29) using AM-IGE, whereas it ranged from 0.51 to 0.55 (0.12 to 0.14) with AM-IGEi, depending on the behavior trait used to define social interactions. These results confirm the contribution of IGEs to the total heritable variation of ADG. Moreover, important differences between models were observed in EBV rankings. The percentage of coincidence of top 10% animals between AM and AM-IGEi ranged from 0.44 to 0.89 and from 0.41to 0.68 between AM-IGE and AM-IGEi. Based on the goodness of fit and predictive ability, social models are preferred for the genetic evaluation of ADG. Among models including IGEs, when the pair-specific degree of interaction was defined using feeding behavior phenotypes we obtained an increase in the accuracy of genetic parameters estimates, the better goodness of fit and higher predictive ability. We conclude that feeding behavior variables can be used to measure the interaction between pen mates and to improve the performance of models including IGEs.

Heat stress has severe impacts on rabbit performance because they have difficulty getting rid of excess heat. The interaction between genetic and environmental factors plays a vital role in the adaptation process. The current study aimed to evaluate the effects of interaction between the genotype and heat conditions (G×H) on litter size and growth traits. Two rabbit lines were used in the current study, Egyptian maternal line (APRI) and New Zealand White (NZW). The rabbits were raised under normal (22°C) or heat stress (35°C) conditions. The does were raised in individual cages, and their kits were reared under the same conditions. Negative effects of heat conditions were noted for litter traits, as heat stress had significantly reduced the number of born alive, total born, and marketed number rabbits by about 16, 11, and 25%, respectively. Moreover, growth traits were reduced under heat stress conditions compared to normal temperature conditions. Significant differences between genotypes were observed, APRI rabbits were higher than NWZ rabbits in the total born, number of born alive, number weaned, and the number of marketed rabbits by 10, 8, 11, and 10%, respectively. Genotype by heat conditions interaction effects were observed, APRI litters under normal conditions showed significantly higher litter traits than NWZ litters, whereas relevant reductions in litter traits of APRI line than NZW litters when reared in heat conditions. For growth traits, the differences between the two lines under normal conditions were reduced when animals of the two lines were raised under heat stress, for BW₂₈, BW₄₂, BW₆₃, ADG₂₈₋₄₂, and DF₂₈₋₄₂ while these differences were increased for ADG₄₂₋₆₃, DF₂₈₋₄₂ and feed conversion during the whole fattening period. Based on the observed G × H interactions, it is important to select the animals under the same production conditions to raise rabbits that can cope with the expected global warming conditions.

This study evaluates the impact of thermal load on the weights of Tunisian local kids using 24 models with cubic and quadratic Legendre polynomials, based on daily temperatures (Tmin, Tmax, and Tavg) on the day of weight recording and averaged over 7, 14, and 21 days before weighing. The deviance information criterion (DIC) consistently shows that cubic polynomial models offer a better fit than quadratic models, highlighting their superior accuracy in studying the effects of thermal load on kid weights. The models with the best fit utilized average or maximum temperatures over 14 or 21 days. The patterns of response were similar across the temperature variables and periods, showing a stable weight response at lower temperatures (thermoneutral region) followed by a decline as the temperatures increased. The weight loss was −125 g/°C beyond the moderate heat stress threshold (Tavg21 = 17.7 °C) and −450 g/°C beyond the severe heat stress threshold (Tavg21 = 25.3 °C) for Tavg21. The heat stress thresholds for moderate heat stress (HS1) were 8.6 °C for Tmin14, 27.4 °C for Tmax14, and 18.6 °C for Tavg14; moreover, for acute heat stress (HS2), they were 17.2 °C for Tmin14, 32.4 °C for Tmax14, and 25.5 °C for Tavg14. High variability in individual responses was observed, with differences in the slope of response ranging from 2.0 kg/°C for moderate heat stress to around 3.0 kg/°C for severe heat stress for Tavg. The correlations between the weights under different temperatures were low, indicating that rankings based on weight could change with varying heat conditions. The animals with larger weight levels generally demonstrated better heat tolerance, and those with good heat tolerance under moderate conditions were also likely to have good tolerance under severe conditions.

Recirculating aquaculture systems (RAS) represent a sustainable approach to intensive fish production, but their operational complexity poses significant challenges for fish farmers. Addressing these challenges requires managing the complexity of multi-objective optimization with shifting priorities, ensuring interpretability of sophisticated AI models for practical users, and adapting control strategies to the dynamic requirements of fish development stages. This study introduces an adaptive multi-objective reinforcement learning framework that dynamically adjusts priorities across different growth stages, addressing a critical gap in current RAS management approaches. The proposed system implements a hierarchical Deep Deterministic Policy Gradient architecture with growth stage-specific policies, complemented by an interpretable visualization framework that makes complex AI decisions accessible to aquaculture practitioners. Experimental validation in a commercial-scale RAS facility with tilapia demonstrated substantial improvements in both production efficiency and system usability. Key findings include an 18.7% improvement in feed conversion ratio compared to standard implementations, while maintaining exceptional water quality stability (98.3% of time within optimal ranges). Particularly noteworthy was the system’s performance during growth phase transitions—historically challenging periods in RAS management—where adjustment time was reduced by 42.5% while maintaining stable environmental conditions. User studies revealed that the interpretable visualization components significantly enhanced farmers’ trust and understanding, with 89% of participants expressing confidence in the system compared to 65% for previous approaches. These results demonstrate how adaptive reinforcement learning with appropriate interpretability tools can bridge the gap between advanced AI techniques and practical aquaculture management, potentially accelerating adoption of intelligent control systems in commercial aquaculture.