This experiment was carried out at the experimental rabbit farm at the College of Agriculture and Veterinary Medicine, Qassim University, Saudi Arabia, during the hot season. The daily variation in ambient temperature (high and low) during the experimental period was recorded. The average of the maximum and minimum temperatures (mean±standard error [SE]) was 34.2°C±0.4°C and 19.1°C±0.3°C, respectively and the relative humidity ranged from 19% to 42% (SE = 0.9). Two genotypes of male rabbits aged five weeks, the Saudi local breed (Jabali; J) and an imported Spanish rabbits (V-line; V) with an average body weight of 940.0 and 916.6±15.9, respectively were used in the current study. The Spanish V-line rabbits used in this study were imported from Valencia Polytechnic University in Spain to be crossed with a population of Saudi local rabbits in 2000 [18]. The V-line was originally established from four different synthetic maternal populations in 1984, crossing crossbred males of one type with crossbred females of another type [19]. The experimental animals were assigned and confirmed to have no abnormalities and were of good health status. In total, 250 animals (125 each) were housed in individual wire mesh cages (50×40×40 cm) in a semi-closed rabbitry. The cages were equipped with feeding hoppers and drinking nipples. Feed and water were provided ad libitum. All animals were housed under similar housing, management, and environmental conditions. All the experimental rabbits were fed a basal diet (Table 1). All experimental procedures, animal care, and handling were performed according to the animal care instructions of scientific research deanship and approved by the committee of health research ethics and animal care (# 190208), Qassim University, Saudi Arabia.

Growth traits and feed intake on a dry matter basis (DMI) were individually recorded for the rabbits over 6-week experimental trail. The FCR and the ADG were calculated during the experimental period. FCR was calculated as DMI divided by body weight gain. The ADG was computed by dividing the body weight gain by 42 (the number of days of the experimental period). At the end of the experimental period, the rabbits were fasted for 12 h and then slaughtered in the morning and weighed. After bleeding, they were dissected according to Blasco and Ouhayoun [20]. The slaughtered rabbits were skinned, and hot carcasses were weighed and recorded. Organs including the liver, heart, kidney, spleen, and thymus were removed and weighed. Carcass parts (fore, mid, and hind parts) were weighed. All collected data were calculated relative to live body weight as a percentage.

At the end of the experiment, blood samples for biochemical analysis were collected at the time of slaughter from each rabbit into heparinized tubes. The blood samples were centrifuged (1,500×g for 12 min) at 4°C and the harvested plasma was stored at −20°C for further analysis. Biochemical parameters, including total protein, albumin, total cholesterol, and triglycerides, were determined using commercial kits (Biomerieux, Craponne, France). The globulin concentration was determined by subtracting the albumin concentration from the total protein concentration.

When rabbits were ten weeks old, phytohemagglutinin P (PHA-P; Sigma Chemical Co., St Louis, MO, USA) was injected intradermally at a dose of 100 μg/kg in 0.1 mL of sterile saline into the left ear. The site of the needle was marked with permanent black ink upon injection. The ear swelling was measured before the injection and 24, 48, and 72 h afterwards using a constant tension dial micrometer (Ames, Waltham, MA, USA). The cell-mediated index was expressed as the difference between the ear thickness before and after injection. To assess the humoral immune response, total antibodies against sheep red blood cells (SRBC) were determined. Each animal was injected with 1 mL of a 10% suspension of SRBC. Sera of blood samples were collected one-week post injection to determine total immunoglobulins (Ig) using an agglutination assay. The results of antibody titers were expressed as log2 of the reciprocal of the highest dilution, giving a positive reaction. Estimation of log2 hemagglutinin titer was similar to that described by Fathi et al [21].

Expected feed intake was computed using mid-metabolic body weight (BW^0.75) to represent maintenance requirements and the ADG to represent production requirements using multiple regression analysis. Drouilhet et al [15] suggested a formula to calculate RFI according to the following equation:

In the current study, partial regression coefficients of each rabbit genotype were computed according to the following equation from the multiple regression analysis:

Where, OFI_ij, observed feed intake on a dry matter basis (g) of rabbit ith belonging to genotype jth; EFI_ij, expected feed intake (g) of rabbit ith belonging to genotype jth; BW_ij^0.75, mean metabolic body weight of rabbit ith belonging to genotype jth (g^0.75); ADG_ij, average daily gain (g) of rabbit ith belonging to genotype jth; a and b, partial regression coefficients; c, intercept.

The difference between observed (OFI) and expected feed intake (EFI) for each rabbit represents the RFI. RFI was calculated for each experimental animal by using the PROC REG procedure of JMP Ver. 11, SAS Institute [22]. The results are presented as means and the pooled standard error of the mean (SEM). The mean differences between the two genotypes of all studied traits were checked using Student’s t-tests. The PROC CORR procedure was applied to compute the correlation coefficient between RFI and the other traits within each genotype.

The slightly lower weaning weight observed in the V-Line rabbits could attribute to a possible differences between the two genotypes in litter size at weaning. However, this difference was more pronounced and became significant in advancing age. The results obtained for FCR fell within the normal range of most previous research studies, at 2.4 and 2.8 for Jabali and V-line, respectively. Obviously, the early fattening period showed a higher FCR than was observed during the latter period and near slaughter weight. Gidenne et al [5] recorded an FCR in the range of 2.1 to 2.3 in rabbits during the post-weaning period. In terms of RFI, the negative value recorded in the Jabali breed means a lower DMI may be expected compared to the V-line rabbits. The current results suggest that RFI is considered a good indicator of feed efficiency, but there is a disadvantage in that slow-growing animals consume a relatively small amount of feed. These animals show a more favorable RFI value [23,24]. In addition, selection for RFI results in a reduction in the FCR, accompanied by a decrease in ADG [10]. Many factors affect FCR in rabbit production. Genotype, management, feed quality, and ambient temperature are the most critical factors. When exotic breeds are transferred to another location, they may be expected to acclimatize to the new environment. Although the exotic rabbit breed (V-line) had spent a long time under the prevailing environmental temperatures in Qassim province, its performance, including ADG and FCR, was still lower than that of the local rabbits (Jabali). It seems that the V-line genotype is less adapted to ambient environmental factors. It is well documented that the imported breeds exhibit a deterioration in growth performance under high ambient temperatures compared to the local breeds. Iraqi et al [25] reported that the imported breed (V-line) showed lower body weights and weight gain than the Egyptian Gabali breed. They attributed their results to high temperatures affecting the growth rate of imported rabbits during the experimental period. Similar results were found in an experiment conducted on different genotypes by Fathi et al [2]. They stated that the imported rabbits (V-line) exhibited the lowest values of weight gain and feed intake when fed on a diet containing probiotic under summer conditions.

The superiority observed in carcass characteristics of Jabali genotype concurs with previous studies of [2,26]. They found that the local breed and their crosses had higher carcass percentages than exotic breeds. It is well documented that there is a genetic difference between genotypes and breeds of rabbits for carcass characteristics under different environmental factors [2,7,27]. Likewise, Paci et al [28] and Belabbas et al [29] confirmed that the genetic origin significantly affects carcass traits. In an investigation to study the difference between local and foreign rabbit breeds, Abdelsalam et al [30] found a higher (p<0.03) relative percentage of spleen in the Saudi local breed compared to New Zealand White breed. They suggested that local rabbits may have a better immune response compared to their foreign counterparts under high environmental temperatures. A possible explanation for the poor performance of carcass traits associated with imported rabbit breeds could be due to the unadapted genetic potential under hot environmental conditions.

Presently, the results obtained on blood parameters determined in the two genotypes are in agreement with the previous reports conducted under the same environment. In two rabbit genotypes fed a diet supplemented with eucalyptus leaves, Fathi et al [31] reported that both genotypes (local and exotic rabbits) performed similarly for all blood biochemical parameters except for cholesterol content, where V-line had a significantly higher concentration. Concerning the immune response, it could be concluded that the Jabali breed recorded a significant improvement in cellular mediated response after 24 h of PHA-P injection. However, little literature is available concerning the effect of breed or genotype on rabbit immunity under high environmental temperatures. A slight increase in the cellular immunity of local breeds compared to exotic breeds of rabbits raised under high ambient temperature (35°C) has previously been documented [2]. A higher inflammatory response in a line of rabbits selected for higher litter size over ten generations was observed under diverse environmental challenges [32].

The importance of daily body weight gain in deriving the multiple regression equation for RFI in many species has been reported [10,11,33]. A reduction in the FCR and the ADG was found in rabbit lines selected for RFI over 10 generations [10]. In the current study, the significant positive correlations between RFI and DMI or FCR are in alignment with the finding of Drouilhet et al [15]. They found a high genetic correlation between RFI and FCR (0.96) in a rabbit line selected for RFI under an ad libitum feeding regimen. Similarly, Fathi et al [7] reported a strong positive correlation between RFI and both FI and FCR in two varieties of Japanese quail. In a random population of Pekin ducks, Zhang et al [8] found that the RFI was strongly correlated with FI (0.82±0.03). In addition, they recorded a positive phenotypic correlation between RFI and FCR (0.55±0.04). A negligible or low correlation was found between RFI and body weight gain (close to zero) in Japanese quail and Pekin ducks [7,11]. The phenotypic correlation between residual feed consumption and the ADG was close to zero in a commercial line of rabbits selected on post-weaning growth rate [13]. Feki et al [34] found that a negative correlated response in feed conversion was expected when selecting for daily gain. According to low correlation between RFI and immune response found in our results, selection rabbits have a lower RFI may improve immune response if the genetic correlation has the same sign. The negative relationship between RFI and the hind part of carcass is consistent with the finding of Larzul and de Rochambeau [13], who found a strong negative correlation between RFI and the hind part percentage in a commercial line of rabbits. Furthermore, Al-Saef et al [18] reported that V-line rabbits were selected for improved growth performance and carcass composition. The last findings confirmed that the V-line may possess a genetic potential beyond that of Jabali rabbits, although the local breed gains more body weight under hot environmental conditions. It is obvious that genetic structure may play an essential role in the relationship between RFI and both carcass traits and blood parameters. However, the inverse results due to genotype effect need further investigation. Using RFI instead of FCR in a selection index does not introduce new information to the index [35]. Furthermore, if RFI is directly selected without its inclusion in an index, profits would be lower, as the component traits are not weighted to obtain the maximum benefit, as the index does. Additionally, RFI in the rabbit seems to have a genetic correlation with FCR close to one, but a lower heritability [15]. Therefore, using RFI in selection would be less efficient in improving feed efficiency than directly measuring FCR [36].