Natural induction of spontaneous liver steatosis in Greylag Landaise geese (Anser anser) 1

Transcription

1 Published December 3, 2014 Natural induction of spontaneous liver steatosis in Greylag Landaise geese (Anser anser) 1 G. Guy,* L. Fortun-Lamothe, G. Bénard,# and X. Fernandez 2 *INRA, UE89 Palmipèdes à Foie Gras, Domaine d Artiguères, F Benquet, France; INRA, UMR1289 Tissus Animaux Nutrition Digestion Ecosystème et Métabolisme, F Castanet-Tolosan, France; Université de Toulouse INPT ENSAT, UMR1289 Tissus Animaux Nutrition Digestion Ecosystème et Métabolisme, F Castanet-Tolosan, France; Université de Toulouse INPT ENVT, UMR1289 Tissus Animaux Nutrition Digestion Ecosystème et Métabolisme, F Toulouse, France; and #Université de Toulouse INPT ENVT, Unité de Mycotoxicologie, F Toulouse, France. ABSTRACT: The present experiment aimed at demonstrating that the Greylag geese (Anser anser), commonly used for the production of foie gras, are able to develop spontaneous hyperphagia and subsequent liver steatosis under specific handling conditions and without overfeeding. One hundred and eighty male geese were used in this experiment. After a period of feed restriction, at the age of 19 wk, corn was provided ad libitum. From wk 21 to 23, the daylight duration was progressively reduced from 10 to 7 h and kept as such until the end of the experiment (wk 31). Thirty birds were slaughtered at wk 19, 23, 25, 27, 29, and 31. During the first 2 wk after corn delivery, the average consumption rose up to 600 g/bird/d and decreased slowly thereafter to reach 270 g at wk 31. The liver weight increased from 95 (wk 19) to 514 g (wk 31), and most of these changes were due to the increase in liver lipid content from 6 to 50% of liver weight. There was no mortality during the experimental period. Histological observations indicate that the accumulation of fat in the livers occurred through a large increase in the size of the hepatocytes without modification of the cell boundaries and without any sign of inflammation or degeneration. Our data clearly show that under specific management conditions of feeding and photoperiod, the geese are able to initiate spontaneous liver steatosis. These results demonstrate their natural ability to store fat in the liver without any visible sign of tissue alteration. However, the variability in the response remains very high (at wk 31, the CV in liver weight was 45%). Further research is needed to better understand the origin of this variability. Key words: geese, liver steatosis, spontaneous fattening 2013 American Society of Animal Science. All rights reserved. J. Anim. Sci : doi: /jas INTRODUCTION The practice of overfeeding is a very ancient tradition, originating from Egypt in 2500 BC. The initial 1 The present study was supported by funding from PHASE Division of the Institut National de la Recherche Agronomique (INRA France). The authors thank Jean-Bernard Laverze at UE Palmipèdes à Foie Gras (INRA Artiguères-Bordeaux) for expert assistance in animal care, and Stéphane Seidlinger at UMR TANDEM (INRA Toulouse) for taking care of the chemical and biochemical analyses. We also acknowledge the members of the scientific advising committee of this project: Dr. Elisabeth Baeza (INRA, URA, Tours, France), Dr. Christel Marie-Etancelin (INRA, SAGA, Toulouse, France), Dr. Thierry Astruc (INRA, QuaPA, Clermon-Fd, France), and Prof. Stéphane Davail (CNRS, UMR 5254 IPREM, Mont-de-Marsan, France). 2 Corresponding author: fernandez@ensat.fr Received May 23, Accepted October 11, goal was to fatten birds to provide energy food for human consumption. Nowadays however, force feeding has the main objective to produce foie gras (fatty liver), a high quality delicatessen item with a strong added value. It is generally stated that the efficiency of waterfowl in their response to overfeeding originates from the natural abilities of migrating birds to overconsume food to store energy before long travels (Odum, 1960). Indeed, in birds traveling for long distances, the premigratory increase in BW can reach +100% of the lean mass, and the major store of energy for migration is fat (Bairlein, 1987). However, another school of thinking states that birds used for foie gras production are far from wild migratory birds. In addition, it is generally admitted that the spontaneous premigratory fattening is mostly localized in subcutaneous tissue, and little energy is stored in the liver (Pond, 1978). Thus, the

2 456 Guy et al. link between liver fattening of waterfowls and their migratory origin remains under debate. The Greylag Landaise geese come from the wild Greylag goose, which is a migratory bird originating from Europe (Kahlert et al., 1996). The domestication of Greylag geese by the Egyptians started probably 5,500 yr ago (Farrell, 2004). These birds show a good capacity for liver steatosis (Guy et al., 1995), related to a high lipogenic capacity in the liver which is, in birds, the major site of de novo lipogenesis (Leveille et al., 1975). Premigratory fattening has been demonstrated in geese and it is closely related to the annual cycle of reproduction (review by Raveling, 1978). The premigratory increase in BW of Canada female geese may reach +46%, among which lipid content shows a +209% rise (Raveling, 1979). Meanwhile, liver weight was doubled and liver lipid contents showed a five-fold increase. The photoperiod is known to be the major environmental factor controlling migration itself and the premigratory fattening process (review by Bairlein and Gwinner, 1994). In particular, the shortening of daylength occurring before winter migration accelerates the development of fat deposition. It has long been demonstrated that body fattening could be experimentally induced in some migrant species by changes in the photoperiod, and that the level of fattening may reach that observed in corresponding wild birds (Odum, 1960). Previous observations have shown that geese were able to spontaneously increase feed consumption to accumulate sufficient body reserves before laying and subsequent egg incubation (Raveling, 1979). We developed the hypothesis that spontaneous fattening could be naturally induced in Landaise geese in breeding conditions by combining a short photoperiod and a stimulation of a high energy food intake. We also speculated that a substantial part of this fattening would occur in the liver. The aim of this experiment was to demonstrate the natural ability of domestic waterfowl, at least geese, to spontaneously develop liver steatosis in breeding conditions. MATERIALS AND METHODS The experiments described here fully comply with legislation on research involving animal subjects according to the European Communities Council Directive of November 24, 1986 (86/609/EEC). The investigators were certificated by the French governmental authority for carrying out these experiments (Agreement No ). All experimental facilities used here were also certificated for the breeding, care and slaughtering of animals (Agreement No. A40624). Table 1. Chemical composition and physicochemical traits of the corn Item Value Water, % of fresh grain 25 CP, % DM 9.20 Starch, % DM Cellulose, % DM 2.52 Ether extract, % DM 3.86 ph Value 5.70 Weight of 1000 kernals, g 272 Animals and Experimental Treatment Birds and Diets. This trial was performed from July 1, 2009 to February 2, 2010 at the INRA experimental unit on waterfowl (Benquet, France) on d-old male Landaise goslings of the Palmsire strain. Birds were fed a commercial diet from birth to 19 wk (2,800 kcal/kg ME and 180 g/kg CP from birth to 6 wk and 2,800 kcal/kg ME and 160 g/kg CP from 7 to 9 wk; Maïsadour, Benquet, France) to meet the NRC requirements (NRC, 1994). From 20 wk onward, the birds were fed a fresh harvested corn stored in a flexible silo (preservation in oxygen-free conditions). The corn was of standard commercial type except that it remained at the initial stage (i.e., at time of harvesting), of DM (DM = 75%). The characteristics of this corn are given in Table 1. The choice of this type of corn was based on its high palatability for geese (Guy, unpublished observations) and its high starch content, which favor fattening (Hermier et al., 1999). Feeding Program. The feeding program was inspired from those used for reproducers. It was based on a period of feed restriction which aims to ensure normal growth and prevent excessive fattening (McLandress and Raveling, 1981) and to induce hyperphagia when stopped. The details of the alimentary schedule were that the birds were fed ad libitum from birth to 5 wk of age, then 300 g/d in wk 6, 350 g/d in wk 7 to 9, 300 g/d in wk 10 to 11, 280 g/d in wk 11 to 14, and 250 g/d in wk 15 to 19. When the birds reached the age of 19 wk (i.e., from wk 20 onward), the birds had ad libitum access to corn. Housing and Management Conditions. From hatching to 4 wk, the birds were kept in a 100-m 2 pen. At the age of 5 wk, the animals were allocated into 6 groups of 30 birds each. The allocation was performed to ensure live BW homogeneity between the 6 groups. They were then housed inside 6 collective rooms of the same windowless building, equipped with a small outdoor yard. From birth to 15 wk, the birds were submitted to natural lighting. During this period the duration of daylight fell from 14 h 11 min (beginning of July) to 11 h 13 min (15 September). At 15 wk of age, the free access to the outdoor area was suppressed and the geese

3 Spontaneous liver steatosis in geese 457 were then kept under artificial light using a cycle of 10 h light:14 h dark with an average light intensity of 30 lux from 0700 to 1700 h. At the start of wk 21, the light duration was gradually decreased by 30 min every 2 d until reaching a cycle of 7 h light:17 h dark at almost 23 wk of age. This short photoperiod was maintained up to the end of the experiment. During the experimental period, no additional heating was provided in the pens, and the temperature was registered daily at 0730 h. Slaughtering, Sampling, and Measurements. The individual BW was registered at 15 wk. At that time, 6 groups of 5 birds were identified in each room to constitute 6 equivalent groups of 30 birds each to be slaughtered at different stages. Thirty animals were slaughtered at 19 wk of age (just before corn feeding), and were considered as the reference group before the experimental stimulation. Thereafter, a group of 30 animals originating from the 6 pens was slaughtered at the ages of 23, 25, 27, 29, and 31 wk, corresponding to 4, 6, 8, 10, and 12 wk of corn consumption, respectively. The collective feed intake was registered daily in each pen from 19 to 31 wk (i.e., during the feeding period to corn). Mortality was checked daily during the same period. The same slaughtering procedure was applied at each slaughter age. The food was not withdrawn before slaughter. The birds were crated in transport cages (4 birds/cage) and transported to the experimental slaughterhouse (5 min transport). The animals were uncrated and weighed. This BW was computed as slaughter weight. They were then suspended individually from a shackle with their head downward and electrically stunned using headonly scissor tongs delivering a 90 V, 50 Hz AC current for 5 s. All birds were slaughtered by a ventral cut of neck blood vessels within 5 s after the end of the stun. The carcasses were weighed after bleeding and plucking, but before evisceration. The corresponding weight was computed as carcass weight. At about 20 min postmortem, the carcasses were eviscerated and the liver was carefully removed and weighed. Abdominal fat was also collected and weighted. The carcasses were then stored overnight in a +4 C chilling room. The color of the liver was measured immediately after slaughter along the ventral face of the big lobe using the trichromatic CIE Lab coordinates system (L*, a*, b*) using a CR 300 Minolta chromameter (Minolta, Osaka, Japan). The average of 3 measures done at different sites (top, middle, and bottom of the large lobe) was computed on each liver. A sample of approximately 50 g was taken from the central part of the big lobe. This sample was divided into 2 parts which were immediately frozen in liquid N 2, vacuum packed, and stored at 80 C until further chemical analyses. A sample was collected and stored in 10% neutral buffered formalin until histological analysis. The day after slaughter, the carcasses were anatomically dissected and these weights were recorded: breast muscle (Pectoralis major), breast skin + subcutaneous fat, thigh + shanks meat and bone, and corresponding skin + subcutaneous fat. Chemical and Biochemical Analyses Liver samples were analyzed for water, protein (total N) and lipid contents. Water content was determined by oven drying at 105 C for 24 h to constant weight (JOCE, 1971). Nitrogen was determined according to the Dumas combustion method using the Leco auto-analyzer (model FP-428, Leco Corp., St Joseph, MI) and converted to CP using 6.5 as conversion factor. Lipid content was estimated on ground liver samples by nearinfrared spectroscopy on a NIR-System 6500 spectrophotometer (FOSS NIRSystems, Silver Spring, MD; Marie-Etancelin et al., 2011). For the analysis of glycogen, glucose, and lactate, about 1 g of frozen liver tissue was homogenized in 10 ml of 0.5 M perchloric acid, and 0.5-mL aliquots of the homogenate were taken after removal of the fat cake, for the enzymatic determination of glycogen and glucose after glycogen hydrolysis with amyloglycosidase (Dalrymple and Hamm, 1973). The rest of the homogenate was centrifuged (20 min at 2,500 g) and 4 C, and the supernatant was used for glucose solely, and lactic acid (Bergmeyer, 1974) determination. The content of glycogen was calculated as the difference between the results of the 2 sets of glucose determination. The results were expressed in μmol/g of liver. Histological Observations Histological observations were made on liver samples taken at 20 min post mortem. Sections of liver (0.5 cm in thickness) were fixed in 10% neutral buffered formalin. Once the specimen was properly fixed, it was placed into disposable plastic cassettes and embedded in paraffin wax. Block was clamped onto the cutting frame of the microtome and sliced in pieces of 3 μm and placed onto glass slides for removal of the paraffin and staining. Slices were stained with May-Grünwald Giemsa stain (Lillie-Pasternack staining; Gabe, 1968, p. 249), periodic acid Schiff, and Masson s trichrome. Samples were examined by light microscopy (Nikon eclipse 80i, Nikon France, Champigny-sur-Marne, France). Images were captured by camera Nikon DS Fi1 and analyzed with NIS Elements software (Nikon France).

4 458 Guy et al. Statistical Analyses Analyses of variance were performed using the GLM procedure (SAS Inst. Inc., Cary, NC). The model included the main effect of slaughter age and breeding room, and the interaction between these 2 factors. Where appropriate, differences between means were tested using Duncan s multiple range test. In an attempt to normalize percentage data, Log(x) or arcsin (x) transformations were performed according to the range of values. Arcsin (x) was preferred in the case of data ranging along low values (0 to 10%). RESULTS AND DISCUSSION Time-related Changes in Ambient Temperature and Corn Consumption The experiment was performed during winter time and as shown in Fig. 1, the temperatures inside the breeding rooms were globally low. They ranged from 1 to +13 C along the 3-mo experimental period, and the average temperature was +6.5 C. Some differences between the breeding rooms were observed (Fig. 1), and it seems also obvious that Rooms 1 and 6 exhibited the lowest and highest ambient temperature, respectively, in any experimental week. The average ambient temperature differed by 2.2 C between these 2 rooms. On average, the temperatures in Rooms 1 and 6 were 1 C lower and +1.2 C higher, respectively, than the mean temperature among the 6 breeding rooms. Figure 2 presents the changes in weekly averaged corn consumption for each of the breeding rooms. As already stated in the Materials and Methods section, these data represent the collective consumption in a single room, recorded daily, but averaged for a single animal. The initial response in terms of corn consumption was Figure. 1. Time-related changes in the ambient temperature in each of the 6 breeding rooms along the experimental period (temperature was recorded every morning at 0730 h and averaged for each week). Figure 2. Time-related changes in the weekly average consumption of corn by individual geese according to the breeding room (the consumption was recorded for the group but averaged by subject; n = 30). unequal among the different rooms. Indeed, 4 rooms out of 6 exhibited a consistent food intake as soon as the first day of corn delivery. On the other hand, and despite a previous feed restriction, the birds in Rooms 1 and 2 were apparently disturbed by the sudden dietary switch from complete pelleted feed to corn whole grains. The birds from these 2 rooms did not eat the corn during the first 3 d (results not shown), which explains their decreased consumption (250 g vs. 500 g corn/d) on the first week of control. Thereafter they progressively adapted to this new food, and the feed intake was the greatest in these 2 rooms on the second week of stimulation. To our knowledge, data on food consumption by geese under ad libitum feeding are scarce. In 6- to 8-wk-old Landaise geese kept under standard conditions, Arroyo et al. (2012) reported an average consumption of 380 g/d/animal. This level fell to 200 g at the age of 14 wk with only 2 h per day of feeding access. In adult laying Gray Lag geese, Sauveur et al. (1988) reported food consumption ranging from 340 to 380 g/d/animal, according to the energy level of the diet delivered ad libitum. In the present work, overall, the maximum food intake was observed during the second week of corn ad libitum feeding and several rooms exhibited during this week a consumption equivalent to an average of 600 g per day and per bird, which is much greater than the levels of spontaneous ingestion reported in the above cited literature. Thus, present results demonstrated that hyperphagia could be initiated after a feed restriction in adult geese. After 12 wk of stimulation and in spite of a regular decrease in corn consumption over the experimental period, geese had ingested an average of 31.3 kg fresh corn, i.e., 369 g/day per bird, corresponding to 23.5 kg DM (i.e., 277 g MS/d/bird). Whether the change in the photoperiod had influenced slowing the decrease in food consumption cannot be concluded from the present work because control birds without light stimulation were not available. Experiments are currently running in our laboratory to dispel this ambiguity.

5 Spontaneous liver steatosis in geese 459 Table 2. Time-related changes in the geese body composition along the experimental period (19 to 31 wk of age; all data expressed as g except mentioned) Slaughter age (corn ad libitum feeding starting at wk 19) P-value 1 Item wk 19 wk 23 wk 25 wk 27 wk 29 wk 31 SEM Room Week BW at 15 wk ns ns BW at slaughter 6006 c 7634 b 7549 b 8078 ab 8141 ab 8267 a 1092 ns <0.001 BW gain 2 41 a 1636 b 1549 b 2094 a 2204 a 2341 a 817 ns <0.001 Carcass weight 4072 b 5177 a 5139 a 5460 a 5363 a 5402 a 691 ns <0.001 Liver weight (LW) 95 d 191 c 181 cd 268 bc 304 b 514 a 171 ns <0.001 LW/BW at slaughter, % 1,57 d 2,43 cd 2,30 cd 3,14 bc 3,56 b 6,20 a 0,17 ns <0.001 Abdominal fat 192 c 447 b 458 b 536 a 552 a 596 a 137 ns <0.001 Breast muscle 290 c 337 ab 339 ab 352 a 341 ab 320 b 44.8 ns <0.001 Breast skin and fat 101 c 199 ab 186 b 204 ab 213 ab 217 a 48.7 ns <0.001 Thigh muscles and bones 360 c 408 ab 407 ab 412 a 395 ab 386 b 44.9 ns <0.001 Thigh skin and fat 149 b 286 a 250 a 269 a 284 a 269 a 70.9 ns <0.001 Body fattening indicator c 1416 ab 1331 b 1482 ab 1546 a 1567 a 352 ns <0.001 a d Within row, different superscripts indicate significant difference between means (P < 0.05). 1 Level of probability of the effects of the rearing room (Room) and age at slaughter (Week). None of the interaction between the 2 main effects was significant. ns = nonsignificant. 2 BW gain calculated by difference between the weight at 15 wk old and slaughter weight. 3 Body fattening indicator is an approximation of total body fat calculated as abdominal fat + 2 (breast skin and fat) + 2 (thigh skin and fat). Time-related Changes in the Body Composition of Geese Production performances are presented in Table 2. No mortality was observed throughout the duration of the experiment. At the age of 15 wk, before any experimental treatment, the average BW was 5968 g. In male geese from the same genetic type and raised under similar standard conditions, Arroyo et al. (2012) reported a BW of 5888 g at the age of 14 wk. During the experimental period, the BW (measured at slaughter) increased. As compared with the BW at 15 wk, the weight gain ranged from +41 g (wk 19) to +2,340 g (wk 31), corresponding to approximately +25% of the initial BW. It is worth noting that in wild Canada geese, premigratory hyperphagia has been shown to increase BW by +36% and +26% in female and male geese, respectively (McLandress and Raveling, 1981). The increase in BW recorded in our experiment for male geese during 4 wk (wk 20 to 23; +25%) is therefore similar to the premigratory BW gain reported in wild geese for a similar duration (1 mo). Our data showed, however, that the difference in slaughter weight from wk 23 to 31, though significant, was of low magnitude (600 g, corresponding to approximately 7% of BW; Table 2). Thus, most of the changes in BW occurred between the beginning of the experimental period (wk 20) and wk 23. Breast skin and subcutaneous fat, and thigh skin and subcutaneous fat, were significantly increased between wk 19 and 23 (Table 2). The corresponding levels showed an almost twofold increase when considered on an absolute mass basis. As relative to slaughter weight, the corresponding proportions were 1.7 and 2.5% for breast and thigh skin + fat, respectively, at wk 19, and these values reached 2.6 and 3.8% of BW at wk 23, respectively, but they did not change during the rest of the experimental period (data not shown). These results suggest that the maximum subcutaneous fattening seems to be reached early after the start of the stimulation, in accordance with the observed changes in BW. The situation in terms of deep fattening appreciated by abdominal fat weight was somewhat different. Indeed, the initial (192 g) amount of abdominal fat was increased more than twofold at wk 23 and 25 (447 and 458 g, respectively) and it showed a second step of increase in wk 27 to 31, with values ranging from 536 to 596 g, but without significant differences between the 3 last stages. Expressed as a percentage of BW, the abdominal fat proportion ranged from 3.2 to 7.2% at wk 19 and 31, respectively, with gradual increases between these 2 steps (5.9, 6.1, 6.6 and 6.8% for wk 23, 25, 27, and 29, respectively). In 14-wk-old Greylag geese, abdominal fat represented 3.3% of total BW (Arroyo et al., 2012), a value very similar to the 1 we obtained at wk 19 in the present work (3.2%). In overfed geese, abdominal fat may reach 4.7% of total BW (Hermier et al., 1999), a value which remains less than those recorded in the present work along the experimental period from wk 23 (5.9%) to wk 31 (7.2%). In wild Canada geese, Mc Landress and Raveling (1981) showed that the BW gain during premigratory hyperphagia was composed of 61% lipids, 10% proteins, and 35% water. Body fattening is the main consequence of hyperphagia in wildlife, and we confirm this pattern in domestic geese under our

6 460 Guy et al. experimental conditions. Data on breast and leg muscle weights (Table 2) show a pattern of changes in 2 steps. From wk 20 to 23, there was a significant increase in muscle mass, in breast as well as in thigh; though for the latter, the distinction between muscle and bone weight was not done. Thereafter, and more specifically from wk 27, there was a significant decrease in muscle weight. Indeed, the protein content in corn is less than 10%, which is under the nutritional recommendation for adult geese (13 to 14%; Farrell, 2004). Time-Related Changes in Liver Weight At wk 19, before the start of the experimental treatment, the liver weight (CV 15%) corresponded to 1.57% of BW (Table 2). These data are consistent with previous works in domestic geese showing liver weights between 1.5 and 2% of BW in Greylag male geese (Arroyo et al., 2012). Thereafter, the liver weight increased from 191 g at wk 23 to 514 g at wk 31. Correspondingly, because changes in BW were of low magnitude, the liver weight expressed as a percentage of BW gradually increased from 2.43% at wk 23 to 6.20% at wk 31. At wk 23 and 25, the liver weights showed a twofold increase compared with wk 19. However, the CV was dramatically enhanced from 15% to 55 and 62%, respectively, thus indicating a strong interindividual variability in the liver weight changes. This heterogeneity can be further illustrated by the minimum and maximum liver weight observed: 65 vs. 532 g at wk 23 and 68 vs. 590 g at wk 25. This high variability probably explains the lack of statistical differences in liver weight between the first steps of the experimental treatment. Thereafter, there was a progressive increase in liver weight until wk 31, where the maximum value was reached (514 g). The variability in liver response also increased at wk 27 (CV 71%) and 29 (CV 74%), but it was considerably reduced at wk 31 (CV 45% ). It is interesting to note that 20 livers of the 30 birds slaughtered at wk 31 exhibited a weight over 400 g, this weight being the legal limit in France for the denomination foie gras given to goose liver fattened after force feeding in commercial conditions. due to the decrease observed in protein content, the latter being independent of the mode of expression (as a percentage of total or fat-free liver weight; Fig. 3). A further illustration of these links is given in Fig. 4. There was a very strong linear relationship between DM and lipid contents in the livers (R 2 = 0.99; P < 0.001). In addition, Fig. 4 also shows that liver weight and lipid content were closely related (R 2 = 0.87; P < 0.001). As the weight and the lipid content of the livers increased during the experimental period, there were concomitant changes in the color (Fig. 5): the lightness (L*) and the yellowness (b*) increased progressively, though the significant changes for these 2 items were found be- Liver Composition and Histological Characteristics As shown in Fig. 3, the pattern of changes in lipid content closely followed the changes in liver weight. The level of lipids rose from less than 10% (wk 19; lean livers) to about 50% at wk 31. The content of DM also increased along the experimental period, when expressed as a percentage of total liver weight. This increase was mostly due to the rise in lipid content. Indeed, when expressed as a percentage of fat-free liver weight, DM decreased along the experimental period, probably Figure 3. Time-related changes in the chemical composition of male geese livers along the experimental period (n = 30/time point). Data are expressed as percentage of total, or fat-free, liver weight. Vertical bars show the SE of the mean. Different letters indicate significant differences between times at P < 0.05.

7 Spontaneous liver steatosis in geese 461 Figure 5. Time-related changes in the trichromatic coordinates of male geese livers along the experimental period (n = 30/time point). Vertical bars show the SE of the mean. Different letters indicate significant differences between times at P < Figure 4. Relationships between (A) total lipid content and DM and (B) in liver. tween wk 19 and 23 on the 1 hand and between wk 29 and 31 on the other hand (Fig. 5). The observed changes in the redness (a*) were significant but they did not show a clear relationship with the corresponding evolution of liver weight. The L* and b* values reached on average 52 and 38, respectively, in the livers from geese slaughtered at wk 31. In overfed geese showing liver weights between 850 and 900 g, greater L* values (66 to 75) and lower b* values (approximately 25) have been reported (Fernandez et al., 2003, 2010). After 12 wk of corn consumption under the conditions described in the present work, the liver developed spontaneous fattening and their weight increased from less than 100 to 514 g. Under standard overfeeding conditions in Landaise geese, liver weight usually reach values above 800 g with lipid contents of 50 to 55% (Salichon et al., 1994; Leprettre et al., 1997). These values for lipid content are very close to those we recorded in the present study for spontaneously fattened livers (~50%). It is interesting to note that there was a positive relationship between liver weight and increase in BW from the start of the experiment (wk 19) and slaughtering (R 2 ranking between 0.50 and 0.58 for wk 23 to 29 included, results not shown). However, the relationship between liver weight and BW gain had totally disappeared (R 2 = 0.03) at wk 31. The sharp increase in liver weight (304 vs. 514 g) between wk 29 and 31 was unexpected, particularly because the corn consumption was at the minimum level during this last period (278 g per day and per bird). This consumption was low comparatively to the average consumption over the whole experimental period of 369 g. A daily 278-g consumption of corn at 25% moisture does not seem sufficient to explain this increase in liver weight, considering that in the same time, the content of lipids sharply increased (as shown in Fig. 3). It could be hypothesized that the lipids which accumulated in the liver during the last 2 wk could originate mainly from another body source. However, there was no evidence for the mobilization of peripheral fat because subcutaneous or abdominal fat did not decrease significantly during the final 2 wk. During the experimental period, the concentration of glycogen in the liver rose from wk 19 to 29, showing a twofold increase when expressed as a percentage of fatfree liver weight (Fig. 6). At the same time, the concentrations of lactate and free glucose showed only slight, if any, variation. During the last experimental period, however (from wk 29 to 31), there was a significant increase in the concentration of lactate whereas glycogen content tended to decrease, but the difference was not significant when expressed as a percentage of fat-free liver weight. In overfed geese with a short interval between the last meal and slaughter (9 h), Leprettre et al. (1997) reported glycogen concentrations of approximately 1.4% of liver weight (with livers weighing on average 857 g). This is less than in the present study where the concentrations of glycogen ranged from 3.6 (wk 19) to 5.4% (wk 29). In our experiment, food was available until the animals were collected for slaughter and this probably explains part of the difference between our data and those from Leprettre et al. (1997) because these authors showed that glycogen concentrations decrease very quickly in geese livers as the time elapsing from the last meal increases.

8 462 Guy et al. Figure 6. Time-related changes in the contents of glycogen, glucose, and lactate in male geese livers along the experimental period per gram of liver (solid line) or gram of fat free liver (dotted line; n = 30/time point). Vertical bars show the SE of the mean. Different letters indicate significant differences between times at P < It could also be inferred that the increase in glycogen concentration accompanying the increase in liver weight that we observed up to wk 29 in the present study could be a consequence of the spontaneous process of storing energy. Indeed, it has been shown in sparrows that premigratory hyperphagia led to an increase in liver glycogen which is thought to be the primary fuel for takeoff (review by Bairlein and Gwinner, 1994). The increase in lactate content observed between wk 29 and 31 could be linked a better uptake of circulating lactate involved in liver gluconeogenesis and lipogenesis through its oxidation to pyruvate. Some illustrations of the histological observations are presented in Fig. 7, for 3 representative livers showing various weights. Liver A (Fig. 7A) was harvested on wk 23 and showed a slight increase in weight, as compared with lean livers collected at wk 19 (122 g vs. 91 g on average for wk 19). In this liver, the hepatic architecture was well preserved. Some hepatic cells showed optically empty vacuoles which correspond to an overload of fat. The number of cells with this fat accumulation was still limited. The presence of a few clusters of mononuclear cells was noted but no fibrosis reaction were identified (Fig. 7A). In a heavier liver (269 g), the deposition of fat was more regular and uniform throughout the lobule (Fig. 7B). In the centrilobular area, the cells were hypertrophied and the fat accumulated in microvacuoles displaced the nucleus to the periphery (Fig. 7C), whereas in the periportal area, the fat was localized in larger vacuoles (Fig. 7D). Liver C was collected after 6 wk of corn feeding and weighted 790 g. The cell size was sharply increased but the cell membranes remained well preserved (Fig. 7E). The hepatocytes contained large fat vacuoles surrounded by a very thin layer of cytoplasm (Fig. 7F). No inflammatory reaction could be evidenced. Overall, the histological observations indicate that the observed accumulation of fat in the livers occurred through a large increase in the size of the hepatocytes without modification of the cell boundaries and without any sign of inflammation or degeneration. These observations suggest an adaptation of the liver to the accumulation of fat as already demonstrated in previous studies concerning overfeeding (Bénard et al., 2006). Our data clearly show that, under specific management conditions of feeding and photoperiod, the geese are able to initiate spontaneous liver steatosis. Such results demonstrate their natural ability to store fat in the liver without any visible sign of tissue alteration. It is of interest to note that along the 12 wk of experimental treatment, the mortality rate was 0. Though some livers may reach weights similar to those obtained with overfeeding (over 800 g) after 12 wk of corn consumption, the variability in the response remain very high and the average level of liver steatosis reached under our experimental conditions is far below the standard of overfed animals. This is a pioneer work on the spontaneous fattening of waterfowl liver, but it raises many questionings of scientific interest: i) the maintaining of fattening livers over a longer period, ii) the source of the individual variability in the ability to hyperphagia and to develop spontaneous fattening, and iii) the precise moment and conditions in which light stimulation must be used. Some projects are currently running in our laboratory to answer these questions.

9 Spontaneous liver steatosis in geese 463 Figure 7. Lillie-Pasternack (A, B, C, E, and F) and Periodic Acid Schiff (PAS; D) staining of 3 geese livers differing in weight (122 g for liver on A; 269 g for liver on B, C, and D; and 790 g for liver on E and F). LITERATURE CITED Arroyo, J., A. Auvergne, J.-P. Dubois, F. Lavigne, M. Bija, and L. Fortun-Lamothe Influence of feeding sorghum on the growth, gizzard development and carcass traits of growing geese. Animal 6: Bairlein, F Nutritional requirements for maintenance of body weight and fat deposition in the long-distance migratory garden warbler Sylvia borin (Boddaert). Comp. Biochem. Physiol. A 86:

10 464 Guy et al. Bairlein, F., and E. Gwinner Nutritional mechanisms and temporal control of migratory energy accumulation in birds. Annu. Rev. Nutr. 14: Bénard, G., T. Bengone, D. Prehn, S. Durand, C. Labie, and G. Bénard Physiology of ducks during force-feeding: Study of hepatic steatosis. Bull. Acad. Vet. Fr. 159: Bergmeyer, H. U Methods of enzymatic analysis. G. H. Bourne, editor. New York Academic Press, New York. p. 1127, 1196, 1238, Dalrymple, R. H., and R. Hamm A method for the extraction of glycogen and metabolites from a single muscle sample. J. Food Technol. 8: Farrell, D Management, nutrition and products of domestic geese: A review. Proc. Australian Poultry Sci. Symp. 16: Fernandez, X., E. Lahirigoyen, M. Bouillier-Oudot, Z. Vitezica, and A. Auvergne The effects of stunning methods on product qualities in over-fed ducks and geese. 2. Quality of Fatty Liver. Animal 4: Fernandez, X., S. Leprettre, J.-P. Dubois, A. Auvergne, and R. Babilé The influence of current parameters during the water-bath stunning of overfed geese (Anser anser) on blood loss and on fatty liver and meat downgrading. Anim. Res. 52: Gabe, M Techniques Histologiques. Masson, Paris. Guy, G., D. Rousselot-Pailley, and D. Gourichon Comparaison des performances de l oie, du canard mulard et du canard de Barbarie soumis au gavage. Ann. Zootech. 44: Hermier, D., M. R. Salichon, G. Guy, R. Peresson, J. Mourot, and S. Lagarrigue Hepatic steatosis in waterfowl: Metabolic basis and genetic susceptibility. INRA Prod. Anim. 12: JOCE Dosage de l humidité. Journal Officiel des Communautés Européennes L279/8. Eur-Op News, Luxembourg,Luxembourg. Kahlert, J., A. D. Fox, and H. Ettrup Nocturnal feeding in moulting Grey Lag Geese Anser Anser Anti-predator response? Ardea 84: Leprettre, S., A. Auvegne, H. Manse, R. Babilé, M. Candau, and J. P. Dubois Influence of starvation before slaughter and weight of fatty livers on biochemical hepatic composition in geese. Proc. 11th European Symp. on Waterfowl, 8 to 10 September World Poultry Sci. Assoc., Nantes, France. p Leveille, G. A., D. R. Rosmos, Y. Y. Yeh, and E. K. O Hea Lipid biosynthesis in chick. A consideration of site of synthesis, influence of diet and possible regulatory mechanisms. Poult. Sci. 54: Marie-Etancelin, C., B. Basso, S. Davail, K. Gontier, X. Fernandez, Z. Vitezica, D. Bastianelli, E. Baeza, M. D. Bernadet, G. Guy, J. M. Brun, and A. Legarra Genetic parameters of product s quality and hepatic metabolism in fattened mule ducks. J. Anim. Sci. 89: McLandress, M. R., and D. G. Raveling Changes in diet and body composition of Canada geese before spring migration. Auk 98: National Research Council Nutrient requirement of poultry, 9th rev. ed. Natl. Acad. Press, Washington, DC. Odum, E. P Premigratory hyperphagia in birds. Am. J. Clin. Nutr. 8: Pond, C. M Morphological aspects and the ecological and mechanical consequences of fat deposition in wild vertebrates. Annu. Rev. Ecol. Evol. Syst. 9: Raveling, D. G The timing of egg laying by northern geese. Auk 95: Raveling, D. G The annual cycle of body composition of Canada geese with special reference to control of nutrition. Auk 96: Salichon, M. R., G. Guy, D. Rousselot, and J. C. Blum Composition des 3 types de foie gras: Oie, canard mulard et canard de Barbarie. Ann. Zootech. 43: Sauveur, B., D. Rousselot-Pailley, and P. Larrue Alimentation énergétique de l oie reproductrice. INRA Prod. Anim. 1: