Diet selection and eating behaviour of lactating goats subjected to time restricted feeding in choice and single feeding system

Small Ruminant Research 78 (2008) 41–47

Available online at www.sciencedirect.com

Diet selection and eating behaviour of lactating goats subjected to

time restricted feeding in choice and single feeding system

Murat G¨org¨ul¨u

, Mustafa Bo˘ga

, Ahmet S¸ahin

, U˘gur Serbester

,

Hasan R¨us¸t¨u Kutlu

, Suat S¸ahinler

a_{C¸ ukurova University, Agricultural Faculty, Department of Animal Science, Adana, Turkey} b_{Mustafa Kemal University, Agriculture Faculty, Department of Animal Science, Hatay, Turkey}

Received 16 January 2008; received in revised form 14 March 2008; accepted 8 April 2008 Available online 21 May 2008

Abstract

This study was carried out to investigate diet selection and eating behaviour of lactating German Fawn× Hair Crossbred goats in different feeding methods and levels. Twenty German Fawn× Hair first backcross does (B1) were allocated into 4 treatment groups

(2 feeding methods single (TMR) and choice feeding× 2 feeding levels ad libitum and restricted) with 5 replicates. Restricted feeding was applied only 4 h feed allocation during day. Barley, corn, soybean meal, corn gluten meal, wheat bran and alfalfa hay were feed ingredients for single and choice feeding. Eating patterns, milk yield and composition were determined for 8 weeks. The following results were obtained: (1) the meal criteria for goats restricted single and choice-fed, ad libitum single and choice-fed were determined as 1.00 and 0.63, 12.88 and 10.23 min, respectively. (2) Ad libitum feeding increased meal size, meal length, intermeal interval, total eating duration and decreased eating rate and meal number, compared to restricted feeding (P < 0.01). Choice feeding decreased meal size (P < 0.05), meal length (P < 0.01) and increased eating rate and meal number (P < 0.01), compared to single feeding. Restricted fed goats decreased intermeal interval in single feeding compared to choice feeding (P < 0.01), but increased meal number in choice feeding (P < 0.01). (3) Ad libitum choice-fed does made a diet containing 12.79% corn, 35.41% barley, 13.21% wheat bran, 5.35% soybean meal, 1.28% corn gluten meal and 29.80% alfalfa meal while restricted choice-fed does made a diet having more corn (27.69%), corn gluten meal (5.62%) and wheat bran (16.17%) and less barley (14.37%) and soybean meal (4.51%). (4) Choice feeding decreased RUP intake (P < 0.05) without affecting milk protein, irrespective to feeding levels, while having a tendency to increase in milk yield (14.2%) and 4% FCM (8.8%). (5) Restricted feeding decreased DM, ME, ADF and NDF intakes (P < 0.05) with concomitant decreases in 4% FCM, total milk solid, ash and fat compositions (P < 0.05), irrespective to feeding methods. (6) Choice-fed goats changed their preferences for a possible synchronized nutrient intake during a daytime, as sorted barley, soybean meal and alfalfa hay from early morning to late afternoon.

It could be concluded that choice-fed goats have the ability to make their diet to meet nutrient requirements and had a tendency to increase in milk yield. Restriction in feeding time resulted in lower feed intake and milk yield, although the animal changed their feed preference in favour of high quality ingredients and eating pattern with lower meal criterion and intermeal interval.

© 2008 Elsevier B.V. All rights reserved.

Keywords: German Fawn× Hair crossbred goats; Dietary choice; Time restriction; Milk production

∗_{Corresponding author. Tel.: +90 322 3386463.} E-mail address:gorgulu@cu.edu.tr(M. G¨org¨ul¨u).

1. Introduction

The goats offered feed ingredients simultaneously in ad libitum feeding condition can select their optimum 0921-4488/$ – see front matter © 2008 Elsevier B.V. All rights reserved.

diet to meet their nutrient requirements, correspond-ing to their physiological status. They consume high amount of concentrate without suffering any digestive problem by balancing fiber intake with creating syn-chronicity among feed ingredients in respect to energy and nitrogen supply to the rumen and/or host animal (Fedele et al., 2002; G¨org¨ul¨u et al., 2003; Yurtseven and Gorgulu, 2004). Choice feeding improves milk yield in dairy goats when offered feed ingredients with proper combinations maintaining their nutrient requirements (Fedele et al., 2002; G¨org¨ul¨u et al., 2003; Yurtseven and G¨org¨ul¨u, 2007). Thus choice feeding could be evaluated an economic and alternative feeding system for small ruminants.Rodriguez et al. (2007)tested choice feeding for lambs and proved that it was an alternative to con-ventional feeding system for fattening lamb in feedlot condition.

In conventional goat production, the goats are pas-tured in early and late hours of the daytime and feeding with restricted amount of concentrate during milking or after pasture. This system may lead goats to suffer from inadequate nutrient intake and consequently poor performance. The ability and/or good wisdom of small ruminants for diet selection in choice feeding situation may prevent the above referred problems by selecting a higher quality food to counteract the time restriction in feeding. There is, however, insufficient information about the ability of feed preferences and lactating perfor-mance in goats when feed ingredients are offered freely in ad libitum and restricted conditions throughout a day.

On the other hand, few studies on diurnal eat-ing pattern (daily feed intake, number of visits, meal length, intermeal interval, eating rate) of dairy cows (Tolkamp et al., 1998, 2000; DeVries et al., 2003) in single feeding condition, but not for dairy goats, either in single or choice feeding condition have been reported. Intermeal interval for dairy cows was deter-mined as 20 min (Metz, 1975; Tolkamp and Kyriazakis, 1999) by using frequency distribution. Rossi et al. (1999a,b) defined “meal” as feed removal exceeding 5 g and separated to two meals between each other by taking a 15 min period of not-eating as the cri-terion. However, there is still insufficient information regarding meal pattern of dairy goats subjected to dif-ferent feeding managements. Therefore, the aim of the present study was to investigate (1) whether goats would have the ability to maintain milk yield and nutrient intake during either free or restricted access to feed-stuff, (2) to determine meal criterion and diurnal feeding pattern in dairy goats under the above feeding condi-tion.

2. Materials and methods

In the study, 20 German Fawn× Hair first backcross does (B1), having twin kids, in the second lactation were

used. Twenty does weighing 54.8± 1.6 kg, yielding daily 2.72± 0.20 kg milk and being 25 ± 0.7 days after postpartum, were allocated into four treatment groups. Each goat within each treatment was one replicate.

Treatment groups were set up according to 2× 2 factorial arrangements to evaluate two feeding methods (single feed-ing and choice feedfeed-ing) in two feedfeed-ing levels (ad libitum and restricted feeding) as main effects. Restricted feeding was applied by allowing the goats to reach feed only from 09:00 to 11:00 and from 15:00 to 17:00 in a day, while there were no time restrictions applied in ad libitum groups. Barley, corn, soybean meal, corn gluten meal, wheat bran and alfalfa hay were used as feed ingredients for both single and choice-fed groups. Sin-gle diet (TMR) was prepared according to diet selection data obtained from the study ofYurtseven and G¨org¨ul¨u (2007)for the goats in same breeds.

Each goat was housed in a 1.5 m× 1.5 m sized pen having one trough sizing 0.4 m× 1.2 m with 15-L bucket for fresh water. Feed troughs of animals in choice feeding groups were divided into six equal parts to offer feed ingredients separately but simultaneously. The does were divided into four groups of similar mean live weight and milk yield.

The experiment lasted for 8 weeks. Before the experi-ment, choice does were habituated organoleptic properties and metabolic consequences of feed ingredients for a week. Wheat bran, corn and barley, soybean meal, and corn gluten meal along with alfalfa hay (1–2 cm particle size) were offered to choice does in the first, second and third day of the week. This proce-dure was repeated again for following three days to guarantee familiarisation of feedstuffs by goats. On 7th day, all ingredi-ents were offered simultaneously to be continued throughout experimental period. At the same time, single-fed does were accustomed to organoleptic properties of single diet. Each feed ingredient, except alfalfa hay, was mixed with limestone, salt and vitamin–mineral mixture at the levels present in the control diet in order to ensure the micronutrients intakes of choice does and, also, to prevent any possible effect of micronutrients on feed selection. Fresh water was available during experimental period.

The chemical composition of feed ingredients was deter-mined according to the standardAOAC (1998)procedures. NDF and ADF were analyzed by using ANKOM fiber ana-lyzer (Van Soest et al., 1991). Metabolisable energy contents of the diets were calculated using the table values of the feed ingredients published byNRC (1981). Ingredient composition of TMR and its chemical contents are given in the first column ofTable 1.

Live weight change, milk yield and feed intake were deter-mined weekly after training period. Feeds were offered ad

libitum to individual goat everyday, but recorded weekly.

Daily refusals were given to the animal on the following day by adding fresh feed or feed ingredients, but on last

Table 1

The proportional content of feed ingredients and macronutrients in single diet (TMR) or diets made by goats subjected to ad libitum and restricted feeding (mean± S.E.D., %)

Feeding methods Single feeding Choice feeding P-valuee

Feeding levels Restricted and ad libitum Restricted Ad libitum Feed ingredients/diets Single diet (TMR)d _{Goat-made diet}

Corn (C) 15.00 27.69± 10.03 12.79± 2.82 0.19

Barley (B) 26.39 14.37± 6.64 35.41± 3.80 0.02

Wheat bran (WB) 16.06 16.17± 7.36 13.21± 1.18 0.70

Soybean meal (SBM) 3.00 4.51± 2.41 5.35± 3.10 0.83

Corn gluten meal (CGM) 7.39 5.62± 1.97 1.28± 0.56b _0.06

Alfalfa hay (A) 30.00 29.49± 1.64 29.80± 0.65 0.85

Limestonea _{1.55 1.55± 0.04 1.55± 0.01 0.84} Salta _{0.51 0.51± 0.04 0.51± 0.00 0.87} Premixa,c 0.10 0.09± 0.00 0.10± 0.00 0.14 Macronutrient content (%) DM 90.19 89.87± 0.18 90.35± 0.09 0.04 ME (Mcal/kg DM) 2.58 2.61± 0.06 2.58± 0.01 0.62 CP 18.73 17.82± 2.03 16.68± 1.27 0.64 RUP 6.93 5.99± 0.95 4.34± 0.33b _0.13 ADF 19.75 19.39± 1.17 19.62± 0.19 0.85 NDF 32.55 31.65± 2.32 32.63± 0.39 0.68 Crude ash 10.24 10.47± 0.48 9.94± 0.13 0.31 Ether extract 3.08 3.30± 0.19 2.87± 0.08b _0.07

a_{These feed additives were added to single diet while they were selected indirectly with barley, corn, soybean meal and corn gluten meal.} b_{Denotes the difference between the diets selected by does and single diet based on one sample t-test (P < 0.05).}

c_{Each kilogram vitamin–mineral premix provides 8.000.000 IU vitamin A, 1.000.000 IU vitamin D}

3, 30 g vitamin E, 50 g Mn, 50 g Zn, 50 g Fe, 10 g Cu, 150 mg Co, 800 mg I and 150 mg Se.

d_{Man-made diet was offered to the both ad libitum fed and the feed restricted goats.}

e_{Shows the significance in differences between the diets selected by the does fed ad libitum and restricted.}

day of week, animal was allowed to finish the refusals and, then, offered fresh feed or feed ingredient. Animals were milked by hand at 07:00 in the morning and at 14:00 in the afternoon and milk samples were taken from morning milk and, then, milk fat were determined by Gerber method. Milk samples were also analyzed for dry matter, ash, milk pro-tein, NPN, casein nitrogen according to AOAC (1998). The rest of the nitrogen fraction was determined by calculation. Lactose was obtained by subtracting fat plus protein from milk organic matter as described by Sanz Sampelayo et al. (1998).

To record diurnal eating pattern of does, a prototype record-ing system was developed and set up includrecord-ing six scales 30± 0.005 kg capacity and connected to a computer with serial multiplier. The system recorded meal size and meal length for each balance with getting stability after 5 s from a shift in the balance weight for 24 h in a day for ad

libi-tum fed does and for 4 h (from 09:00 to 11:00 and from

15:00 to 17:00) in a day for time restricted fed does. As only one prototype system was owed in the experimental place, a doe from each treatment group was kept in a real time data recording system once a week for a day during 8 weeks experimental period. The data recorded for each week were taken as replicates for the doe from each treatment

groups. Meal criteria were calculated as the point at which the probability density functions of the final two populations cross from the parameters of the two and three-population Gaussian models which minimizes the mis-assignment of intervals to the wrong populations (Tolkamp et al., 1998; Tolkamp and Kyriazakis, 1999; Yeates et al., 2001). The mod-els were fitted using nonlinear curve fitting in GraphPad Prism (2005, Version 4.03) to the pooled and loge-transformed

inter-val lengths (expressed in second) between feeding events. Meal criterion for experimental groups restricted single feed-ing, restricted choice feedfeed-ing, ad libitum single feeding and

ad libitum choice feeding were determined as 1.00, 0.63,

12.88 and 10.23 min, respectively and accordingly, eating patterns of does (eating rate, meal size, meal length, inter-meal interval, inter-meal number and total eating duration) were determined for these experimental groups. Eating frequency in Fig. 1 was determined by counting number of visits to feed container by each doe kept in the recording sys-tem.

Data obtained in the experiment were analyzed by GLM procedure of SPSS with Duncan’s Multiple Range Test. The differences between the diets selected by does and single diet was tested by one sample t-test (Windows version of SPSS, release 10.01).

Fig. 1. Time dependent eating frequency of the does fed ad libitum (CF: choice feeding, TMR: single feeding).

Fig. 2. The proportional intake of feed ingredients by ad libitum choice-fed does. A: alfalfa hay, B: barley, CGM: corn gluten meal, C: corn, WB: wheat bran, SBM: soybean meal.

Fig. 3. The proportional intake of feed ingredients by does subjected to restricted feeding. A: alfalfa hay, B: barley, CGM: corn gluten meal, C: corn, WB: wheat bran, SBM: soybean meal.

3. Results

The experimental results are given inTable 1(diet selection),Table 2(milk yield and composition),Table 3 (eating pattern) andFigs. 2 and 3(diet selection).

Table 1shows that ad libitum choice-fed does made a diet containing less corn, corn gluten meal, and wheat bran and higher barley and soybean meal proportion-ally in comparison to those of the group’s offered the single diet without any statistical difference. Also, they did not change CP, ADF and NDF intakes. Choice-fed does tended to increase DMI (2.98 kg day−1 vs. 2.86 kg day−1), reflecting to a tendency to increase in fat corrected milk (FCM) yield (2.56 kg day−1 vs. 2.18 kg day−1) but poorer milk production efficiency (0.91 MY/DMI vs. 1.18 MY/DMI) (Table 2), compared to ad libitum single-fed goats.

Choice feeding decreased RUP intake (P < 0.05) in both feeding levels ad libitum and restricted feeding without affecting other macronutrient intakes in lactating does.

Restricted choice-fed does made a diet contain-ing higher corn, soybean meal and less barley and corn gluten meal proportionally compared to restricted single-fed does without any statistical difference. When compared to ad libitum choice-fed does, restricted choice-fed does tended to select higher amounts of corn and corn gluten meal, but similar CP, ADF and NDF intakes (Table 1). However, restricted choice-fed does were not able to increase FCM yield (1.66 kg day−1vs. 2.56 kg day−1, P < 0.05) as closer to what ad libitum choice-fed does did (Table 2).

Restricted feeding decreased DM, ME, ADF and NDF intakes (P < 0.05) with resulting decreases in 4% FCM, total milk solid, ash and fat compositions (P < 0.05) irrespective of feeding methods whereas choice feeding had no effects on milk yield and com-position.

Table 3 shows that ad libitum choice-fed does increased eating rate (13.04 g min−1 vs. 7.82 g min−1, P < 0.01) and meal numbers (17.63 vs. 13.25, P < 0.01) while decreased intermeal interval (50.72 min vs. 76.91 min, P < 0.05), meal size (167.07 g vs. 203.42 g per meal, P < 0.05), meal length (18.78 min vs. 33.08 min, P < 0.01) and total eating duration (328.94 min vs. 434.74 min, P = 0.08) compared to ad libitum single-fed does.

Restricted choice-fed does increased meal number (33.25 vs. 20.50, P < 0.01), intermeal interval (11.12 min vs. 6.03 min, P < 0.05) and eating rate (19.97 g min−1 vs. 12.70 g min−1, P < 0.01) while decreased meal size (63.60 g vs. 80.35 g per meal, P < 0.05), meal length (3.79 min vs. 6.97 min per meal, P < 0.01), and total eat-ing duration (125.94 min vs. 135.52 min, P = 0.08) in comparison to restricted single-fed does (Table 3).

Feeding method and feeding level interaction affected intermeal interval (P < 0.01) and meal number

signif-Table 2

The effect of single or choice feeding in ad libitum or restricted level on milk yield and composition Feeding methods (FM) Single feeding (TMR) Choice feeding (CF) P-value

Feeding level (FL) Restricted Ad libitum Restricted Ad libitum S.E.D. FM FL FM× FL

Milk yield (MY, kg/day) 2.31 2.63 2.30 3.34 0.35 0.34 0.07 0.32

4% FCM (kg/day) 1.69b _2.18ab _1.66b _2.56a _{0.27 0.52 0.02 0.45} BWC (g/day) 32.78 33.85 86.64 60.85 26.05 0.14 0.64 0.61 DMI (kg/day) 2.43b _2.86ab _2.63ab _2.98a _{0.15 0.30 0.01 0.76} MPE (DMI/MY) 1.05 1.18 1.40 0.91 0.20 0.85 0.38 0.14 MEI (Mcal/day) 6.28b _7.41ab _6.88ab _7.70a _{0.39 0.26 0.02 0.69} CPI (g/day) 455.35 537.35 470.89 499.69 45.85 0.81 0.24 0.57 RUPI (g/day) 168.48ab _198.82a _157.12ab _129.15b _{16.98 0.03 0.94 0.10} ADFI (g/day) 479.99 566.43 512.03 582.88 33.50 0.48 0.03 0.81 NDFI (g/day) 791.14 933.62 841.00 972.63 60.86 0.47 0.03 0.93 Milk composition Total solid (%) 10.88ab _11.28a _10.43b _11.07a _{0.18 0.08 0.01 0.50} Ash (%) 0.61ab _0.64ab _0.60b _0.67a _{0.02 0.55 0.03 0.32} Protein (%) 3.81 3.59 3.54 3.54 0.10 0.14 0.30 0.30 Fat (%) 2.29ab _2.86a _2.14b _2.50ab _{0.18 0.18 0.02 0.57} Lactose (%) 4.18 4.20 4.15 4.37 0.09 0.41 0.17 0.27

FCM: fat corrected milk, BWC: body weight change, DMI: dry matter intake, MPE: milk production efficiency, MEI: metabolisable energy intake, CPI: crude protein intake, RUPI: rumen undegradable protein intake, ADFI: ADF intake, NDF: NDF intake, DM: dry matter, NPN: non-protein nitrogen.a,b_{Different superscripted letters within the same row show statistical difference between them with respect to either feeding method or} feeding level (P < 0.05).

icantly (P < 0.05), as evidenced that restricted goats decreased intermeal interval in single feeding compared to choice feeding (P < 0.01) but increased meal number markedly (P < 0.01) in choice feeding (Table 3). 4. Discussion

The results obtained in the current study showed that lactating goats were able to make a diet to meet their nutritional requirements when they are offered feed ingredients on self selection basis, even though when

they were subjected to time restriction for feeding. The success in diet selection by choice-fed goats can be explained by hourly arrangements in feed intake and diet selection as shown inFigs. 1–3.

The does fed ad libitum exhibited intensive feed-ing activity near sunrise (08:00–11:00) and sunset (17:00–19:00) (Fig. 1). Fedele et al. (2002) observed that does maximized feed intake at 07:00 and 17:00. The reason of high feeding activity during sunrise and sunset is, more likely, an optimal solar radiation and tempera-ture during these periods because of warm season when

Table 3

Eating behaviour of goats in different feeding methods (single vs. choice feeding) and levels (ad libitum vs. restricted) set in the real time recording system throughout 8 weeks

Feeding methods (FM) Single feeding (TMR) Choice feeding S.E.D. P-value

Feeding level (FL) Ad libitum Restricted Ad libitum Restricted FL FM FL× FM

Meal criterion*(min) 12.88 1.00 10.23 0.63 – – –

Eating rate (g/min) 7.82c 12.70b 13.04b 19.97a 1.15 0.00 0.00 0.38

Meal size (g intake/meal) 203.42a 80.35c 167.07b 63.60c 12.04 0.00 0.04 0.42 Meal length (min/meal) 33.08a 6.97c 18.78b 3.79c 2.64 0.00 0.00 0.05 Intermeal interval (min) 76.91a 6.03c 50.72b 11.12c 4.76 0.00 0.04 0.00

Meal number 13.25c 20.50b 17.63bc 33.25a 1.87 0.00 0.00 0.04

Total eating duration (min/day) 434.74a _135.52c _328.94b _125.94c _{31.84 0.00 0.08 0.14} a,b,c_{Different superscripted letters within the same row show the statistical difference between them with respect to either feeding method or feeding} level (P < 0.05).

the current experiment was conducted. Also, ad libi-tum fed does, especially those on choice-fed, extended their feeding activity much wider time in the afternoon compared to the morning, more likely, due to reaching satiety by consuming fast and high amount of feed in the morning.

The does kept in the real time recording system pre-ferred barley in early time, around 08:00 am in ad libitum (Fig. 2) and 09:00 am in restricted feeding condition (Fig. 3). The preferences of feed ingredient were sorted by animals as barley, soybean meal and alfalfa hay from the beginning to the end of the cycle. Barley preference decreased during feeding cycle gradually and soybean meal (protein) preference increased in the same course by the choice-fed does irrespective of the feeding lev-els. This preference was supported byLeibowitz (1992) andKronberg and Malechek (1997), who reported that mammals prefer carbohydrate during early hours of the feeding cycle, whereas appetite for protein increases gradually over the course of active cycle. Barley is a highly degradable starch source and it can create fast and effective feedback signal in the rumen and body of animals. Other preferences during later phase of the day were concomitant with the minimization of discomfort in the rumen condition and nutrient balance in the rumen and/or host animal (Forbes, 2001; G¨org¨ul¨u et al., 2003; Yurtseven and Gorgulu, 2004).

Restricted does consumed feed very quickly in the morning, extended their feeding activity to longer period in the afternoon and selected corn (having high ME content with slow degradable starch source) and CGM (having high protein content with slow rumen degrad-able protein source) among the feed ingredients, more likely, to compensate the energy and protein require-ments for rumen microorganisms and host animal during feed restriction periods. Although the diet selected by the does in choice feeding group is highly nutritious, those does did not respond better to milk yield. These preferences could be attributed to needs optimizing rumi-nal condition mainly and changes to nutrient partition between mammary tissue and body fat tissues. This could be seen from the tendency in higher live weight gain in choice-fed does if they fed restricted (Table 2). The corn-based diets could supply more propionate and glu-cose resulting in increase in plasma insulin, which can increase body condition and fat accretion (Schmidely et al., 1999; Goetsch et al., 2001).

Ad libitum choice-fed does did make a diet having lower RUP content than other experimental groups. This may be explained that ad libitum choice-fed does may have consumed protein for their symbiotic microorgan-ism because of having more time to compensate their

protein requirement for milk, but time restricted choice-fed does tried to increase RUP intake which is known to be ready for absorption as the animal had a limited time to do so.

Restricted does increased meal numbers what ad libi-tum fed goats did, with a concomitant increase in corn consumption. If they had preferred barley to corn, they could have suffered from acidosis due to their highly degradable starch (Aldrich et al., 1993; Khorasani et al., 1994; Casper et al., 1999). It is well known that ruminants could change their diet preference to minimize imbal-ances in ruminal condition (Forbes, 2001; Fedele et al., 2002; G¨org¨ul¨u et al., 2003).

Expectantly, restricted feeding decreased DM, ME, ADF and NDF intakes, reflecting to lower total solid, fat and ash contents in milk in comparison to ad libi-tum feeding, irrespective to feeding methods. Of course, restricted does were allowed to eat feed 6 times lesser period than that of ad libitum does. In the current study, choice feeding improved nutritional status of the does by increasing DM, ME, ADF and NDF intakes, irrespec-tive to feeding levels. This was evidenced by a moderate increase in milk yield about 14.2% and 4% FCM about 8.8% with an improved MPE when compared to single-fed does. This increase in FCM and milk fat content with the increase in feeding level could be a reflection of the gained nutritional status of the does by ad libitum feeding (Sanz Sampelayo et al., 1998; Yurtseven and Gorgulu, 2004). The results obtained in this study and our previous studies (G¨org¨ul¨u et al., 2003; Yurtseven and G¨org¨ul¨u, 2007) showed that choice feeding have a potential to increase lactating performance. It has also been reported that choice feeding increased fattening performance in lambs (G¨org¨ul¨u et al., 1996). Additionally,Rodriguez et al. (2007)reported that choice feeding could be an eco-nomic and applicable feeding method in practical feedlot condition.

Choice feeding increased eating rate (13.04 vs. 7.82 for ad libitum feeding; 19.97 vs. 12.70 for restricted feeding) compared to single feeding. These could be explained that more choices and ad libitum feed avail-ability may give advantages to the does to do feeding activity more liberally. Furthermore, goats are more selective than other livestock in feeding activity (Van Soest, 1982); these may be incentive for goat to con-sume feed in shorter interval and higher incidence. More likely, to make precise dietary choices, present feed restricted goats increased meal number by making sec-ondary meals (Table 3) compared to ad libitum fed goats. Above all, the meal criterion was determined as 0.63 min in restricted choice-fed goats. It was clearly understood that restricted goats only had totally 4 h in a day for

feed allowance to compensate their nutritional require-ments for milk production and used about 52–56% of this allowed time for eating while ad libitum fed goats used only 22–30% of this time. Actually, goats eat during two long periods per day called main meals separated by sev-eral smaller meals called secondary meals (Abijaoude et al., 1999). Goats substituted one feed with another making a precise choice (Fedele et al., 2002).

It could be concluded that choice-fed goats have the ability to make their diet to meet nutrient requirements and had a tendency to increase in milk yield. Restriction in feeding time resulted in lower feed intake and milk yield, although the animal changed their feed preference in favour of high quality ingredients (for energy; corn vs. barley, for protein; corn gluten vs. soybean meal) and feeding pattern with lower meal criterion (0.82 min vs. 11.6 min) and intermeal interval (8.6 min vs. 63.8 min). Acknowledgements

The authors are thankful to staffs of Research and Application Farm of Cukurova University, Agricultural Faculty for their technical contribution and also to Scien-tific Research Fund of Cukurova University for financial support under the project number ZF2004BAP6. References

Abijaoude, J.A., Morand-Fehr, P., Bechet, G., Brun, J.P., Tessier, J., Sauvant, D., 1999. A method to record the feeding behaviour of goats. Small Rumin. Res. 33, 213–221.

Aldrich, J.M., Muller, L.D., Varga, G.A., Griel, L.C., 1993. Non-structural carbohydrate and protein effects on rumen fermentation, nutrient flow, and performance of dairy cows. J. Dairy Sci. 76, 1091–1105.

AOAC, 1998. Official Methods of Analysis. 16th ed., 4th Revision, Association of Official Analytical Chemists, Washington DC. Casper, D.P., Maiga, H.A., Brouk, M.J., Schingoethe, D.J., 1999.

Syn-chronization of carbohydrate and protein sources on fermentation and passage rates in dairy cows. J. Dairy Sci. 82, 1779–1790. DeVries, T.J., von Keyserlingk, M.A.G., Weary, D.M., Beauchemin,

K.A., 2003. Measuring the feeding behavior of lactating dairy cows in early to peak lactation. J. Dairy Sci. 86, 3354–3361.

Fedele, V., Claps, S., Rubino, R., Calandrelli, M., Pilla, A.M., 2002. Effect of free-choice and traditional feeding systems on goat feed-ing behaviour and intake. Livest. Prod. Sci. 74, 19–31.

Forbes, J.M., 2001. Consequences of feeding for future feeding. Comp. Biochem. Physiol. Part A 128, 463–470.

Goetsch, A.L., Detweiler, G., Sahlu, T., Puchala, R., Dawson, L.J., 2001. Dairy goat performance with different dietary concentrate levels in late lactation. Small Rumin. Res. 41, 117–125. G¨org¨ul¨u, M., Kutlu, H.R., Demir, E., ¨Ozt¨urkcan, O., Forbes, J.M.,

1996. Nutritional consequences of free choice among feed ingre-dients by Awassi lambs. Small Rumin. Res. 20, 23–29.

G¨org¨ul¨u, M., G¨uney, O., Torun, O., ¨Ozuyanık, O., Kutlu, H.R., 2003. An alternative feeding system for dairy goats: effects of free-choice

feeding on milk yield and milk composition in early lactation of Damascus goats. J. Anim. Feed Sci. 12, 33–44.

Khorasani, G.R., de Boer, G., Robinson, B., Kennelly, J.J., 1994. Influence of dietary protein and starch on production and metabolic responses of dairy cows. J. Dairy Sci. 77, 813– 824.

Kronberg, S.L., Malechek, J.C., 1997. Relationship between nutrition and foraging behavior of free-ranging sheep and goats. J. Anim. Sci. 75, 1756–1763.

Leibowitz, S.F., 1992. Neurochemical–neuroendocrine systems in the brain controlling macronutrient intake and metabolism. Trends Neurosci. 15, 491–497.

Metz, J.H.M., 1975. Time Patterns of Feeding and Rumination in Domestic Cattle. Communications Agricultural University, Wageningen.

NRC, 1981. Nutrient Requirements of Goat. National Academy Press, Washington, DC.

Rossi, R., Del Prete, E., Scharrer, E., 1999a. Effect of the H1-histamine receptor agonist betahistine on drinking and eating behaviour in pygmy goats. Physiol. Behav. 66, 517–521.

Rossi, R., Del Prete, E., Rokitzky, J., Scharrer, E., 1999b. Circadian drinking during ad libitum and restricted feeding in pygmy goats. Appl. Anim. Behav. Sci. 61, 253–261.

Rodriguez, A.B., Bodas, R., Fernandez, B., Lopez-Campos, O., Man-tecon, A.R., Giraldez, F.J., 2007. Feed intake and performance of growing lambs raised on concentrate-based diets under cafeteria feeding systems. Animal 1, 459–466.

Sanz Sampelayo, M.R., Perez, L., Boza, J., Amigo, L., 1998. Forage of different physical form in the diets of lactating Granadina Goats: nutrient digestibility and milk production and composition. J. Dairy Sci. 81, 492–498.

Schmidely, P., Lloret-Pujol, M., Bas, P., Rouzeau, A., Sauvant, D., 1999. Influence of feed intake and source of dietary carbo-hydrate on milk yield and composition, nitrogen balance, and plasma constituents of lactating goats. J. Dairy Sci. 82, 747– 755.

Tolkamp, B.J., Allcroft, D.J., Austin, E.J., Nielsen, B.L., Kyriazakis, I., 1998. Satiety splits feeding behaviour into bouts. J. Theor. Biol. 194, 235–250.

Tolkamp, B.J., Kyriazakis, I., 1999. To split behaviour into bouts, log-transform the intervals. Anim. Behav. 57, 807–817.

Tolkamp, B.J., Schweitzer, D.P.N., Kyriazakis, I., 2000. The biologi-cally relevant unit for the analysis of short-term feeding behavior of dairy cows. J. Dairy Sci. 83, 2057–2068.

Van Soest, P.J., 1982. Nutritional Ecology of the Ruminant. O. and B. Books, Corvallis, OR.

Van Soest, P.J., Robertson, J.B., Lewis, B.A., 1991. Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in rela-tion to animal nutrirela-tion. J. Dairy Sci. 74, 3583–3597.

Yeates, M.P., Tolkamp, B.J., Allcroft, D.J., Kyriazakis, I., 2001. The use of mixed distribution models to determine bout criteria for the analysis of animal behaviour. J. Theor. Biol. 213, 413– 425.

Yurtseven, S., Gorgulu, M., 2004. Effects of grain sources and feeding methods, free-choice vs total mixed ration, on milk yield and com-position of German Fawn× Hair crossbred goats in mid lactation. J. Anim. Feed Sci. 13, 417–428.

Yurtseven, S., G¨org¨ul¨u, M., 2007. The effects of multiple choices for grain and protein sources differing in ruminal degradability on diet selection and performance of lactating dairy goats. J. Anim. Prod. 48, 7–14.