Behaviour reflecting immediate and prolonged postprandial satiety

Stereotypies (also often referred to as self-directed behaviour or non-feeding oral activities) are displayed in close relationship with a meal, and particularly shortly after feeding⁽Reference Terlouw, Wiersma, Lawrence and MacLeod ^{17 )}. As both feeding levels and fibrous diets have been shown to effectively reduce these behaviours after a meal (for example, see Terlouw et al. ⁽Reference Terlouw, Lawrence and Illius ^{8 )}; Brouns et al. ⁽Reference Brouns, Edwards and English ^{11 )}; Bergeron et al. ⁽Reference Bergeron, Bolduc, Ramonet, Meunier-Salaun and Robert ^{12 )}; Ramonet et al. ⁽Reference Ramonet, Robert, Aumaître, Dourmad and Meunier-Salaün ^{18 )}) post-meal stereotypies may reflect (lack of sufficient) satiety. When rooting material is present on the floor, which partly allows the expression of natural foraging behaviour, substrate-directed rooting and nosing may also be used as immediate postprandial indicators of satiety. Several studies report increased rooting when access to feed is restricted in pigs⁽Reference Beattie and O'Connell ^{19 –}Reference Jensen, Kyriazakis and Lawrence ^{21 )}. As manipulation of pen components seems to replace manipulation of rooting material in its absence⁽Reference De Leeuw and Ekkel ^{22 ,} Reference Whittaker, Spoolder, Edwards, Lawrence and Corning ^{23 )}, this behaviour may also reflect components of natural appetitive behaviour and hence be used as an indicator of immediate postprandial satiety. This behaviour may also be performed in a stereotypic manner (e.g. bar-biting).

Apart from stereotypic behaviour and rooting or manipulative behaviour, physical activity may also reflect satiety. It has been shown that when feed is restricted there is an increase in activity⁽Reference Beattie and O'Connell ^{19 )}, which corresponds with the observation that locomotion is an integral part of appetitive foraging strategies in pigs⁽Reference Stolba and Wood-Gush ^{5 )}. Theoretically, in contrast with stereotypies, which appear to be mainly elicited by a meal⁽Reference Terlouw, Wiersma, Lawrence and MacLeod ^{17 )}, physical activity and rooting or manipulative behaviour may be shown not only just after a meal, but also several hours later, anticipating the next meal. Thus, these behaviours may also reflect prolonged satiety.

Voluntary feed intake, and particularly time between subsequent feeding bouts, may also be a potential indicator of prolonged satiety. Data on voluntary feed intake patterns over the 24 h cycle, however, are often lacking in studies with dietary fibre. Thus, in the present paper, the focus will be on behaviours that potentially reflect satiety.

Hypotheses

It is hypothesized that (1) bulky fibres (including fibres with a high water-holding capacity), irrespective of their fermentability, reduce mainly stereotypies, but also physical activity and substrate-directed rooting and manipulative behaviours in the first hours after a meal in pigs, indicating enhanced immediate postprandial satiety; (2) highly-fermentable fibres, irrespective of their bulkiness properties, reduce physical activity and substrate-directed behaviours for many hours after a meal, indicating a prolonged postprandial satiety. As stereotypies appear to be more closely related to the immediate postprandial period, prolonged effects on these behaviours are not expected.

Immediate postprandial effects

In all studies self-directed behaviour, substrate-directed behaviour, general oral behaviour and physical activity directly after feeding are either reduced or not influenced by the inclusion of dietary fibre in the diet of sows (Table 2). The effects are not dependent on the fermentability of the fibre source, as both poorly-fermentable (e.g. OTH) and highly-fermentable (e.g. SBP) fibres reduce abnormal behaviours and physical activity. As no fibre sources with a low bulkiness (e.g. RS) were included in the studies reported, no firm conclusions of the effect of bulkiness per se can be drawn.

Self-directed behaviour immediately after a meal is reduced by SBP⁽Reference Brouns, Edwards and English ^{11 ,} Reference Zonderland, De Leeuw, Nolten and Spoolder ^{36 )} and a mixture of fermentable fibre sources⁽Reference De Leeuw, Zonderland, Altena, Spoolder, Jongbloed and Verstegen ^{37 ,} Reference Van der Peet-Schwering, Spoolder, Kemp, Binnendijk, den Hartog and Verstegen ^{38 )}, but not by OTH⁽Reference Bergeron, Bolduc, Ramonet, Meunier-Salaun and Robert ^{12 )} (Table 2). Self-directed behaviour was not measured in studies with wheat bran (WB). General oral behaviour (including self- and substrate-directed stereotypies), however, is reduced by both OTH⁽Reference Bergeron, Bolduc, Ramonet, Meunier-Salaun and Robert ^{12 ,} Reference Robert, Bergeron, Farmer and Meunier-Salaün ^{39 )} and WB⁽Reference Ramonet, Robert, Aumaître, Dourmad and Meunier-Salaün ^{18 )}. WB is not as effective as SBP in reducing oral stereotypies⁽Reference Ramonet, Robert, Aumaître, Dourmad and Meunier-Salaün ^{18 )}; however, this finding is only reported in an observation period that includes feeding, and when SBP is compared with WB and the control diet there is a higher feeding time, leaving less time for non-feeding activities.

All fibre sources tested reduce substrate-directed behaviour after feeding. Increasing levels of a mixture of fibre sources does not influence substrate-directed behaviour, but does linearly reduce total oral behaviour after a meal⁽Reference De Leeuw, Zonderland, Altena, Spoolder, Jongbloed and Verstegen ^{37 )} (data not shown in Table 2). This finding is in line with the hypothesis that postprandial oral behaviour may be a good indicator of the satiating effects of dietary fibres. In contrast, immediately-postprandial satiety is not always clearly reflected in general physical activity. For instance, in some studies SBP⁽Reference Brouns, Edwards and English ^{11 ,} Reference Zonderland, De Leeuw, Nolten and Spoolder ^{36 ,} Reference Braund, Edwards, Riddoch and Buckner ^{40 ,} Reference Danielsen and Vestergaard ^{41 )}, a mixture of fibre sources⁽Reference Van der Peet-Schwering, Spoolder, Kemp, Binnendijk, den Hartog and Verstegen ^{38 )} and OTH⁽Reference Bergeron, Bolduc, Ramonet, Meunier-Salaun and Robert ^{12 ,} Reference Robert, Matte, Farmer, Girard and Martineau ^{42 )} reduce physical activity, whereas in other studies they do not⁽Reference Whittaker, Spoolder, Edwards, Lawrence and Corning ^{23 ,} Reference de Leeuw, Jongbloed and Verstegen ^{34 ,} Reference De Leeuw, Zonderland, Altena, Spoolder, Jongbloed and Verstegen ^{37 ,} Reference Ramonet, Meunier-Salaün and Dourmad ^{43 –}Reference Whittaker, Edwards, Spoolder, Lawrence and Corning ^{45 )}. WB does not affect immediately-postprandial physical activity, but it was only tested in two studies, in one of which OTH was not found to be effective⁽Reference Robert, Rushen and Farmer ^{44 )}.

With the exception of two studies⁽Reference Brouns, Edwards and English ^{11 ,} Reference Ramonet, Robert, Aumaître, Dourmad and Meunier-Salaün ^{18 )}, feeding time was excluded or corrected for when measuring oral behaviour directly after a meal. For physical activity it is not always clear whether authors made this correction. As dietary fibres prolong eating time⁽Reference Brouns, Edwards and English ^{11 ,} Reference Bergeron, Bolduc, Ramonet, Meunier-Salaun and Robert ^{12 ,} Reference Ramonet, Robert, Aumaître, Dourmad and Meunier-Salaün ^{18 ,} Reference Whittaker, Spoolder, Edwards, Lawrence and Corning ^{23 ,} Reference Robert, Bergeron, Farmer and Meunier-Salaün ^{39 ,} Reference Danielsen and Vestergaard ^{41 ,} Reference Ramonet, Meunier-Salaün and Dourmad ^{43 ,} Reference Robert, Rushen and Farmer ^{44 ,} Reference Holt, Johnston, Baidoo and Shurson ^{46 )} and sows feed in standing posture, postprandial physical activity in sows fed fibrous diets could be overestimated when feeding time is not taken into account (for example, see Ramonet et al. ⁽Reference Ramonet, Meunier-Salaün and Dourmad ^{43 )} and Robert et al. ⁽Reference Robert, Rushen and Farmer ^{44 )}).

In growing pigs studies on behavioural effects of dietary fibres are lacking. Several studies, however, have measured the effects of dietary fibre on energy expenditure associated with physical activity. During the first hours after meal time activity-related heat production (AHP) is reduced in group-housed restrictedly-fed growing pigs, which has been observed for SBP-silage⁽Reference Schrama, Verstegen, Verboeket, Schutte and Haaksma ^{47 )} (Fig. 2) and RS⁽Reference Bolhuis, Van den Brand, Staals, Zandstra, Alferink, Heetkamp and Gerrits ^{48 ,} Reference Schrama and Bakker ^{49 )} (Fig. 3). Immediately-postprandial effects of RS⁽Reference Bolhuis, Van den Brand, Staals, Zandstra, Alferink, Heetkamp and Gerrits ^{48 ,} Reference Schrama and Bakker ^{49 )} cannot be attributed to its bulkiness (including water-holding capacity), which is low (Table 1) and does not differ from the bulkiness of the digestible starch used in the control diets, but are probably caused by other characteristics of RS. Unfortunately, effects of RS on behaviour have not been measured in sows.

Fig. 2. Daily pattern of activity-related heat production (AHP) in group-housed growing pigs fed a high-starch diet (○) or a high-NSP diet (•). BW, body weight; , Feeding-time; —, period of dark. Mean values for the time of day were significantly different between diets: *P<0·05, †P<0·10. (From Schrama et al. ⁽Reference Schrama, Verstegen, Verboeket, Schutte and Haaksma ^{47 )}; reproduced with permission of the Journal of Animal Science.)

Fig. 3. Daily pattern of activity-related heat production (AHP) in group-housed growing pigs fed a pregelatinized maize starch (○) or native potato starch (•) diet, calculated from data of Schrama & Bakker⁽Reference Schrama and Bakker ^{49 )}. BW, body weight; , Feeding-time; —, period of dark. Mean values for the time of day were significantly different between diets: *P<0·05, **P<0·01, ***P<0·001, †P<0·10.

Prolonged postprandial effects

Fermentable fibre sources clearly reduce physical activity for several hours after feeding, indicating prolonged postprandial satiety. The effects of poorly-fermentable sources such as OTH and WB on physical activity are less conclusive (Table 3). Self-directed behaviour was measured for a longer period post-feeding only in studies in which fermentable fibre sources were fed, and the levels shown are very low. In general, substrate-directed behaviour and total oral behaviour are reduced by both highly- and poorly-fermentable fibre sources.

Table 3. Review of prolonged postprandial effects of dietary fibre on behaviour in sows, measured during a preprandial (PP) or interprandial (IP; i.e. >4 h after a meal and not directly preceding a meal) period or over 12–24 h in different studies

SBP, sugarbeet pulp; SBPS, sugarbeet-pulp silage; OTH, oat hulls; WB, wheat bran; PK, palm-kernel expeller; CP, citrus pulp; SH, soyabean hulls; near-AL, sows were allowed to feed ad libitum during a 30 min period; LE, low energy; BE, same diet as LE, but balanced for energy by adding fat; PP1, before the first of two meals; PP2, before the second of two meals; ↑, ↓, ↔, increased, decreased and not affected by dietary treatment respectively.

* No. of meals per d.

† Feeding frequency was part of a 2×2 factorial design.

‡ Time period when lights were on.

§ Tendency.

As was expected, self-directed stereotypic behaviour is a poor indicator of prolonged postprandial satiety. Except for the immediate postprandial period (until 3 h after feeding), the levels of this behaviour are generally very low. SBP, however, is able to reduce self-directed behaviour before a meal to even lower levels, but only in one study in which sows that received the SBP-rich diet were allowed to eat large amounts⁽Reference Zonderland, De Leeuw, Nolten and Spoolder ^{36 )}. General oral behaviour before a meal and over the 24 h cycle (including both self- and substrate-directed behaviours) is reduced by a mixture of fermentable fibre sources⁽Reference De Leeuw, Zonderland, Altena, Spoolder, Jongbloed and Verstegen ^{37 ,} Reference Ramonet, Meunier-Salaün and Dourmad ^{43 )} and by OTH⁽Reference Bergeron, Bolduc, Ramonet, Meunier-Salaun and Robert ^{12 )}.

In contrast to the hypothesis, poorly-fermentable fibres from OTH and WB also reduce substrate-directed behaviour in the longer term⁽Reference Robert, Matte, Farmer, Girard and Martineau ^{42 ,} Reference Robert, Rushen and Farmer ^{44 )}; it should be noted that only chain manipulation was observed and measurements took place mainly just before feeding time in these studies. It is possible that manipulation of the chain, which was located near the trough, may have been compensatory behaviour, anticipating the lack of satisfaction of chewing needs associated with the upcoming fibre-poor meal.

Physical activity is clearly reduced by fermentable fibre sources⁽Reference Ramonet, Robert, Aumaître, Dourmad and Meunier-Salaün ^{18 ,} Reference de Leeuw, Jongbloed and Verstegen ^{34 ,} Reference Zonderland, De Leeuw, Nolten and Spoolder ^{36 ,} Reference De Leeuw, Zonderland, Altena, Spoolder, Jongbloed and Verstegen ^{37 ,} Reference Braund, Edwards, Riddoch and Buckner ^{40 ,} Reference Ramonet, Meunier-Salaün and Dourmad ^{43 ,} Reference Whittaker, Edwards, Spoolder, Lawrence and Corning ^{45 ,} Reference McGlone and Fullwood ^{50 ,} Reference Rijnen, Verstegen, Heetkamp, Haaksma and Schrama ^{51 )}; the only exception is a study in which time spent lying over the 24 h cycle is increased by soyabean hulls (see Holt et al. ⁽Reference Holt, Johnston, Baidoo and Shurson ^{46 )}). Physical activity over the 24 h cycle is reduced by the poorly-fermentable WB as effectively as SBP in one study⁽Reference Ramonet, Robert, Aumaître, Dourmad and Meunier-Salaün ^{18 )}. However, WB has no effect on physical activity⁽Reference Robert, Matte, Farmer, Girard and Martineau ^{42 ,} Reference Robert, Rushen and Farmer ^{44 )}, with even an increase in the frequency of posture changes before the second meal during the second pregnancy⁽Reference Robert, Matte, Farmer, Girard and Martineau ^{42 )}. In three studies physical activity is reduced by OTH, although only during the 12 h light period in one study⁽Reference Bergeron, Bolduc, Ramonet, Meunier-Salaun and Robert ^{12 )} and not over 24 h in the other two studies⁽Reference Robert, Bergeron, Farmer and Meunier-Salaün ^{39 ,} Reference Robert, Matte, Farmer, Girard and Martineau ^{42 )}. During the second pregnancy frequency of posture changes is affected by OTH, while standing is increased⁽Reference Robert, Matte, Farmer, Girard and Martineau ^{42 )} (not shown in Table 3). In another study physical activity is not affected by OTH⁽Reference Robert, Rushen and Farmer ^{44 )}. Thus, results of the poorly-fermentable OTH and WB on physical activity over the 24 h cycle are not conclusive. Highly-fermentable fibres, therefore, have a higher potential to prolong postprandial satiety. This conclusion is supported by a study in which the daily voluntary intake of several high-fibre diets was measured, which shows that the ingestion of diets with SBP (40–65%, w/w) is much lower (2·3–5·0 kg/d) than that of diets with OTH (37%; 7·7 kg/d) or WB (67%; 7·1 kg/d) ⁽Reference Brouns, Edwards and English ^{52 )}. The SBP-rich diets thereby reduce voluntary feed intake, with digestible energy intake approximating to levels that are acceptable for optimal reproduction. A study of the effect of a diet containing 45% (w/w) SBP fed ad libitum to gestating sows on their performance has concluded that feeding a high level of fermentable NSP ad libitum during three successive reproduction cycles has no negative effects on reproductive performance⁽Reference van der Peet-Schwering, Kemp, Plagge, Vereijken, den Hartog, Spoolder and Verstegen ^{53 )}.

Experimental evidence suggests that some (SBP, RS), but not all (soyabean hulls, solvent-extracted coconut meal), highly-fermentable fibre sources reduce AHP in group-housed growing pigs (Table 4). SBP silage⁽Reference Schrama, Verstegen, Verboeket, Schutte and Haaksma ^{47 )} (Fig. 2) and SBP⁽Reference Rijnen, Van den Borne, Schrama, Gerrits, Souffrant and Metges ^{54 )} tend to reduce AHP over the 24 h cycle in growing pigs, but the effect is predominantly present close to feeding. RS reduces AHP over the 24 h cycle, with the effects being significant not only after feeding, but also before the afternoon meal⁽Reference Bolhuis, Van den Brand, Staals, Zandstra, Alferink, Heetkamp and Gerrits ^{48 ,} Reference Schrama and Bakker ^{49 )} (Fig. 3). In one study all treatments increase AHP about 5 h before the second meal (i.e. 4 h after the first meal), with a decrease shortly after each meal⁽Reference Bolhuis, Van den Brand, Staals, Zandstra, Alferink, Heetkamp and Gerrits ^{48 )}. This finding is in agreement with those of a study that reports an interprandial increase in physical activity in sows starting 6–7 h before the second meal (i.e. 5–6 h after the first meal)⁽Reference de Leeuw, Jongbloed and Verstegen ^{34 )}. The highest AHP is shown by pigs fed the control diet and housed in straw-bedded pens⁽Reference Bolhuis, Van den Brand, Staals, Zandstra, Alferink, Heetkamp and Gerrits ^{48 )} (2×2 factorial design). The authors attribute this outcome to the presence of straw being an outlet for the increased motivation to display foraging behaviour in pigs fed the control diet. However, the addition of milled wheat straw to the diet, in order to increase its bulkiness, does not significantly reduce AHP over the 24 h cycle⁽Reference Schrama and Bakker ^{49 )}. In sows on a high-fibre diet (SBP and soyabean hulls) AHP is reduced between meals, but not when expressed as a 24 h average, probably because of increased eating time of the high-fibre diet⁽Reference Ramonet, Van Milgen, Dourmad, Dubois, Meunier-Salaün and Noblet ^{55 )}. SBP does not have a clear effect on AHP over the 24 h cycle, but a comparison of the two lowest fibre levels with the two highest fibre levels (contrast analysis), shows a decrease in AHP in the pigs on the highest-fibre diets⁽Reference Rijnen, Verstegen, Heetkamp, Haaksma and Schrama ^{51 )}.

Table 4. Review of prolonged effects of dietary fibre on activity-related heat production (AHP) in studies with group-housed growing pigs (GP) and sows (S)Footnote *

SBP, sugarbeet pulp; SBPS, sugarbeet-pulp silage; T, tapioca; GMS, gelatinized maize starch; SECM, solvent-extracted coconut meal; SH, soyabean hulls; RS, resistant starch; GPS, gelatinized potato starch; W+SB, wheat+soyabean meal; ↓, ↔, decreased and not affected by dietary treatment respectively.

* Animals were group-housed in all experiments, except Ramonet et al. ⁽Reference Ramonet, Van Milgen, Dourmad, Dubois, Meunier-Salaün and Noblet ^{55 )}. Feeding levels were 1·3 and 2·5×maintenance requirement for sows and growing pigs respectively, except Schrama et al. ⁽Reference Schrama, Verstegen, Verboeket, Schutte and Haaksma ^{47 )} (2×maintenance requirement) and Ramonet et al. ⁽Reference Ramonet, Van Milgen, Dourmad, Dubois, Meunier-Salaün and Noblet ^{55 )} (sows received 37·2 MJ digestible energy/d).

† Dose–response study, the highest inclusion level is given.

‡ Tendency.