Review: Chestnut and quebracho tannins in pig nutrition: the effects on performance and intestinal health

Article history: Received 17 January 2020 Received in revised form 19 August 2020 Accepted 21 August 2020 Available online 17 December 2020 Natural extracts are frequently adopted as a valuable alternative to antibiotics in intensive animal farming. Their diverse bioactive constituents such as phytosterols, glucosinolates, carotenoids and polyphenols have shown antioxidant, anti-inflammatory and antibacterial effects. Tannins are the largest class of polyphenol compounds of plant extracts, which can be classified into two hydrolysable or condensed subgroups. Poultry and swine nutrition are the most important sectors in which tannins have been used, firstly adopting tannin-rich feedstuffs and more recently, using tannin extracts from different plants. Several commercial products are available containing tannins extracted from the European chestnut tree (Castanea sativaMill.) and the American quebracho (Schinopsis spp.). Tannins extracted from these plants have been applied on intensive swine farms due to their ability to improve animal performance and health. These positive and prominent effects are frequently associated with the antinutritional effects in reducing feed palatability, digestibility and protein utilization of feed. Some criticisms and contrasting results regarding pig performance and intestinal health have been reported. This paper provides an overview of the effects of chestnut and quebracho tannins on growth performance and intestinal health of pigs in order to clarify the appropriate dosage and response in the various physiological stages. © 2020 The Authors. Published by Elsevier Inc. on behalf of The Animal Consortium. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).


Introduction
The increase in antibiotic resistance has led to the development of protective actions (e.g. prudent use of antimicrobials) based on the 'One Health' approach (European Food Safety Authority (EFSA), 2017; World Health Organization (WHO), 2017). The future actions and the World Health Organization action plan against antimicrobial resistance are based on best practices in implementing and monitoring the 'one health' plans, supporting novel solutions to prevent and treat infections, thereby increasing the efforts in terms of combating antimicrobial resistance and related risks worldwide.
In pig farming, the development of antibiotic resistance can lead to multifactorial infections (Rossi et al., 2013;Rossi et al., 2014a). Antimicrobials and/or additives based on pharmacological levels of zinc and copper oxide (2 000-3 000 ppm) are commonly used during the postweaning phase (Hejna et al., 2019;Hejna et al., 2020). However, many international organizations have recommended decreasing the use of copper and zinc oxide due to the suspected resistance to certain bacteria and risks associated with the impact of these heavy metals on the environment (European Food Safety Authority (EFSA), 2017; World Health Organization (WHO), 2017).
Alternatives to in-feed antibiotics and zinc oxide are thus required in order to prevent antimicrobial resistance and natural extracts could usefully replace antibiotics in swine industry (Onelli et al., 2017;Dell'Anno et al., 2020;Sotira et al., 2020). Natural extracts have several biologically active constituents, which have different bio-activities on animal metabolism (Huang et al., 2018;Perricone et al., 2020). Polyphenolic compounds are ubiquitous in all plant organs and commonly distributed in legumes, cereals and fruits (Huang et al., 2018). Tannins are an important class of polyphenol compounds and are mainly classified into two hydrolysable (HTs) or condensed (CTs) subgroups. They are present in a wide variety of animal feed ingredients such as forage, legumes, fruits, cereals and grains (Jansman, 1993).
Poultry and swine are the most important sectors in which tannins have been tested, firstly adopting tannin-rich feedstuffs (such as sorghum) and more recently using tannin extracts from different plants (Jansman, 1993). In poultry, early studies were conducted on the effects of tannin-rich feedstuffs on growth performance, feed efficiency and antibacterial capacity (Huang et al., 2018). These promising effects led researchers to also test tannins in swine industry (Huang et al., 2018). However, in the early 1980s most studies performed in swine were based on the effects of sorghum, barley, maize, faba bean tannins on pigs' digestibility, protein use and growth performance (Mitaru et al., 1984). Since then, various sources and tannin extracts have been adopted in pigs, such as tannins from grape seeds, grape pomace, tannic acid and acorns (Huang et al., 2018).
Other relevant sources of tannins are the European chestnut tree (Castanea sativa Mill.) and the American quebracho (Schinopsis spp.). Traditionally, chestnut and quebracho extracts have been adopted in leather tanning industry. Although, in the last decade these extracts have been applied in animal nutrition, due to the high concentration of HTs and CTs contained in chestnut and quebracho extracts. Hence, raised attention has been made on chestnut and quebracho tannins application as feed supplements. In addition, several additives containing chestnut and quebracho tannins are available in the market, confirming the great interest of animal feed sector. The increasing application of these extracts is also related to the continued availability of such products over the year, compared with other sources (e.g. grape pomace).
In monogastric animals, supplementation with tannins at different concentrations has positive effects because of their antioxidant, antiinflammatory and antibacterial activities (Biagi et al., 2010;Huang et al., 2018). Tannins extracted from chestnut and quebracho have been exploited on intensive swine farms and adopted in the different physiological stages of swine (post-weaning and growing/finishing phases) to enhance growth performance, modulate intestinal microbiota and decrease the incidence of diarrhoea in particular during the post-weaning period (Girard et al., 2019;Girard and Bee, 2020). Moreover, contrasting results on the effective supplementation of chestnut and quebracho on animal performance and intestinal health have been observed and, therefore, reported and discussed in the present review. This review analyses the effects of chestnut and quebracho tannins on the growth performance and intestinal health of pigs in order to clarify the appropriate dosage and response in the various physiological stages of pigs.

Classification and composition of tannins
Tannins are polyphenolic secondary metabolites generally defined as 'water-soluble phenolic metabolites' or 'macromolecular phenolic substances' (Khanbabaee and van Ree, 2001). They can be classified by: i) their molar mass between 300 and 3 000 Da; ii) their molecular structures, divided into two main groups (hydrolysable and condensed tannins) and iii) their structural properties (gallotannins, ellagitannins, complex tannins and condensed tannins) (Khanbabaee and van Ree, 2001). Due to the wide differences in the classification of tannins, for practical reasons we adopt the 'hydrolysable and condensed tannins' classification.
Hydrolysable tannins can be fractionated hydrolytically into their components (Khanbabaee and van Ree, 2001) by treatment with hot water or with the enzyme tannase, which catalyses the hydrolysis of tannins. Hydrolysable tannins are composed of esters of gallic (gallotannins) or ellagic (ellagitannins) acid and glucose (Smeriglio et al., 2017). Therefore, the term 'hydrolysable tannins' includes both gallotannins and ellagitannins. In fact, ellagitannins are not hydrolysable, but for historical reasons are still classified as HTs (Khanbabaee and van Ree, 2001).
Condensed tannins are non-hydrolysable oligomeric and polymeric proanthocyanidins, consisting of coupled catechin units. The coupling pattern of the catechin units in condensed tannins can vary extensively. Thus, different tannins present diverse structures and degrees of polymerization. The tannin synthesis occurs in the plants as protection against insects, diseases or grazing animals. Plants accumulate a considerable amount of tannins in the bark, roots, wood, leaves and fruits. Condensed tannins can be found in ferns, fern allies, gymnosperms and many dicotyledonous and monocotyledonous flowering plants (Constabel et al., 2014). However, HTs are limited to dicotyledonous plants (Constabel et al., 2014).
Tannins can also be found in several other plants conventionally used in animal nutrition, such as forage, shrubs, cereals and medicinal herbs (Huang et al., 2018). Several tannin-rich by-products such as grape pomace, olive, peanut, green tea, fruit and vegetable coproducts have been extensively studied in animal nutrition (Brenes et al., 2016). In addition, tannins can be directly extracted from different plants such as mimosa, oak, chestnut, quebracho trees or from byproducts such as grape seeds, and acorns (Krisper et al., 1992;Huang et al., 2018). Several studies have reported the use of tannins extracts from the chestnut and quebracho. Compared to other tannin-rich commercial products, chestnut and quebracho tannin extracts are widely studied in animal nutrition worldwide. The increasingly utilization of these compounds and the growing of plant extracts market have led to develop specific regulations. In light of this, tannins are recognized as safe ingredient by US Food and Drug Administration (FDA), 2017 (listed in 21 CFR 184.1097;21 CFR 173.310). Chestnut and quebracho extracts are classified as 'Natural flavouring substances and natural substances used in conjunction with flavours' in the Food Additives Status List (FDA,21 CFR 172.510), whereas in the European Union Register of Additives (annex I, 2019) chestnut (Castanea sativa Mill.) and quebracho (Schinopsis spp.) tannin extracts are under the subclassification 'Natural productsbotanically defined'. Despite the large application of various tannin extracts in animal nutrition, chestnut and quebracho extracts are the only ones listed in the European Union Register of additives, although limited authorizations were enquired to the European Food Safety Authority for the authorization of commercial products (European Food Safety Authority (EFSA), 2005 and 2016).

Chestnut tannins
Chestnut trees are a source of hydrolysable tannins, in which ellagitannins and gallotannins are the main representative classes. Ellagitannins are chemically characterized by at least two galloyl units (C-C coupling) and do not contain a glycosidically linked catechin unit (Landete, 2011). The hydroxydiphenoyl residue undergoes lactonization to produce ellagic acid, which is not easily hydrolysed because of the further C-C coupling of the polyphenolic residue with the polyol unit. The main ellagitannins are vescalagin, vescalin, vescalignin, castalagin, castalin and castaligin, while the most representative complex tannins are acutissimin A (Khanbabaee and van Ree, 2001). Gallotannins are composed of galloyl units or their meta-depsidic derivatives bound to different polyol-, catechin-or triterpenoid units (Landete, 2011). The hydrolysis of gallotannins produces glucose and gallic acid.

Chestnut tree (Castanea sativa Mill.)
Chestnut, Castanea sativa Mill., (Fagaceae) is the predominant sweet chestnut tree in Europe (Živković et al., 2009). Sweet chestnut has an important economic value in timber production, besides being a valuable food and feed source. Several chestnut-based products are on the food market (traditional dried chestnuts, chestnut flour, marrons glacés, frozen chestnut products, chestnut flakes, beer or liquors) and feed market (chestnut flour, chestnut extracts).
The high content of tannins also increases the value of sweet chestnut trees. The tannin concentration largely depends on the season, and the age and part of the tree. Živković et al. (2009) showed a higher content of total phenolic compounds and total condensed tannins in catkin (3.28%), chestnut bark (3%), the red internal seed shell (2.82%) and brown seed shell (1.19%). The bark contains a high tannin concentration, from 5 to 10%, due to its protective function against vermin and UV light. Krisper et al. (1992) reported that the tannin concentration of wood increases linearly with the age of the tree.

Quebracho tree (Schinopsis spp.)
Quebracho trees (Schinopsis spp., family Anacardiaceae) grow in South America, mostly in northern Argentina and eastern Paraguay. The most representative quebracho trees belong to Schinopsis lorentzii and Schinopsis balansae species, characterized by red hardwood. A third tree species is the white quebracho (Aspidosperma quebrachoblanco, family Apocynaceae). However, the 'true quebracho' is referred to as Schinopsis lorentzii and Schinopsis balansae from the Anacardiaceae family (Venter et al., 2012). These trees are converted into different products such as railroad ties, boards, beams, poles, piles, fence posts and cross-arms. Quebracho trees are not generally used for construction proposes due to their heavy weight and short length. Quebracho trees are principally used for their hardwood content of tannins, which are used in the tanning of high-grade leathers and adhesive manufacturing (Pizzi, 2019). Quebracho extract obtained from Schinopsis lorentzii contains 15-21% of pure tannin, whereas the extract obtained from Schinopsis balansae has a pure tannin content of 20-21% (Venter et al., 2012).

Extraction methods
The extraction process is similar to tannins extracted from chestnut. The heartwood is stripped of its bark and chipped. The obtained wood chips are extracted using hot water (100°C), with a bisulphite solution. The addition of bisulphite solution to the hot water increases the extraction rate of tannins (Venter et al., 2012). The quebracho extract is composed of 95% condensed tannin or proanthocyanidins and 5% water soluble sugars on a dry basis.
The most representative CTs in quebracho tannins are proanthocyanidin, typically the red colour of CTs is due to the reaction of the high temperature and aqueous acid (Venter et al., 2012). The first application of CTs tannins in leather, and more recently tannin applications as bio-based adhesives, has led to an increase in tannin use as bio-based materials in widespread industrial processes and sectors (Pizzi, 2019).
Several by-products are produced during industrialized tannin extraction, such as exhausted wood biomasses, which are reintegrated into the industrial process to produce pellets for heating and energy production. In addition, chestnut wood, leaves and shells can be recycled in the food and feed industries, and in pharmaceutical and cosmetic manufacturing. Chestnuts and quebracho trees are thus a valuable source of tannins not only for industrial purposes but also for nutrition. The renowned characteristics and properties of tannins can be applied from feed to food. These approaches are in line with the circular economy and offer a new life and alternative prospective for recovering these by-products with beneficial effects for the environment and the economy.

Benefits and challenges of tannin supplementation in pigs
Tannins have a considerable impact on swine health and productivity. Numerous field studies have highlighted their benefits and challenges (Huang et al., 2018). The effects of chestnut or quebracho tannin supplementation in swine are different due to the different commercial feed additives adopted that contain chestnut and quebracho tannin extracts (Van Parys et al., 2010;Bilić-Šobot et al., 2016;Bee et al., 2017).
The heterogeneity of commercial products is associated with the use of chestnut or quebracho individually or in mixtures with different percentages of chestnut and quebracho, and hence different amounts of HTs and CTs. The commercial products tested on pig are therefore extremely different also in terms of the concentrations of tannins in the extract which range from 54 to 82%, thus the presence of different percentages and types of HTs or CTs (Supplementary Table S1).
The beneficial effects of tannin supplementation in pig farming are related to their antimicrobial, antioxidant and radical scavenging, antiinflammatory activities and on the immune status (Huang et al., 2018). In swine farming, pathogenic bacteria such as E. coli are mostly associated with post-weaning diarrhoea (PWD), a multifactorial disease that occurs after weaning (Rossi et al., 2014a;Rossi et al., 2014b). The use of tannin extracts can be a valuable alternative for the control of PWD (Smeriglio et al., 2017;Huang et al., 2018;Girard and Bee, 2020). In fact, the bacteriostatic activity of tannins has been shown in both gram-positive (Listeria monocytogenes, Staphylococcus aureus, Bacillus subtilis and Enterococcus faecalis) and gram-negative bacteria (Citrobacter freundii, E. coli, Pseudomonas aeruginosa, Salmonella enteric ser. Typhimurium) (Smeriglio et al., 2017). Tannins have antioxidant and radical scavenger effects as also demonstrated by the study conducted by our group . Although HTs may enhance the antioxidant capacity of liver and plasma in animals (Huang et al., 2018), the mechanism underlying the tannin action as an antioxidant compound is not fully understood.
The anti-inflammatory activity and animals' immune response are mostly associated with tannins antioxidant activities. To date, the in vivo study conducted by Stukelj et al. (2010) administered to growing pigs 0.15% HTs associated with 0.15% four acids which did not show differences on immune response. Nevertheless, several studies conducted on other animals' species identified a positive immune response with tannins supplementation. In particular, in both in vivo and in vitro, the response of tannins on anti-inflammatory and immune status showed promising effects when animals were exposed to stressors such as lipopolysaccharide challenge, bacterial challenge, heat stress or intestinal cells exposed to hydrogen peroxide and dextran sodium sulphate (Park et al., 2014;Liu et al., 2018;Reggi et al., 2020). Tannins, HTs and CTs, are able to modulate intestinal pro-inflammatory cytokine expression acting also on gut barrier and tight junctions . However, there is a lack of in vivo studies on the action mode of tannins in the swine gut, especially in weaning and post-weaned piglets.
The bioactive characteristics of tannins can affect the palatability, digestibility and protein use of feed. The ability to bind proteins and carbohydrates in monogastric animals is associated with the antinutritional effects of tannins in reducing feed palatability (Bee et al., 2017). An increased concentration of proline in the parotid glands of animals treated with HTs from acorns may lead to the release of higher amounts of the tannin-protein complex in the saliva (Cappai et al., 2013). Proanthocyanidins, which are CTs, have a high affinity with proline-rich protein, and the strength of the interaction depends on both the nature of the protein and the proanthocyanidin molecule. An animal's ability to tolerate the antinutritional effects of HTs or CTs is thus an essential defensive mechanism that ensures its beneficial nutritional and extra-nutritional effects (Candek-Potokar et al., 2015). Animals adapted to tannin ingestion, such as ruminants, have therefore shown minimized negative effects, such as regular nutrient digestibility, better tannin absorption and minor antinutritional effects in general.
Animals not fully adapted to tannin ingestion, typically monogastric animals, are less able to tolerate the astringency effects triggered by the tannin-protein complex. Consequently, a higher astringency effect decreases the feed palatability and ingestion (Bee et al., 2017). However, several studies have shown discordant findings when tannins were evaluated on amino acid and protein digestibility (Antongiovanni et al., 2007;Galassi et al., 2019). The capacity of tannins to bind proteins is not necessarily negative. In fact, the formation of a tannin proteincomplex in the digestive tract may protect proteins, carbohydrates and lipids from oxidative damage during digestion (Cappai et al., 2013).
In general, digestibility is a key factor in determining tannins' effect at gastrointestinal level. Then, the biodegradation and the absorption of HTs along the gastrointestinal tract have been investigated by several studies, whereas the fate and absorption of CTs seems to be more complex due to their structural complexity (Mueller-Harvey, 2006). As reported by Reggi et al. (2020), the antimicrobial and antioxidant activity of HTs and CTs was reduced after in vitro digestion, which may be due to a lower bio-accessibility or to the excessive degradation of antimicrobial and antioxidant molecules. The authors, however, reported a beneficial effect of in vitro digested HTs and CTs when administered to experimentally damaged intestinal swine cells, suggesting that a trophic effect at intestinal epithelium occurred. In light of this, the bioavailability of tannins after oral supplementation plays a crucial role and should be considered and further investigated. Therefore, the mechanism of action of tannins or tannins degradation-molecules exploitation should be deeply investigated at intestinal level (tissue or cells).

Chestnut and quebracho tannin effect on growth performance and intestinal heath in pig
In the light of the activities and properties of tannins, identifying the correct application dose is essential in order to maximize the beneficial effects of tannins and minimize the antinutritional effects on animal growth performance. The selection criteria of the studies discussed in this review were i) studies performed on cross bred pigs from postweaning to finishing phases; ii) studies evaluating diets supplemented with tannins from chestnut or quebracho and iii) study outcomes evaluating zootechnical performance or intestinal health.
Effects of Chestnut and quebracho tannin supplementation during the postweaning phase

Growth performances
The anti-inflammatory and antibacterial activities of tannin supplementation could enhance growth performance and alleviate PWD, one V. Caprarulo, C. Giromini and L. Rossi Animal 15 (2021) 100064 of the greatest disorders to occur during the first 2 weeks after weaning (Rossi et al., 2013 and2014a;Huang et al., 2018). The studies we reviewed adopted a supplementation of tannins ranging from 0.11 to 3%; hence, the tannin dosage was considered low for <1% of tannin inclusion, medium for ≥1-2% and high for ≥3%. Several studies have been performed on tannin supplementation during the post-weaning phase ( Table 1). The inclusion of low tannin doses of 0.11, 0.23 and 0.45% HTs, for a period of 28 days, did not influence the animal's live weight (BW) and average daily feed intake (ADFI; Biagi et al., 2010). However, the same study reported a higher average daily gain (ADG) and increased feed efficiency throughout the 28 days of the trial compared with the control group (no tannin supplementation).
The inclusion in the diets of a medium tannin dose of 1% HTs/CTs had no effect on BW and feed efficiency ratio (Girard et al., 2018). However, tannin supplementation at 2% HTs/CTs showed a positive effect on daily feed intake and ADG (Girard et al., 2019). The use of medium tannin doses of 2% HTs/CTs may result in improved daily feed intake and average body gain, whereas doses under 1% HTs/CTs seem not to affect animals' growth performance. It is possible that the application of low tannin doses, especially a dietary inclusion of below 0.5% HTs/CTs, could have a limited effect on animals. This limitation could establish a cut-off limit for tannin efficacy in relation to the dose meriting further investigation.
The combination of tannins with other bioactive compounds could be beneficial for enhancing animal performance. The combined application of organic acids within pig feed led to better growth and affected intestinal microbiota. The administration of 0.19% HTs and 0.16% of fatty acids in the diet improved BW, ADG and feed conversion ratio (FCR; Brus et al., 2013b). This positive effect on growth performance could be related to the effect of the organic acids in decreasing gastric pH, improving nutrient digestion and acting as an energy source for the gastrointestinal tract.
The synergistic effect of tannins with sodium salicylate, a nonsteroidal anti-inflammatory drug, was tested in weaned piglets (Girard et al., 2019). However, the use of sodium salicylate in combination with a medium tannin dose of 2% HTs/CTs did not affect growth. Therefore, the synergistic effect of tannins and organic acids or other compounds merits further study. Organic acids may alleviate the negative effects of tannins on feed intake, growth and feed efficiency.
The medium supplementation of tannins seems to be more effective on the growth of post-weaned piglets compared to a low inclusion (Biagi et al., 2010;Frankič and Salobir, 2011;Brus et al., 2013b;Girard et al., 2018 and2019). However, a low dose combined with 0.16% of fatty acids showed beneficial effects on animal performance. Most of the beneficial effects of tannins are related to an improved ADG and ADFI or FCR associated with the medium tannin dose. Literature data show that low (< 1%) or medium (≥ 1-2%) doses of tannin supplementation may positively influence growth during the post-weaning phase.
Exactly how tannins improve performance is not fully understood. In the studies here considered, the duration of tannin supplementation seems to vary drastically, from day 14 to day 104. Despite the positive effect of the combination of tannins and organic acids, where the trial lasted 104 days (Brus et al., 2013b), the beneficial effects were detected with a 14-or 28-day supplementation (Biagi et al., 2010;Frankič and Salobir, 2011;Girard et al., 2018 and2019).
Another important factor which can modulate animal response to tannins may be the feeding regime adopted in the different studies (ad libitum vs restricted) and the basal diet administered which may contain different level of tannin from basal ingredients. The majority of studies during the post-weaning period adopted ad libitum feeding. However, Frankič and Salobir (2011) used a feed-restricted regimen, which is not common practice in piglet trials, and the obtained results may be more associated with the tannin dosage, which ranged from 0.075 to 0.30%, rather than the duration of supplementation. The basal diets adopted during the post-weaning phase in the different studies analysed contained different amount of basal ingredients, such as corn meal, wheat meal, barley meal and wheat starch which are intrinsically rich in tannin. The utilization of tannin-rich feedstuffs can contribute to increase the overall concentration of tannins in the diets. The latter Table 1 Effects of tannins on pigs' zootechnical performances during the post-weaning phase. factor is not always considered in animal trials but it can hide or modulate animal performance and response.

Intestinal health
The optimal growth performance in pig is directly linked with the health of the gastrointestinal system. The post-weaning phase is critical in the lifetime of piglets, during which gastrointestinal morphology and physiology undergo several changes. Bioactive compounds such as tannins could positively affect these changes at physiological and microbial levels ( Table 2). There are few in vivo studies related to the effect of tannins on intestinal health during the post-weaning. The in vivo administration of low tannin doses, at 0.11, 0.23 and 0.45% HTs, did not alter the Enterococcus spp. count, but tended to increase the Lactobacillus spp. count in the jejunum and coliforms in the caecum (Biagi et al., 2010). In addition, no effects of tannins were detected on villous height and crypt depth of the intestinal mucosa. The same author reported that an in vitro experiment did not show the effects of tannins (0.75, 1.5 and 6 HTs g/l concentrations in the medium) on coliform count, nevertheless coliforms increased with tannins at 3 HTs g/l. The Lactobacillus spp. count was significantly reduced by the highest tannin concentration (6 HTs g/l), and the Enterococcus spp. count increased significantly. In vitro fermentation experiments have shown that tannins decreased the total gas production, ammonia concentrations and total volatile fatty acids (acetic acid, propionic acid, iso-butyric acid, n-butyric acid, iso-valeric acid and the n-butyric to iso-butyric acid ratio).
The effects of tannins on intestinal morphology were not significant to induce a beneficial improvement. Thus, these findings on Lactobacillus spp. and coliforms seem contrasting. Girard et al. (2018) demonstrated that HTs/CTs did not affect Escherichia coli abundance in faeces on infected post-weaning piglets with Escherichia coli (ETEC) F4 stain supplemented with a medium 1% dose of HTs/CTs (Girard et al., 2018). Nevertheless, the supplementation of a medium 1% dosage of HTs/CTs did not prevent outbreaks of diarrhoea or ETEC shedding. However, the prevalence of diarrhoea, faecal score and days of diarrhoea was lower in the HTs/CTs group. Similar results were found when a medium dose of HTs/CTs was used (Girard et al., 2019); piglets challenged with Escherichia coli (ETEC) had a lower average faecal score, a lower percentage of piglets with diarrhoea and less ETEC shedding in faeces. Both medium tannin doses of 1% HTs/CTs and 2% HTs/CTs decreased the incidence of diarrhoea in weaned piglets. However, only 2% of HTs/CTs inclusion in the diet was able to decrease E. coli shedding.
The combination of a low dose of HTs with organic acids decreased faecal Escherichia coli, Campylobacter spp. count and increased the lactic acid bacteria count (Brus et al., 2013b). These findings could be explained by the antagonist effect of lactic acid bacteria on Escherichia coli and Campylobacter spp. growth. However, the addition of organic acids used in the Brus et al. (2013b) and the duration of tannin supplementation (104 days) could have influenced the results. These results are interesting because they highlight the beneficial effects in reducing the bacterial count and E. coli shedding. The low and medium supplementation of HTs/CTs or HTs seems to reduce some bacteria populations.
The mechanisms behind the in vivo antimicrobial effects of tannins at the intestinal level, especially how tannins modulate intestinal health in weaned piglets, are not fully understood. Some hypotheses regarding the antimicrobial activity of tannins include the presence of ellagitatannin which showed protein binding, enzyme inhibition, substrate deprivation, complex formation with cell walls, membrane disruption and metal ions (Girard and Bee, 2020). The antimicrobial activity of tannins may be attributed to the oxidation of tannins and the liberation of hydrogen peroxide. In addition, the high affinity of tannins to bind proteins could increase the number of hydroxyl groups, subsequently increasing the antimicrobial activity (Mueller-Harvey, 2006).
The tannin dosage used during the post-weaning phase needs further investigation, since the rate that tannins move to the small intestine is not fully understood. Hydrolysable tannins could also be hydrolysed in the stomach and in the small intestine, thus freeing several metabolites, such as ellagic acid. Hydrolysable tannins metabolites may also be involved in the antibacterial activities. In addition, the studies adopted HTs alone or the combination of HTs and CTs; thus, the synergic effect of HTs and CTs cannot be excluded.
Overall, the application of chestnut and quebracho based supplements in the intestinal health of post-weaning animals requires further studies to test the correct dosage, time of supplementation and mechanism of actions. In this regard, the use of a combination of in vitro experimental models, such as those described by Giromini et al. (2019) and Reggi et al. (2020), and animal trials to test the same HTs or HTs/CTs products could provide a more complete overview of both mechanisms of action in vitro and the efficacy at gastrointestinal level in animals. Finally, there are currently few studies on the modulation of pigs' microbiota by tannins during post-weaning. There are few in vivo studies on these effects and the intestinal population studied is restricted to the main pathogen bacteria in weaned piglets (Biagi et al., 2010;Brus et al., 2013b;Girard et al., 2018;Girard and Bee, 2020). In general, tannins' antimicrobial activity could increase the knowledge on the effects of tannins on specific pathogens. However, the effects of HTs/CTs or HTs on the entire microbiome need to be clarified which could then lead to new nutritional strategies for the best use of this promising additive.
Effects of Chestnut and quebracho tannin supplementation during the growing and finishing phases

Growth performances
In intensive pig farming, during the growing and finishing phases, nutrition plays a key role in achieving better nutrient abortion and digestibility, better feed conversion and efficiency, and improving growth performance and health. Tannins derived from chestnut and quebracho could improve animals' performance and health status due to their antimicrobial and antioxidant effects ( Table 3). The use of low doses has revealed different response on animal performance. The low tannin dose adopted by Antongiovanni et al. (2007) (0.25 and 0.50% HTs) did not impact on ADG and FCR. Similar results were reported by Prevolnik et al. (2012), as 0.20% of HTs supplementation did not modulate ADG, ADF or FCR in pigs. In line with this, low dose HTs effects on animal performance were found in the trial conducted by Galassi et al. (2019)). However, Brus et al. (2013a) reported an increased ADG in the finishing phase (90-120 kg) and throughout the entire trial (30-120 kg), with a low supplementation of HTs (0.20%). Bee et al. (2017) reported that FCR was positively affected by medium, 1.5%, and high, 3%, doses of HTs/CTs, while BW, ADG and ADFI were not influenced. Conversely, a high 3% dose of HTs decreased ADFI without modifying the BW, ADG and FCR (Candek-Potokar et al., 2015).
Other studies have reported the administration of tannins together with other compounds, such as natural extracts rich in polyphenols (Ranucci et al., 2015), acids (Stukelj et al., 2010) and polyunsaturated fatty acid (Tretola et al., 2019). However, the combination of oregano extract and HTs did not enhance pigs' performance, whereas it increased the antioxidant status and lipid oxidation (Ranucci et al., 2015). A low HTs dose combined with the supplementation of 0.15% of a mixture of four acids did not affect the growth performance of pigs (Stukelj et al., 2010). Tretola et al. (2019) reported that the use of a high HTs/CTs dose (3%) in combination with 2% PUFA did not improve animal performance. The synergistic effect of tannins with other compounds could increase the interest in view of the various beneficial activities of several compounds, especially phenolic compounds. V. Caprarulo, C. Giromini and L. Rossi Animal 15 (2021) 100064 Most studies adopted commercial additives based on tannins extracts; however, an alternative source of tannins is the direct use of dried chestnuts or chestnut meal in animal diets. Chestnut meal contains over 75% of HTs, and when used in pig diets slightly increased the tannin content in the diet. The inclusion of chestnut meal from 5 to 25% translated into a tannin concentration in the diet ranging from 0.17 to 0.19% on a DM basis (Lee et al., 2016). Thus, supplementing dried chestnut or chestnut meal could be comparable to a low dose of tannin extract.
The supplementation of 0.3 and 0.5% (DM) chestnut meal improved the average BW and feed efficiency (Joo et al., 2018). Lee et al. (2016) reported an increased DM intake for the experimental group that received 10% of chestnut meal. In addition, the beneficial effect of tannins was highlighted by supplementing 15% of dried chestnuts. The tannin group had a greater BW compared to the control group (De Jesus et al., 2016). In contrast with other studies, which showed a marginal or no effect of tannins supplementation on zootechnical performance, the use of dried chestnuts or chestnut meal revealed promising results.
In the studies reported in Table 3, the duration of tannin supplementation during the growing and finishing phases varies drastically, from day 21 to day 270. Tannin supplementation during this life stage revealed dissimilar effects on growth performance. In particular, tannins extract did not increase the zootechnical parameters except for ADG. A 30-day study by Brus et al. (2013a) adopted a low dose of HTs, while other studies supplemented similar higher doses of HTs and HTs/CTs (from 0.15 to 3%) for a longer period.
Results of studies conducted during the growing and finishing phases have thus shown heterogeneous responses probably related to the duration of the tannin supplementation, type of tannin product (extracted or chestnut meal) used, age of animals and dietary basal ingredients naturally rich in tannin. In fact, tannins are present in several feedstuffs and ingredients for animal nutrition, such as corn, wheat and barley. Thus, tannin-rich feedstuffs could increase the concentration of tannins in the diet fed to animals.
It is possible that the higher 3% dose of HTs and HTs/CTs used in some studies (Candek-Potokar et al., 2015;Bee et al., 2017;Tretola et al., 2019), in addition to the basal diet content of tannins, could increase the tannin intake resulting in no effect on growth performance.
Among the studies reported in the present review, only Lee et al. (2016) reported the basal diet concentration of tannins which were 0.16% of DM. As per post-weaning phase, also in finisher studies investigating the role of HTs and CTs on growth performance and intestinal health, the concentration of HTs and CTs in the basal diet should be taken into account. Further investigations are needed to evaluate the correct administration of tannins irrespectively of the fattening and finishing period, in relation to the basal diet content of tannins.

Intestinal heath
The intestinal morphology and histological characteristics of the small intestine may be modulated by HTs supplementation (Table 4; Brus et al., 2013a;Candek-Potokar et al., 2015;Bilić-Šobot et al., 2016). Low tannin doses of 0.20% HTs showed no effects on the height of mucosa, height of villi, height of epithelium and proportion of necrosis (Brus et al., 2013a). In this latter study, necrosis was observed in all experimental groups, probably due to the negative correlation with daily gain. Normally, intestinal necrosis leads to a decreased feed intake and subsequent lower body gain. However, animals in the HTs group showed an equal level of necrosis as the control group, but a higher daily gain. It is not clear how HTs affect intestinal morphology, though it seems that they have a trophic effect in intestine maintenance in terms of absorbing nutrients.
Medium and high HTs supplementation from 1 to 3% affected the morphology of the duodenum (Bilić-Šobot et al., 2016). In particular, 3% of HTs supplementation increased the villus height and villus perimeter, while 1 and 3% increased mucosal thickness (Bilić-Šobot et al., 2016). Only the duodenum was affected by HTs supplementation, which is an important digestive site in which an important stage of digestion takes place together with the absorption of fat and fat-soluble vitamins. Digestion and absorption in the small intestine are the main functions in which villous height and crypt depth are involved. The higher villus height and perimeter thus increased the surface area of nutrient absorption. However, Candek-Potokar et al. (2015) showed the positive effect of HTs supplementation on growth performance, with 1 and 2% leading to an increased daily feed intake. These results were explained by the higher villus height and perimeter, which led to an increased nutrient absorption. There are few in vivo studies on intestinal morphology, and data on histological evaluations of tannin administration are not sufficient. It is therefore not possible to hypothesize what level of tannin dose could influence intestinal morphology. Van Parys et al. (2010) showed that Salmonella typhimurium shedding and colonization were not affected by HTs in Salmonella-challenged growing pigs. However, the same author demonstrated in vitro bacteriostatic and bactericidal effects against Salmonella typhimurium. The outcomes obtained in vitro and in vivo were influenced by different experimental conditions (in vitro condition vs in vivo condition). In fact, in vitro experiment condition cannot fully mimic the complexity of the interactions that occur within the entire organism. In the growing and finishing phases, the effectiveness of the antibacterial activity of tannins during in vivo trial is still unclear.
Another important intestinal aspect is how tannins modulate intestinal microbiota. High doses of HTs/CTs (3%) were shown to modulate the intestinal microbiota of growing pigs (Tretola et al., 2019). The alpha diversity analysis revealed that the bacterial community index (Chao1) and operational taxonomic units (OTUs) and phylogenetic diversity of pigs' microbiota decreased with HTs/CTs supplementation. However, HTs/CTs increased the Oscillospira genus of the Ruminococcaceae family, while reducing the Lactobacillales order, Streptococcaceae and Veillonellaceae families. Thus, HTs/CTs were able to modulate some bacterial genuses and families that are considered beneficial in preventing intestinal inflammation together with other families that are usually harmful for intestinal integrity and health. However, the Lactobacillales order was reduced by HTs/ CTs supplementation. Compared to the other studies reported in this review, considering the growing and finishing phases, Tretola et al. (2019) were the only study in which the combination of HTs and CTs was tested. Nowadays, we cannot exclude a synergistic effect of these two type of tannins. In vitro studies on E. coli reported this possible synergistic effect ; however, the in vitro response is not fully understood.
The immune effect of tannin supplementation was not deeply investigated. Considering the studies reported in this review, only Stukelj et al. (2010) investigated the immune effect of tannins on growing and finishing pigs. The blood concentration of neutrophils decreased, while lymphocytes and eosinophils concentration increased after 21 days of 0.15% HTs + 0.15% four acids supplementation. In this study, the synergistic effect of tannins with fatty acids and the effect of tannins alone should be further investigated. Hence, no reports are available on HTs or CTs effect on immune status on pigs; this lack of data should be investigated in order to better clarify the effect of HTs and CTs on immune response and in particular on expression of related genes.
The in vitro and in vivo results highlighted the promising effects of HTs or HTs/CTs on different types of bacteria, such as the Salmonella typhimurium, Oscillospira genus and the Streptococcaceae and Veillonellaceae families. Aside from the effectiveness of HTs or HTs/CTs on numerous bacteria, both gram-positive and gram-negative, there is a lack of studies on the intestinal ecology of growing and finishing pigs. Frequently, the in vitro effects of HTs or HTs/CTs on the intestinal bacteria population are not in line with the results from in vivo studies. Thus, further investigations are needed in order to clarify HTs or HTs/ CTs activity in pigs.

Conclusion
In conclusion, the use of hydrolysable and condensed tannins in swine nutrition could have beneficial effects on the growth performances and gastrointestinal health status of both post-weaning and growing pigs. However, the effectiveness of tannin supplementation in weaned and finishers seems to be highly related not only to the dose administered to animals but also to the duration of supplementation, the presence of other sources of tannins in the basal diet, the feed regimen adopted (ad libitum vs restricted) and to the synergistic effect of hydrolysable and/or condensed tannins with other supplements. Also, to date, the majority of the in vivo studies performed have not fully elucidated the properties of tannins on swine immune status and on the intestinal microbiota. All the latter points require additional investigations.
Finally, although the establishment of the effective tannin supplementation protocol in swine requires further studies, they seem to enhance productivity and animal health and therefore, they could be adopted as feed additives.

Supplementary materials
Supplementary data to this article can be found online at https://doi. org/10.1016/j.animal.2020.100064.

Ethics approval
This review did not require any ethical approval.

Data and model availability statement
None of the data were deposited in an official repository.

Declaration of interest
None.