Animal 15 (2021) 100145 Contents lists available at ScienceDirect Animal The international journal of animal biosciences Complementarity between microhistological analysis and PCR-capillary electrophoresis in diet analysis of goats and cattle using faecal samples J. Pareja a,⁎, C. Espunya b, E. Baraza c, J. Bartolomé a a Department of Animal and Food Sciences, Universitat Autònoma de Barcelona, 08193, Bellaterra, Barcelona, Spain b Independent researcher, 08290 Cerdanyola del Vallès, Spain c Department of Biology, Universidad de las Islas Baleares, 07122 Palma de Mallorca, Mallorca, Spain a r t i c l e i n f o a b s t r a c t Article history: An evaluation is made of the complementarity between two non-invasive techniques, cuticle microhistological Received 8 July 2020 analysis (CMA) and PCR-capillary electrophoresis (PCR-CE) DNA-based analysis, for the determination of herbi- Received in revised form 9 November 2020 vore diet composition from faecal samples. Cuticle microhistological analysis is based on the different microan- Accepted 10 November 2020 atomical characteristics of the epidermal fragments remaining in the faeces. The PCR-CE technique combines Available online 15 December 2020 PCR amplification of a trnL(UAA) genomic DNA region with amplicon length determination by CE, with this Keywords: length being characteristic for each species or taxon. A total of 37 fresh stool samples were analyzed, including Bos taurus 16 from feral goats (Capra hircus) from the Tramuntana mountain range (Mallorca, Baleares) and 11 from Capra hircus Bruna dels Pirineus cattle breed (Bos taurus) from the surrounding Montserrat mountain range (Barcelona, Faecal analysis Spain). All the animals were in a free grazing Mediterranean pine habitat, dominated by Aleppo pine (Pinus Free grazing halepensis). The results showed that both techniques detected a similar number of plant components in the faeces Plant epidermis of goats and cows. In the case of goats, a positive correlation was obtained between the percentage of samples in which a particular taxon is detected by CMA and the percentage of samples in which that taxon is detected by PCR-CE. This correlation was not observed in the case of cows. It is concluded that PCR-CE is a fast and reliable method to detect the different plant components in the faeces of herbivores. However, it cannot be considered as an alternative to CMA, but as a complementary method, since both techniques can detect some taxa that are not detected by the other technique. In addition, CMA detected the presence of the different taxa in a greater number of samples, and at the same time, it enables quantitative data to be obtained for plant diet composition. The species of herbivore also seems to influence the results obtained by PCR-CE, so more studies are required to address this aspect. © 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/). Implications and Wegge, 2006), including the bite or time counting methods (Pinto et al., 2014), near IR reflectance spectroscopy (Coates and Dixon, Diet analysis of animals is a relevant aspect in environmental man- 2008), quantification of n-alkanes in faeces (Ferreira et al., 2007), iso- agement. In this study, two techniques were compared, the cuticle tope stability analysis (Sponheimer et al., 2003; Codron and Brink, microhistological analysis and polymerase chain reaction-capillary elec- 2007), esophageal fistula sampling (Bautista et al., 1996), rumen con- trophoresis DNA-based analysis, in feral cows and feral goats in a Med- tent sampling (Norbury and Sanson, 1992; Bertolino et al., 2009), iterranean environment. We found that both techniques had different microhistological analysis of faeces (Stewart, 1967; Bartolome et al., results in both species, demonstrating a complementary effect of both 1995) and, more recently, molecular analysis based on DNA (Pegard techniques in the analysis of diet. et al., 2009; Soininen et al., 2009). Each method has a number of advan- tages and limitations when interpreting eating habits (Henley et al., Introduction 2001). Among these techniques, cuticle microhistological analysis (CMA) is the most used method to study the diet in herbivores, al- Many analysis techniques have been used in the past years to quan- though it requires a considerable training effort (Holechek and Gross, titatively or qualitatively evaluate the composition of the herbivore diet, 1982), and there could be errors due to epidermal degradation and the most relevant being: the direct observation of the animal (Shrestha the microhistological similarity of some plant epidermis (Henley et al., 2001). For this reason, it is important to carry out a correct processing ⁎ Corresponding author. of the samples and the personnel must first know the microhistological E-mail address: javimev@gmail.com (J. Pareja). fragments of reference, such as the shape and cell arrangement, https://doi.org/10.1016/j.animal.2020.100145 1751-7311/© 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/). J. Pareja, C. Espunya, E. Baraza et al. Animal 15 (2021) 100145 stomata, trichomes and glands (Chetri, 2006). On the other hand, mo- was homogeneously distributed onto three slides with 50% glycerine lecular analyses based on the presence of genomic DNA remnants of and was sealed with DPX resin (Herter Instruments, Barcelona, Spain). ingested plants in faecal samples have been shown to be a useful and To obtain the reference epidermis collection, the leaf samples were sub- rapid tool to determine the components of the herbivore diet jected to a scalpel scrape until the epidermis of both faces (abaxial and (Taberlet et al., 2007; Valentini et al., 2009). Within the molecular tech- adaxial) was obtained. This epidermis was mounted onto slides with niques, the PCR-CE technique (PCR-Capillary Electrophoresis), based on glycerine. The most recalcitrant species, in which epidermal tissue the PCR amplification of a trnL(UAA) gene fragment followed by the de- could not be obtained by this method, were subjected to acid digestion, termination of the length of the amplicon by CE, has been used to qual- following the same protocol as in the case of faecal samples. Prepara- itatively estimate the presence of a small number of species in a tions were observed under an optical microscope (Motic BA210, controlled diet for chamois (Espunyes et al., 2019). However, this tech- MoticEurope SLU, Barcelona, Spain) at 100 and 400 magnifications, nique has not yet been used to determine the nature of the plants con- equipped with a Moticam 2300 camera. By using the Motic Images sumed by herbivores in free ranging situations. Plus 2.0 software, images of cellular forms, trichomes, and stomata The objectives of this work were to determine if the technique of the were obtained from each plant sample, as a reference shown in Supple- PCR-CE is able to identify the taxa of the diet in field grazing animals and mentary Figure S1. Subsequently, the faecal samples were then ob- to compare the results obtained by PCR-CE with those obtained by CMA, served under the same microscope, making linear transects until a by evaluating the complementarity between both techniques. For this total of 200 epidermal fragments per sample were identified. The evaluation, faecal samples of animals that graze in Mediterranean envi- microhistological analysis was performed at the animal production lab- ronments were used, such as cows (Bos taurus) of the Bruna dels Pirineus oratory of the Veterinary Faculty of the Universitat Autònoma de Barce- breed in the surroundings of the Montserrat mountain, Barcelona, and lona (UAB). This laboratory is part of the ASFAC-LAB quality programme feral goats (Capra hircus) on the island of Mallorca. consisting of an exercise of intercomparison of laboratories performing analysis of feed and raw materials for animal feed (http://www.asfac- Material and methods lab.com/en/). Study area Molecular analysis (PCR-capillary electrophoresis) This study was conducted with faecal samples of goats (Capra hircus) Genomic DNA (gDNA) from the most abundant plant species in each and cows (Bos taurus) that were fed in the undergrowth of Mediterra- area studied and from faeces samples was extracted with the DNeasy nean pine forests. Samples of goats were collected in the Tramuntana Plant Mini Kit and the QIAamp DNA faeces Kit (QIAGEN, Hilden, mountain range, on the island of Mallorca (39°48′28″N2°47′37″E), Germany), respectively, following manufacturer's instructions. Geno- where there is a population of feral animals (Vives and Baraza, 2010). mic DNA of plants and faeces was diluted in 100 μl of Milli Q water The mean annual temperature and precipitation in the mountains are and stored at − 20 °C. DNA concentration was measured with a 18 °C and 650 mm, respectively. The cows were located around the NanoDrop 1000 Spectrophotometer (Thermo Scientific, Waltham, Montserrat Mountain Natural Park in the province of Barcelona USA), and the quality of the gDNA was analyzed on a 0.8% agarose gel (41°40′28”N1°46′33″E), where the mean annual temperature is of electrophoresis in 1 × TAE buffer (40 mM Tris pH 7.6, 20 mM acetic 13 °C, and rainfall of 610 mm. In this case, the animals spend the acid, 1 mM EDTA). The sequences corresponding to the chloroplast winter months and part of the spring in the area, then they transhu- gene trnL(UAA) were amplified by PCR from the oligonucleotides trnL- mance to the Pyrenees. The vegetation of both areas is dominated by G (5’gggcaatcctgagccaaatc3’) and trnL-D (5’ggggatagagggacttgaac3’) trees of Aleppo pine (Pinus halepensis M.) with a Mediterranean scrub (Taberlet et al., 2007), using plant and faeces gDNA as a template. The undergrowth with abundance of Erica multiflora L., Pistacia lentiscus L. PCRs contained 1 × MyTaq reaction buffer (Bioline Reagents Ltd., and Rosmarinus officinalis L. The herbaceous stratum is dominated by London, United Kingdom), oligonucleotides at a concentration of 0.4 the poaceae of the genus Brachypodium and, in the case of Mallorca, μM each (Stab Vida, Caparica, Portugal), 1 U MyTaq DNA polymerase also by Ampelodesmos mauritanica (Poiret) T. Durand et Schinz. and 20 ng of gDNA, in a final reaction volume of 20 μl. The trnL-D oligo- nucleotide was labelled with 6-FAM (carboxyfluorescein) fluorochrome Sample collection to allow fluorescent detection by CE. The amplification reaction com- prised different steps: initial denaturation of the gDNA at 95 °C for 1 Faecal samples were collected between the winter of 2016 and min, followed by 35 cycles of denaturation at 95 °C for 15 s, annealing spring of 2017. Goat faeces were obtained directly from the rectum, at 56 °C for 15 s and elongation at 72 °C for 45 s, ending with a final elon- from the animals killed during the culls made by the local administra- gation step at 72 °C for 3 min. The amplifications were performed on a tion. Cow faeces were collected right after their defecation. In order to MJ Research PTC-100 Thermal Cycler (MJ Research Inc., Mississauga, prepare a reference collection for microhistological and molecular anal- Canada). The amplification results were visualized on a 1.5% agarose ysis, 100 g of leaves of the most abundant plant species in each study gel electrophoresis in 1× TAE. area was collected. A total of 27 faecal samples were collected, 11 The length of the amplicons was determined by CE using a Genetic from cows and 16 from feral goats. The samples were labelled according Analyzer 3130xl system (Applied Biosystems, Waltham, USA), using to the ear tag number of each cow. To prevent DNA degradation, sam- the LizS600 marker as standard length. The results were obtained as ples were kept refrigerated during the transport from the field to the electropherograms where the amplified DNA fragment appeared as a laboratory and then stored at −20 °C until analysis. fluorescence peak in the corresponding length. The electropherograms were analyzed using the Peak Scanner version 2.0 software (Applied Cuticle microhistological analysis Biosystems, Waltham, USA). The components of the diet were identified by comparing the lengths of the different amplicons obtained from the To determine the botanical composition of the diet in faeces by CMA, analysis of the faeces of cows and goats, with the reference collection the procedure used by Stewart (1967) was followed. The faecal samples established from the taxa considered in Table 1. were crushed in a mortar, and then 0.5 g of sample was placed in a test The molecular analysis was performed at the Genomics and Bioin- tube, where the content was digested with 3 ml of 65% concentrated ni- formatics service of the UAB, which has a quality management system tric acid in a thermal bath at 80 °C for 2 min. This content was diluted in based on the ISO9001: 2015 standard and complies with the UAB 250 ml of distilled water and successively filtered through two sieves of code of good practice in research (https://sct.uab.cat/genomica- 1 and 0.125 mm pore size. The material obtained in the second sieve bioinformatica/en). 2 J. Pareja, C. Espunya, E. Baraza et al. Animal 15 (2021) 100145 Table 1 Some species that belong to the same genus had the same amplicon Length of the amplicons, in base pairs, of the plant species used as a reference to know the length, and therefore, they were grouped into genus: Dorycnium diet in cows and feral goats using the PCR-Capillary electrophoresis technique. pentaphyllum and Dorycnium hirsutum (Dorycnium sp.); Brachypodium Species Amplicon length (bp) retusum and Brachypodium phoenicoides (Brachypodium sp.); Cistus Aphyllanthes monspeliensis 647 albidus, Cistus monspeliensis and Cistus salviifolius (Cistus sp.); Erica Carex sp. 598 multiflora and Erica arborea (Erica sp.); Quercus coccifera and Quercus Smilax aspera 540 ilex (Quercus sp). Taxa in which the species could not be determined Lolium perenne 524 at the time of collection were also considered at the genus level, as Dorycnium sp. 520 Brachypodium sp. 515 was the case of Carex sp. and Ononis sp. Finally, species of a different Osyris alba 504 genus or family but with identical amplicons were considered in the Ampelodesmos mauritanica 496 same group (Buxus balearica and Genista scorpius; Genista lucida, Cneorum tricoccon 495 Quercus sp. and Sanguisorba minor; Olea europaea and Phillyrea Rhamnus alaternus 494 Arbutus unedo 492 angustifolia). The mean number of taxa identified per sample was higher Cistus sp. 489 when CMA was employed in both species, cows and goats (Table 2). Rubia peregrina 485 Table 3 shows the percentages of epidermal fragments identified in Chamaerops humilis 484 the faeces of cows together with the percentages of individuals in which Psoralea bituminosa 483 a specific taxon analyzed by CMA and/or PCR-CE appears. Cuticle Buxus balearica/Genista scorpius 479 Erica sp. 477 microhistological analysis succeeded in identifying 19 taxa, while 20 Bupleurum rigidum 472 were identified by PCR-CE. Some taxa were only detected with only Rosmarinus officinalis 461 one of the techniques. These were Aphyllanthes monspeliensis, Carex Rubus ulmifolius 460 sp. and Rosmarinus officinalis by CMA and Ephedra fragilis, Genista scor- Genista lucida/Quercus sp./Sanguisorba minor 458 pius, Hieracium pilosella and Osyris alba by PCR-CE. Eight taxa appeared Pinus halepensis 456 Globularia alypum 445 in a greater proportion using CMA technique, and six using PCR-CE. The Olea europaea/Phillyrea angustifolia 444 dominant taxon in the diet of cows was Brachypodium sp., which was Pistacia lentiscus 425 more than a quarter of the diet. The PCR-CE technique forced to group Hieracium pilosella 418 different species that has the same number of amplicon bps, as is the Ononis sp. 395 Juniperus oxycedrus 265 case of Genista lucida, Quercus sp. and Sanguisorba minor, and in the Ephedra fragilis 260 case of Olea europaea and Phillyrea angustifolia. In both cases, CMA bp = Base pairs. only detected fragments of one of the taxa, specifically Quercus sp. and Olea europaea. In addition, considering that Genista lucida is not in the Montserrat study area, this species would no longer be part of the first group, in the case of cows. Statistical data processing Table 4 shows the percentages of epidermal fragments identified in the faeces of goats, together with the percentages of individuals in The analyses were performed separately for each animal species. For which a specific taxon analyzed by CMA and/or PCR-CE appears. In CMA analyses, the abundance of each taxon in the diet was determined goat faeces, a total of 17 taxa were identified by CMA and 15 by PCR- as a percentage of epidermal fragments identified in the faeces. A gener- CE. Smilax aspera was only detected by CMA, and Genista lucida only alized linear mixed model of binomial distribution was performed in by PCR-CE. Most taxa appeared in a greater proportion of individuals order to determine the relationship between % of epidermal fragments using CMA, and only Arbutus unedo using PCR-CE. As in the case of identify in each sample and the probability of being detected through cows, from the group formed by Genista lucida, Quercus sp. and PCR-CE analysis. The percentage of fragments quantified was consid- Sanguisorba minor by PCR-CE, only Quercus sp., were identified by ered as a fixed predictive variable, the probability of being detected by CMA. On the contrary, from the group of Olea europaea and Phillyrea PCR-CE as a response variable, and each species as a random factor (B angustifolia, CMA distinguished fragments of both. (nt/x) = α + β1 percentage of epidermal + γ specie + ε). The glmer No significant correlation was found between the percentage of function from the lme4 (Version 1.1–8) package of R version 3.6.2 samples in which a particular taxon is detected by CMA with the (Bates et al., 2015) was used. percentage of samples that the taxon is detected by PCR-CE no for Because of the lack of normality, a nonparametric Spearman correla- cows (ρ = 0.17; P = 0.44 Spearman correlation test) neither for goats tion was used to analyze the relationship between the percentage of (ρ = 0.43; P = 0.06 Spearman correlation test). Similarly, the percent- samples which detected each taxon by PCR-CE and the percentage of age of fragments of a taxon detected by CMA does not appear to have a sample which was detected by CMA. In addition, the overall composi- significant effect on the probability that taxon is detected by PCR-CE in tion of the analyzed diet was compared with each of the techniques cow faeces (P = 0.4 GLMM-binomial, Fig. 1). But contrary, in the case of using non-metric multidimensional scaling. In this analysis, each sam- goats, this effect is significant (P = 0.006 GLMM-binomial, Fig. 1). ple was considered separately and the distance matrix was calculated based on the binomial distance of Jaccard. The monoMDS function from the vegan package (Version 2.5-2) of R version 3.6.2 (Oksanen et al., 2018) was used. The script used for the statistical analysis is avail- Table 2 able as Supplementary Material S1. Number of taxa identified per faecal sample in cows and feral goats (maximum, minimum and mean) by each technique (CMA and PCR-CE). Results Taxon/sample Cows Feral goats CMA PCR-CE CMA PCR-CE By means of the molecular technique of PCR-CE applied to the 39 se- Maximum 16 13 14 8 lected taxa, it was possible to amplify a fragment corresponding to the Minimum 11 3 10 2 gene trnL(UAA) in all of them, with lengths varying between 647 and Mean 13 9 12 6 260 base pairs (bp) (Table 1). These data were used as a reference, SD 2 4 1 2 shown in Supplementary Figure S2, for comparison with the size of CMA = Cuticle microhistological analysis; PCR-CE = Polymerase chain reaction-Capillary amplicons detected in feces. electrophoresis. 3 J. Pareja, C. Espunya, E. Baraza et al. Animal 15 (2021) 100145 Table 3 Percentage (%) of epidermal fragments in cow faeces determined by CMA and percentages of individuals in which a particular taxon appears by CMA and PCR-CE techniques. Taxon Faecal epidermal fragments (CMA) Individuals with the presence of the taxon (CMA) Individuals with the presence of the taxon (PCR-CE) Aphyllanthes monspeliensis1 5.5 100.0 0.0 Arbutus unedo 1.7 81.8 54.5 Brachypodium sp. 27.6 100.0 63.6 Bupleurum rigidum 0.6 9.1 36.4 Carex sp. 1 6.1 90.9 0.0 Cistus sp. 0.7 63.6 54.5 Dorycnium sp. 4.1 100.0 27.3 Ephedra fragilis2 0.0 0.0 18.2 Erica sp. 4.8 90.9 72.7 Sanguisorba minor 0.0 0.0 (81.8) Quercus sp. 3.5 72.7 Genista scorpius2 0.0 0.0 9.1 Globularia alypum 0.3 18.2 18.2 Hieracium pilosella2 0.0 0.0 9.1 Juniperus oxycedrus 1.3 81.8 72.7 Olea europaea 6.5 100.0 (90.9) Phillyrea angustifolia 0.0 0.0 Ononis sp. 0.05 9.1 27.3 Osyris alba2 0.0 0.0 9.1 Pinus halepensis 6.0 100.0 54.5 Pistacia lentiscus 4.1 90.9 100.0 Rhamnus alaternus 1.5 54.5 54.5 Rosmarinus officinalis1 9.0 100.0 0.0 Rubus ulmifolius 3.7 90.9 27.3 Smilax aspera 1.3 54.5 72.7 Other taxa 11.65 100.0 – CMA = Cuticle microhistological analysis; PCR-CE = Polymerase chain reaction-Capillary electrophoresis. 1 Taxa detected only by CMA. 2 Taxa detected only by PCR-CE. Considering the overall diet composition of each sample in terms of components in the faeces of cows and goats. Similar results have been the presence/absence of each taxon, the diets of both cows and goats are obtained recently in the study of the chamois diet (Rupicapra rupicapra) more similar to each other using the CMA method than using the PCR- (Espunyes et al., 2019), although in this case, with controlled diets and CE, as shown by the results of the non-metric multidimensional scaling using two amplicons in the molecular analysis. The use of a single analysis (Fig. 2). In both cases, the stress function was minimal, 0.05 in amplicon in the PCR-CE technique simplifies the method. Our results the case of cows, and 0.1 in the case of goats, ensuring that the two- validate the PCR-CE method previously developed showing that it can dimensional representation adequately collects the multidimensional also be applied to studies of diet composition of field grazing herbivores. information of all taxa. Thus, the PCR-CE analysis would be a good alternative to CMA, since the latter requires a long training process (Holechek and Gross, 1982), and Discussion can be affected by the subjectivity of the observer and the differential di- gestibility of each plant species (Leslie et al., 1983; Bartolome et al., The results obtained in this work show that the two analysis tech- 1995). For this reason, molecular analysis to determine the composition niques, CMA and PCR-CE, are able to detect a similar number of plant of the diet is becoming widespreadly used, although it is not exempt Table 4 Percentage (%) of epidermal fragments in goat faeces determined by CMA and percentages of individuals in which a particular taxon appears by CMA or by PCR-CE techniques. Taxon Faecal epidermal fragments (CMA) Individuals with the presence of the taxon (CMA) Individuals with the presence of the taxon (PCR-CE) Ampelodesmos mauritanica 12.2 100.0 25.0 Cneorum tricoccon 2.1 62.5 6.3 Arbutus unedo 0.5 25.0 68.8 Brachypodium sp. 9.3 100.0 31.3 Chamaerops humilis 6.1 87.5 31.3 Cistus sp. 19.2 93.8 93.8 Ephedra fragilis 1.0 50.0 6.3 Erica sp. 4.6 93.8 56.3 Genista lucida 0.0 0.0 (43.8) Quercus sp. 2.1 62.5 Sanguisorba minor 0.0 0.0 Globularia alypum 1.2 62.5 12.5 Juniperus oxycedrus 1.0 56.3 6.3 Olea europaea 8.7 100.0 (68.8) Phillyrea angustifolia 0.9 43.8 Pinus halepensis 3.7 93.8 6.3 Pistacia lentiscus 5.7 93.8 93.8 Rosmarinus officinalis 5.1 100.0 18.8 Smilax aspera1 1.8 56.3 0.0 Others 14.8 100.0 – CMA = Cuticle microhistological analysis; PCR-CE = Polymerase chain reaction-Capillary electrophoresis. 1 Taxa detected only by CMA. 4 J. Pareja, C. Espunya, E. Baraza et al. Animal 15 (2021) 100145 Fig. 1. Box diagram showing the median (thick line) and distribution of the percentage of fragments found in faecal samples of cows and goats by Cuticle microhistological analysis (CMA) for taxa not detected (0) and detected (1) by PCR-Capillary electrophoresis (PCR-CE). from limitations, such as their extreme dependence on the degradation difficult to explain and may be based on the particularities of each diges- of the gDNA caused by the animal's own digestion or by environmental tive tract (Garnick et al., 2018; Sugimoto et al., 2018), or on the amplifi- factors (Deagle et al., 2006; Espunyes et al., 2019). However, the results cation bias due to DNA degradation (Valentini et al., 2009). It is also obtained here suggest that both techniques can be complemented worth noting the fact that, as regards the presence or absence of a cer- rather than substituted. This is based on the fact that some species are tain taxon, the samples of both animals are more similar to each other only detected by one of the two methods. when the CMA technique is applied. In this sense, CMA would provide In this work, differences between individuals have also been de- more information than the PCR-CE in the analysis of the components tected when both techniques are used. Thus, in the case of cows, no cor- of the diet. These results agree with those obtained by Murphree relation was observed between techniques, indicating that if a taxon (2012), who, when working with controlled diets, obtained a mean of appears in many individuals applying one technique, the same does 89% of correct identifications using CMA, and only 50% using molecular not have to happen when applying the other. In contrast, a positive cor- analysis. Also, Valentini et al. (2009) indicated that by using the molec- relation was observed in goats, which suggested that both techniques ular method, 50% of the ingested taxa could be identified. will detect a greater or lesser percentage of individuals with a given Finally, it should be noted that the main difference between the two taxon. In addition, the fact that taxa with higher percentages in the av- methods is the possibility of quantifying the components of the diet. erage diet of goats appear in more individuals, and that this does not Thus, proportions of the different fragments detected in the faeces can occur in the case of cows, reinforces the idea that the results of both be obtained by CMA and, in general, it is assumed that these proportions techniques vary depending on the animal species. This difference is correspond to the animal's intake. On the other hand, PCR-CE shows Fig. 2. Non-metric multidimensional scaling (NMDS) representing the distance between faecal samples of cows and goats as a function of the species composition of each sample analyzed by Cuticle microhistological analysis (CMA) in triangle and PCR-Capillary electrophoresis (PCR-CE) in circle, using the Jaccard distance matrix. 5 J. Pareja, C. Espunya, E. Baraza et al. Animal 15 (2021) 100145 severe inaccuracies since the amount of extracted DNA from faeces can- Islands that, through the staff of the Consortium for the Recovery of not be related to the amount of food ingested, probably due to the var- the Fauna of the Balearic Islands (COFIB), provided us with the excre- iability in the degradation of the DNA during digestion (Deagle et al., ments of the feral goats. 2009 and 2010; Valentini et al., 2009). Another limitation is the similar- ity of the amplicon size that can occur between some species. Valentini Financial support statement et al. (2009) recommend, for these “problematic” taxa, to complement with another additional method, such as the CMA, or simply work This work was supported by Ministry of Economy and Competitive- with larger taxonomical groups, such as genders or families. ness (MINECO) of Spain (CGL2015-70449-R) and by the project Life It is concluded that the molecular analysis by PCR-CE, based on the Montserrat (LIFE13 BIO/ES/000094). Javier Pareja was supported by length of a single amplicon, is a fast and reliable method to detect the the National Fund for Scientific, Technological Development and Tech- different plant components in the faeces of herbivores. The type of her- nological Innovation (FONDECYT), the funding branch of the National bivore seems to influence the results obtained with this technique, so Council for Science, Technology and Technological Innovation more studies are required to address this aspect. There is no doubt (CONCYTEC) Peru (Grant contract number N° 236-2015-FONDECYT). that the PCR-CE can complement the CMA by being able to detect some species that leave few epidermal fragments and consequently References are not detected by CMA. However, it cannot be considered as an alter- Bartolome, J., Franch, J., Gutman, M., Seligman, N.G., 1995. Physical factors that influence native, since CMA detects the presence of the different taxa in a greater fecal analysis estimates of herbivore diets. Journal of Range Management Archives 48, number of samples and at the same time is able to obtain percentage 267–270. data of the diet content. Bates, D., Mächler, M., Bolker, B.M., Walker, S.C., 2015. Fitting linear mixed-effects models using lme4. Journal of Statistical Software 67, 1–48. Bautista, J., Medina, J., Mamani, G., 1996. Selectividad y degradabilidad in situ de ingesta Supplementary materials de pastizales nativos en alpacas y llamas en puna húmeda. Revista de Investigación de Camélidos Sudamericanos ALLPAK’A 6, 27. Supplementary data to this article can be found online at https://doi. Bertolino, S., Di Montezemolo, N.C., Bassano, B., 2009. Food-niche relationships within a guild of alpine ungulates including an introduced species. Journal of Zoology 277, org/10.1016/j.animal.2020.100145. 63–69. Chetri, M., 2006. Diet analysis of gaur (Bos gaurus gaurus; Smith, 1827) by micro- Ethics approval histological analysis of fecal samples in parsa wildlife reserve, Nepal. Our Nature 4, 20–28. Coates, D.B., Dixon, R.M., 2008. Development of near infrared analysis of faeces to esti- Not applicable. mate non-grass proportions in diets selected by cattle grazing tropical pastures. Jour- nal of Near Infrared Spectroscopy 16, 471–480. Data and model availability statement Codron, D., Brink, J.S., 2007. Trophic ecology of two savanna grazers, blue wildebeest Connochaetes taurinus and black wildebeest Connochaetes gnou. European Journal of Wildlife Research 53, 90–99. None of the data were deposited in an official repository, but are Deagle, B.E., Eveson, J.P., Jarman, S.N., 2006. Quantification of damage in DNA recovered available upon request only for research use. The model was not depos- from highly degraded samples - a case study on DNA in faeces. Frontiers in Zoology ited in an official repository, but scripts of model are available at Supple- 3, 11. Deagle, B.E., Kirkwood, R., Jarman, S.N., 2009. Analysis of Australian fur seal diet by pyro- mentary Material S1. sequencing prey DNA in faeces. Molecular Ecology 18, 2022–2038. Deagle, B.E., Chiaradia, A., McInnes, J., Jarman, S.N., 2010. Pyrosequencing faecal DNA to Author ORCIDs determine diet of little penguins: is what goes in what comes out? Conservation Ge- netics 11, 2039–2048. Espunyes, J., Espunya, C., Chaves, S., Calleja, J.A., Bartolomé, J., Serrano, E., 2019. Comparing Javier Pareja: 0000-0001-5759-4215. the accuracy of PCR-capillary electrophoresis and cuticle microhistological analysis Carme Espunya: 0000-0002-4581-5006. for assessing diet composition in ungulates: a case study with Pyrenean chamois. Elena Baraza: 0000-0002-2928-3978. PLoS One 14, e0216345. Jordi Bartolomé: 0000-0002-3784-5248. Ferreira, L.M.M., Garcia, U., Rodrigues, M.A.M., Celaya, R., Dias-da-Silva, A., Osoro, K., 2007. Estimation of feed intake and apparent digestibility of equines and cattle grazing on heathland vegetation communities using the n-alkane markers. Livestock Science Author contributions 110, 46–56. Garnick, S., Barboza, P.S., Walker, J.W., 2018. Assessment of animal-based methods used for estimating and monitoring rangeland herbivore diet composition. Rangeland Javier Pareja: Validation, Formal analysis, Investigation, Writing – Ecology & Management 71, 449–457. original draft, Visualization and Writing - review & editing. Carme Henley, S.R., Smith, D.G., Raats, J.G., 2001. Evaluation of 3 techniques for determining diet Espunya: Term, Conceptualization, Methodology, Software, Validation, composition. Rangeland Ecology & Management/Journal of Range Management Ar- Formal analysis, Investigation, Resources, Data curation and Writing - chives 54, 582–588. Holechek, J.L., Gross, B., 1982. Training needed for quantifying simulated diets from review & editing. Elena Baraza: Methodology, Software, Validation, For- fragmented range plants. Rangeland Ecology & Management/Journal of Range Man- mal analysis, Investigation, Data curation, Writing - original draft and agement Archives 35, 644–647. Writing - review & editing. Jordi Bartolomé: Term, Conceptualization, Leslie, D.M., Vavra, M., Starkey, E.E., Slater, R.C., 1983. Correcting for differential digestibil- Methodology, Investigation, Visualization, Supervision, Project adminis- ity in microhistological analyses involving common coastal forages of the pacific northwest. Journal of Range Management Archives 36, 730–732. tration and Funding acquisition. Murphree, J.J., 2012. Evaluation of the efficacy of DNA sequencing and microhistological analysis for determining diet composition in ungulates. MSc thesis. Arizona State Declaration of interest University, Tempe, AZ, USA. Norbury, G.L., Sanson, G.D., 1992. Problems with measuring diet selection of terrestrial, mammalian herbivores. Australian Journal of Ecology 17, 1–7. None. Oksanen, J., Blanchet, F.G., Friendly, M., Kindt, R., Legendre, P., McGlinn, D., Minchin, P.R., O’Hara, R.B., Simpson, G.L., Solymos, P., Stevens, M.H.H., Szoecs, E., Wagner, H., 2018. Vegan: community ecology package, R package version 2.5-2. Retrieved on 2 Febru- ary 2019 from https://CRAN.R-project.org/package=vegan. Acknowledgements Pegard, A., Miquel, C., Valentini, A., Coissac, E., Bouvier, F., François, D., Taberlet, P., Engel, E., Pompanon, F., 2009. Universal DNA-based methods for assessing the diet of graz- Thanks to Miguel Ibáñez, Gabriella Rizzardini, to the owner of the ing livestock and wildlife from feces. Journal of Agricultural and Food Chemistry 57, 5700–5706. farm “El Brunet”, to the hunters of the Big Game Association of Balearic Pinto, R.R., Gómez, C.H., Guevara, H.F., Hernández, S.D., Ruiz, S.B., 2014. Preferencia y Islands and to the Genomics and Bioinformatics Service of the UAB. The conducta ingestiva de ovinos alimentados con frutos arboreos tropicales. Revista Department of Agriculture, Environment and Territory of the Balearic Científica 24, 158–163. 6 J. Pareja, C. Espunya, E. Baraza et al. Animal 15 (2021) 100145 Shrestha, R., Wegge, P., 2006. Determining the composition of herbivore diets in the Sugimoto, T., Ito, T.Y., Taniguchi, T., Lkhagvasuren, B., Oyunsuren, T., Sakamoto, Y., trans-Himalayan rangelands: a comparison of field methods. Rangeland Ecology & Yamanaka, N., 2018. Diet of sympatric wild and domestic ungulates in southern Management 59, 512–518. Mongolia by DNA barcoding analysis. Journal of Mammalogy 99, 450–458. Soininen, E.M., Valentini, A., Coissac, E., Miquel, C., Gielly, L., Brochmann, C., Brysting, A.K., Taberlet, P., Coissac, E., Pompanon, F., Gielly, L., Miquel, C., Valentini, A., Vermat, T., Sønstebø, J.H., Ims, R.A., Yoccoz, N.G., Taberlet, P., 2009. Analysing diet of small herbi- Corthier, G., Brochmann, C., Willerslev, E., 2007. Power and limitations of the chloro- vores: the efficiency of DNA barcoding coupled with high-throughput pyrosequenc- plast trnL (UAA) intron for plant DNA barcoding. Nucleic Acids Research 35, e14. ing for deciphering the composition of complex plant mixtures. Frontiers in Valentini, A., Miquel, C., Nawaz, M.A., Bellemain, E., Coissac, E., Pompanon, F., Gielly, L., Zoology 6, 16. Cruaud, C., Nascetti, G., Wincker, P., Swenson, J.E., Taberlet, P., 2009. New perspectives Sponheimer, M., Lee-Thorp, J.A., DeRuiter, D.J., Smith, J.M., Van Der Merwe, N.J., Reed, K., in diet analysis based on DNA barcoding and parallel pyrosequencing: the trnL ap- Grant, C.C., Ayliffe, L.K., Robinson, T.F., Heidelberger, C., Marcus, W., 2003. Diets of proach. Molecular Ecology Resources 9, 51–60. southern African Bovidae. Journal of Mammalogy 84, 471–479. Vives, J.A., Baraza, E., 2010. La cabra doméstica asilvestrada (Capra hircus) en Mallorca Stewart, D.R.M., 1967. Analysis of plant epidermis in faeces: a technique for studying the ¿Una especie a erradicar? Galemys 22, 193–205. food preferences of grazing herbivores. Journal of Applied Ecology 4, 83–111. 7