Differentiation of foraging areas of Lama guanicoe (Camelidae) in Patagonia using multiproxy analysis of feces: Modern reference dataset for paleoecological reconstruction

Abstract

Paleoecological studies based on camelid coprolites analysis provide valuable information about diet, vegetation, and habitats. In Patagonia, Argentina, relationships between humans and camelids date back to the early Holocene. The archeological site Cueva Milodón Norte 1 (CMN1), Pueyrredón lake, contains camelid coprolites and evidence of human occupation since 8600 years cal. BP. Environmental changes and lake-level fluctuations have influenced the landscape and resource availability. Actualistic studies are essential for building reference models that improve paleoecological interpretations of Holocene coprolites in Patagonia. The aim of this study was to identify and differentiate the foraging areas of guanaco (Lama guanicoe) through the analysis of phytoliths, pollen, and plant fragments from modern feces collected in two distinct environments: the shrub-grass steppe surrounding CMN1, and the sub-shrub steppe near Salitroso lake (SAL), within the lake Posadas-Pueyrredón basin. Multiproxy protocol was applied to identified plant taxa consumed by L. guanicoe. Also, a plant phytolith reference collection was built for this study. The results of the combined analyses of the three proxies in modern feces identified representative taxa for each environment: Poeae, Cerastium arvense, Colobanthus lycopodioides, Colliguaja, Azorella, Fabaceae, and Asteraceae at CMN1 site, and Stipeae, Verbenaceae, Nassauvia, Atriplex sagittifolia, Senecio and Nardophyllum at SAL site, among others. Principal component analyses based on each proxy showed a differentiation between feces from CMN1 and SAL, supporting the identification of discrete foraging areas. These results demonstrate the value of multiproxy approaches in building reference models for interpreting past camelid diets and habitat use in southern Patagonia.

Keywords

phytolith pollen plant fragments actualistic reference model habitat use paleodiet reconstruction

Introduction

Paleodietary studies through coprolite analysis have provided detailed information about the interaction between herbivores and their environment during the past (Horrocks et al., 2002). In South America, a number of studies have contributed to the advancement of knowledge about diet of herbivores, vegetation, and ancient habitats (Mosca Torres et al., 2018; Velázquez et al., 2014; Velázquez and Burry, 2022). These include the construction of plant reference collections (from different vegetation proxies) and the analysis of modern camelid feces, which serve as comparative frameworks (Agliano et al., 2024; D’Antoni, 2008; Martínez Tosto et al., 2016). Several authors have contributed to the development of these frameworks through palynological and micro histological studies of modern camelid feces (Agliano, 2022; Mosca Torres et al., 2022; Puig et al., 1997; Romano, 2012; Velázquez and Burry, 2012), which have shown seasonal differences in foraging and the presence of extra-local pollen, indicative both direct and indirect plant consumption (Velázquez et al., 2014; Velázquez and Burry, 2012, 2019). However, differences in plant digestibility and poor preservation of organic remains can affect their representativeness in the feces record (Benvenuto et al., 2022; Shillito et al., 2013). Phytoliths are an underexplored proxy in modern feces and coprolites from Patagonia; their high stability in sediments and feces makes them a valuable paleoecological indicator (Piperno, 2006; Velázquez et al., 2021). However, in the context of dietary studies their use depends on the consumption of phytolith producing plants and the proper preservation of their remains (Shillito et al., 2013). Therefore, it is important to employ multiple proxies to strengthen interpretations of diet, foraging areas and paleoenvironmental conditions (Wood et al., 2021).

In Patagonia, Argentina, the relationship between humans and camelids dates back to the end of the Pleistocene, as evidenced by camelid bone fragments found at archeological sites, which indicate that these animals were used as a source of food and raw materials (Casamiquela, 1983; Miotti and Salemme, 1999). The archeological site Cueva Milodón Norte 1 (CMN1), near Pueyrredón lake, has showed evidence of human occupation dating to nearly 8000 cal yr BP (Aschero et al., 2019; Sacchi et al., 2016), and camelid coprolites have also been found. Since ~8000 cal yr BP, environmental changes and lake-level fluctuations have shaped the landscape and influenced resource availability (Caruso Fermé and Civalero, 2014; Horta et al., 2011, 2016; Velázquez and Burry, 2022). At present, the guanaco (Lama guanicoe Müller (1776) is distributed across a wide range of environments in Patagonia. This camelid is a generalist, pseudo-ruminant herbivore of great ecological importance in the region, with a diet based on herbs and shrubs (Burgi, 2007; Raedeke, 1980). Historically, the species ranged from Peru to Tierra del Fuego, but its distribution has declined since the 19th century due to competition with domestic livestock, hunting, and adverse climatic factors (Franklin, 1982; Montes et al., 2000). In this context, this study explores the potential of a multiproxy approach based on phytoliths, pollen, and plant fragments identified in modern feces to provide insights into the differentiation of guanaco foraging areas in Patagonia. Vegetation in two distinct environments with different ecological characteristics was characterized: the shrub–grass steppe surrounding the archeological site Cueva Milodón Norte 1 (CMN1) and the sub-shrub steppe of Salitroso lake (SAL). This study also presents a small reference set of plant phytoliths from the most common taxa consumed by guanaco in both environments. The aim of this reference set was to improve the identification of plant taxa in feces samples. The information obtained in this study contributes to a better understanding of guanaco feeding behavior in different ecosystems and establishes the basis for building reference models that could improve paleoecological interpretations of Holocene camelid coprolites in Patagonia.

Study Area

The Posadas-Pueyrredón-Salitroso Lacustrine System (PPSLS; 47° 30’ 24” S, 71° 49’ 06” W) is located in the northwestern region of Santa Cruz Province, Argentina. The climate in the area is cold-temperate and semi-arid, with temperatures ranging between 5°C and 8°C and annual precipitation varying from 200 mm in the west to 150 mm in the east (Oliva et al., 2001). This area is influenced by cold fronts associated with the Southwestern Wind System, which bring precipitation to the southernmost part of South America. As these fronts cross the Andes, they generate a decrease in west-to-east precipitation gradient (Garreaud et al., 2013; Viale et al., 2019). From a phytogeographical perspective, according to Cabrera (1976), the basin is located between the Deciduous Forest District of the Subantarctic province and the Central District of the Patagonian province. The region encompasses the ecological areas of the Andean Complex, Sub-Andean Grassland, and Central Plateau, as described by Oliva et al. (2001). The vegetation reflects the precipitation gradient, ranging from Nothofagus forests in the wetter western areas to shrub-grass steppes in the central region, dominated by species such as Festuca pallescens and Azorella prolifera (ex Mulinum spinosum), among others. Further east, the vegetation transitions to sub-shrub steppes characterized by species like Nassauvia glomerulosa and various Stipa species, including S. speciosa and S. chrysophylla (Oliva et al., 2001). This study focused on two different vegetation units within the Central Plateau region: the shrub-grass steppe surrounding the archeological site Cueva Milodón Norte (CMN1 site; 47° 18’ 22.54” S, 71° 53’ 47.80” W) and the sub-shrub steppe in the Salitroso lake area (SAL site; 47° 34’ 15.56” S, 71° 35’ 38.76” W; Figure 1).

Figure 1.

(a) Location of the study area. Base map of Argentina from the Instituto Geográfico Nacional (IGN), modified by the authors, (b) satellite image (Landsat Copernicus) acquired on 26 November 2025, showing the two sampling sites (CMN1 and SAL), (c) the shrub–grass steppe surrounding the archeological site Cueva Milodón Norte (CMN1), (d) the sub-shrub steppe in the Salitroso lake area (SAL), and (e) on the right, footprints of a guanaco; on the left, guanaco feces.

Materials and methods

Fieldwork

During the spring field campaign, three guanaco (Lama guanicoe) dung piles (i.e. accumulations of feces deposited by different individuals) with clearly defined boundaries were selected within each vegetation unit (CMN1 and SAL sampling sites). These sampling sites were located approximately 30 km apart to increase the likelihood that the contents of the feces from each dung pile reflect the vegetation of the vegetation unit in which they were found (i.e. reducing the likelihood that animals fed at the CMN1 site and defecated at the SAL site). Dung piles were selected within a previously prospected area of 10 ha, entirely contained within a single vegetation unit and avoiding ecotonal zones. A minimum distance of more than 30 m was maintained between selected dung piles to reduce the resampling of the same defecation events. From each dung pile, one composite sample was generated by pooling four fresh feces (CMN1:81, 82, 83; SAL: 88, 92, 97). The feces were assigned to guanaco based on their size, shape, and color (Figure 1e). Despite efforts to ensure sample independence, some feces samples could have been derived from the same individual or from different individuals with similar diets, as they may belong to the same social group, consistent with the species’ territorial behavior (Franklin, 1982). The collected samples were individually packed in plain kraft paper bags to avoid contamination, promote drying, and prevent fungal growth prior to laboratory processing. To characterize the vegetation unit at each site, vegetation surveys were conducted using the line-intercept method. Three transects (10 m in length) were established near the sampled dung piles in patches representative of the dominant vegetation (Mueller-Dombois and Ellenberg, 1974). Representative plant species recorded along the transects were collected to build botanical reference collections and were identified following Flora Argentina (Zuloaga et al., 2023).

Laboratory work

Phytolith reference collection

In order to identify plant taxa based on phytolith production patterns, seven species were selected for phytolith assemblage analysis (Table 1). This selection was based on plant species that were dominant in the vegetation surveys at each site and are known, from previous studies, to be consumed by guanaco (Puig et al., 1997). Leaves from the upper and lower parts of three specimens of each species were collected. Phytolith extraction followed the calcination technique of Campos and Labouriau (1969). The resulting ash was weighed to calculate the phytolith content per gram of leaf tissue (% dry weight = (g phytoliths / g tissue) × 100). The ashes were mounted with immersion oil, and 200–350 phytolith morphotypes were observed and classified using an optical microscope at 400× and 1000× magnification. Phytolith naming followed the International Code for Phytolith Nomenclature 2.0 (Neumann et al., 2019). The phytolith assemblage of each species was determined based on the relative frequency of each morphotype. In addition, to confirm the identification and description of phytoliths in plant tissues, leaves were analyzed according to the methods proposed by Fernández Honaine et al. (2019).

Table 1.

List of plant species to the phytolith reference collection and the percentage of silica content.

Family		Plant species	Growth form	Life-cycle	Sample location	Phytolith content (g/100 g)
Monocotyledons	Poaceae	Festuca pallescens (St.-Yves) Parodi	Herb	Perennial	CMN1	0.90	±	0.116
Monocotyledons	Poaceae	Pappostipa humilis (Cav.) Romasch	Herb	Perennial	SAL	6.03	±	0.803
Dicotyledons	Apiaceae	Azorella prolifera (Cav.) G.M. Plunkett & A.N. Nicolas^a	Shrub	Perennial	CMN1	0.19	±	0.077
	Asteraceae	Nardophyllum bryoides (Lam.) Cabrera	Shrub	Perennial	CMN1	0.52	±	0.126
	Asteraceae	Nassauvia glomerulosa (Lag. ex Lindl.) D. Don	Shrub	Perennial	SAL	2.32	±	0.690
	Euphorbiaceae	Colliguaja integerrima Gillies & Hook.	Shrub	Perennial	CMN1	0.26	±	0.169
	Solanaceae	Lycium chilense Miers ex Bertero var. chilense	Shrub	Perennial	SAL	0.66	±	0.146

Mean value and standard deviation.

Mulinum spinosum (Cav.) Pers.

Feces processing

The feces that constitute each sample were dried at 35°C in a stove for 5 days, weighed, and subjected to morphological description (diameter, width, shape, texture, and color). Surface inclusions were also recorded under a stereoscopic microscope at 4× magnification (Jouy-Avantin et al., 2003). To avoid including non-dietary particles resulting from post-depositional contamination, only the inner portion of the feces was analyzed (hereinafter feces sample).

Phytoliths, pollen, and plant fragments analysis

In order to extract phytoliths, pollen, and plant fragments from each feces sample, the extraction technique of Velázquez et al. (2019) was followed. Feces samples were placed in 15 ml Falcon tubes, and a tablet of Lycopodium clavatum spores (Batch No. 124,961, mean = 12,542 spores/tablet) was added. The feces samples were hydrated with 0.5% trisodium phosphate dodecahydrate solution and stored at 4°C until the material disaggregated. The disaggregated material was then filtered through a 260 μm mesh using trisodium phosphate solution. Plant fragments retained on the mesh were dried at 35°C for 24 h and reserved for macrofossil analysis. The filtrate was concentrated by centrifugation at 2500 rpm for 5 min for phytolith and pollen analysis. Three semi-permanent preparations were made using immersion oil for the microscopic observation of phytoliths at 100× magnification from a 1 ml aliquot of the filtrate. At least 300 diagnostic phytoliths were counted and identified. When articulated phytoliths were observed, the individual silicified cells were counted. Pollen extraction was performed through acetolysis (nine parts acetic anhydride and one part sulfuric acid) following Faegri and Iversen (1989). Semi-permanent slides were prepared with glycerin for the microscopic identification. Identification and counting of pollen grains and spores were conducted at 400 and 1000× magnification. The counting criterion was to achieve a pollen sum of at least 200 grains per sample (Sobolik, 1988). Pollen percentage and pollen concentration (grains/gr) were calculated. According to Faegri and Iversen (1989), anemophilous taxa produce between 10,000 and 70,000 pollen grains per anther, whereas zoophilous taxa produce 1000 grains or fewer per anther. Based on these production differences, pollen concentrations greater than 10,000 grains/g for anemophilous taxa and greater than 1,000 grains/g for zoophilous taxa were considered high and potentially indicative of intentional consumption. Part of the plant fragments recovered from the mesh were rehydrated in water and bleached with 50% sodium hypochlorite for 2 min. They were then washed using distilled water on a 260 μm mesh. The material from each sample was mounted on slides with glycerin jelly for microscopic observation. Observations were conducted at 100× magnification along five random lines per slide and plant fragments from 100 microscopic fields were identified. All proxies were observed with a Zeiss Primo Star microscope. Photographs were taken with a Nikon Coolpix S2900 digital camera. Phytoliths were classified according to the ICPT by Neumann et al. (2019), with specific categorizations for Bilobate morphotypes following Fredlund and Tieszen (1994). The taxonomic identification of pollen types and phytoliths was based on specialized literature (Heusser, 1971; Markgraf and D’Antoni, 1978; Piperno, 2006; Twiss et al., 1969) and reference collections from the Laboratorio de Palinología y Bioantropología (IIPROSAM-CONICET). Plant fragments were identified by comparing them with histological preparations of species collected at the study area, housed in the Laboratorio de Botánica at Universidad Nacional de Mar del Plata. Results were plotted on a percentage diagram using the Tilia v. 1.7.14 software (Grimm, 2011).

SEM and EDS analysis

Samples were analyzed using a Scanning Electron Microscope (SEM) and Energy Dispersive Spectroscopy (EDS) at the Scanning Electron Microscopy and Microanalysis Service (SeMFI – LIMF), Faculty of Engineering, Universidad Nacional de la Plata, Argentina, to determine the elemental composition of certain particles present in the ash of plant material and feces.

Statistical analysis

Principal component analysis (PCA) was conducted: a) to evaluate the relevance of phytolith assemblages in differentiating the plant species analyzed, b) to characterize and differentiate the foraging areas through the analysis of each proxy (phytoliths, pollen, and plant fragments) from the feces. PCA was performed through a correlation matrix. Only isolated morphologies were included for the phytoliths analysis in plants. For the phytolith analysis in feces, categories with less than 3% relative frequency, as well as those classified as low diagnostic value (Elongate columnar/clavate, Elongate dentate, Elongate entire to Oblong entire, Elongate entire, Elongate sinuate, Stomatal complexes, Bloky, Acute bulbosus, Circular, Elongate arcuate, Tracheary annulate/helical), Others (Saddle, Papillate, Cone, Stomatal complexes, and Polylobate), and unidentified categories, were excluded. For the PCA analysis of pollen in feces, pollen types were considered if they exceeded 2% in at least one sample. Pollen types from taxa indicative of anthropogenic impact (Rumex, Brassicaceae, Asteraceae subf. Cichorioideae, Plantago, Gomphrena) and those that were unidentified were excluded. In the PCA analysis of plant fragments in feces, categories ‘Poaceae’ and ’Dicots’ (correspond to other Poaceae and other dicots that could not be identified to a more taxonomic level), and unidentified categories were excluded. To determine whether there were significant differences between the feces from CMN1 and SAL, a Fuzzy C-Means Clustering (FCMC) analysis was conducted using the principal components obtained for each proxy. Unlike traditional clustering methods, this fuzzy approach assigns each sample a degree of membership to different clusters, allowing for a more detailed and accurate interpretation of the cluster structure. These analyses were carried out using the R program (R Core Team, 2024).

Results

Vegetation characterization at CMN1 and SAL

The vegetation cover was 75.1% and 65.5% in CMN1 and SAL, respectively. A total of 11 native vascular plants were identified at the species level. Species richness was higher in CMN1 compared to SAL (Figure 2). The dominant taxa in CMN1 were Azorella prolifera (=Mulinum spinosum; 33.8%) and Festuca pallescens (14.0%), followed by Colliguaja integerrima (8.3%). In SAL, the dominant species were Nassauvia glomerulosa (28.8%), Pappostipa humilis (14.3%), and Lycium chilense (11.4%), followed by Atriplex sagittifolia (7.5%). The other taxa represented less than 5%.

Figure 2.

Relative percentage of taxa and bare ground in the CMN1 and SAL sites.

Reference collection of phytolith in plants

The mean phytolith content in monocotyledons, exclusively represented by the Poaceae (grasses), ranged from 0.9% (Festuca pallescens) to 6.03% (Pappostipa humilis), whereas in dicotyledons, it ranged from 0.19% (Azorella prolifera = Mulinum spinosum) to 2.32% (Nassauvia glomerulosa; Table 1).

Phytolith assemblages

Phytoliths were recovered in the leaves of the seven species studied. The morphotypes appeared in articulated and isolated forms. The relative percentages of morphotypes are available in Supplemental material S1.

Poaceae

Festuca pallescens

Phytolith assemblage was mainly characterized by Rondel (RON), Trapezoid (TRZ), Elongate columnar/clavate (ELO_COL/CLA), Elongate dentate (ELO_DEN) and Acute bulbosus (ACU_BUL). Other morphotypes described were Crenate (CRE) and Elongate arcuate (ELO_ARC; less than 10%; Figure 3a–f).

Figure 3.

Phytoliths of leaves observed under an optical and scanning electron microscope: (a) epidermal tissue with articulated Rondel, (b) Rondel, (c) Acute Bulbosus (short), (d) epidermal tissue with Elongate Columnar/Clavate, (e) Acute Bulbosus (long), (f) Elongate Dentate articulated, (g) Crenate articulated, (h) epidermal tissue with Bilobate_stypa-Type articulated, (i) Tracheary Annulate/Helical, (j) Tracheary Annulate/Helical articulated with Elongate Arcuate, (k) Elongate Entire to Oblong Entire, (l) Tabular Entirearticulated, (m–p) Oblongo to Ovate Granulate articulated (OBL/OVA_GRA), (q) Tabular Entire to Sinuate, (r) Elongate Entire to Oblong Entire, (s) stomatal complex, (t) Tabular Entire Granulate articulated, (u) Tabular Entire Psilate articulated, (v) Spheroid Psilate, (w and x) Spheroid Ornate , (o, p and x) photo from scanning electron microscope and (y) EDAX analysis.

Pappostipa humilis

Phytolith assemblage was mainly characterized by Acute bulbosus, Elongate columnar/clavate, Crenate, Bilobate_Stypa-type (BIL_Stipa type), and Elongate arcuate. Other morphotypes also described were, Elongate entire (ELO_ENT), Elongate dentate (ELO_DEN) and Stomatal complexes (STO_COM_MON; less than 10%; Figure 3g and h).

Apiaceae

Azorella prolifera (=Mulinum spinosum)

Phytolith assemblages were mainly represented by Elongate arcuate (ELO_ARC) and Tracheary annulate/helical (TRA_ANN/HEL) and Elongate entire to Oblong entire (ELO_ENT to OBL_ENT). Acute bulbosus, Tabular entire psilate (TAB_ENT_PSI) and Tabular entire to sinuate (TAB_ENT/SIN) were scarce (in less than 5%; Figure 3i–l).

Asteraceae

Nardophyllum bryoides

Phytolith assemblage was mainly characterized by Tabular entire granulate (TAB_ENT_GRA), Tabular entire psilate (TAB_ENT_PSI) and Oblongo to Ovate granulate (OBL/OVA_GRA) phytoliths. Other phytolith morphotypes observed in less than 10% were Elongate entire, Tracheary annulate/helical and Stomatal complexes (STO_COM_DIC; Figure 3m–p).

Nassauvia glomerulosa

Phytolith assemblage was mainly characterized by Tabular entire to sinuate and Elongate entire to Oblong entire. Other morphotypes observed in less than 10% were Acute bulbosus, Circular (base of hairs; CIR), Elongate crenate (ELO_CRE), Elongate arcuate, Tracheary annulate/helical and Stomatal complexes (STO_COM_DIC; Figure 3q–s).

Euphorbiaceae

Colliguaja integerrima

Phytolith assemblages were mainly represented by Tabular entire psilate and Tabular entire granulate. Other morphotypes present in less than 5% were Elongate entire, Tracheary annulate/helical, Blocky (BLO), Acute bulbosus and Stomatal complexes (STO_COM_DIC; Figure 3t and u).

Solanaceae

Lycium chilense

Phytolith assemblages were mainly represented by Spheroid psilate (SPH_PSI) and Spheroid ornate (SPH_ORN; with diameter of 10–15 µm) and Tabular entire psilate and Tabular entire granulate. Other morphotypes observed in less than 10% were Elongate entire, Tracheary annulate/helical, Acute bulbosus and Stomatal complexes (STO_COM_DIC; Figure 3v–y).

Multivariate analysis (PCA) of plant phytolith assemblages

The principal component analysis (PCA) based on the phytolith assemblage in the leaves of the species explained 87.43% of the total variance, with the first three dimensions accounting for the majority of the variance (43% for Dim1, 26.9% for Dim2, and 17.4% for Dim3). The analysis of variable contributions and their graphical representation of Dim1 versus 2 showed that, the most influential variables were Rondel (RON), Trapezoid (TRZ), Crenate (CRE), Bilobate Stipa type (BIL_Stipa type), Elongate columnar/clavate (ELO_COL/CLA), Elongate dentate (ELO_DEN) and Stomatal complexes of Monocotyledons (STO_COM_MON). To a lesser extent, Tabular (TAB) was also relevant. The variables with the highest contribution to Dim1 versus 3 were Elongate columnar/clavate (ELO_COL/CLA), Elongate entire to Oblong entire (ELO_ENT to OBL_ENT), Elongate arcuate and Tracheary annulate/helical (ELO_ARC + TRA_ANN/HEL) and Tabular (TAB). It is worth noting that the Spheroid psilate/ornate (SPH) variable did not have sufficient representation, as in none of the possible combinations of the three dimensions did it account for over 50% of the variability. The biplots revealed that grasses formed a separate group from the rest of the species and can also be distinguished from each other. Festuca pallescens was primarily associated with Trapezoid (TRZ), Elongate dentate (ELO_DEN), and Rondel (RON), while Pappostipa humilis was associated with Crenate (CRE), Bilobate Stipa type (BIL_Stipa type), and Stomatal complexes of Monocotyledons (STO_COM_MON). In the Dim1 versus Dim3 biplot, Azorella prolifera (=Mulinum spinosum) was mainly associated with Elongate arcuate (ELO_ARC) and Tracheary annulate/helical (TRA_ANN/HEL), and Elongate entire to Oblong entire (ELO_ENT to OBL_ENT), while Nassauvia glomerulosa was mainly associated with the latter. Nardophylum bryoides, Lycium chilense, and Colliguaja integerrima were exclusively associated with the Tabular (TAB) variable (Figure 4).

Figure 4.

Biplot of the principal components analysis (PCA) of plants based on the relative frequencies of phytoliths produced by leaves: (a) Dim1 versus 2 and (b) Dim1 versus 3.

Morphological description of feces

The morphology of the feces examined was predominantly irregular in shape, with some exhibiting a tapered extremity, while others demonstrated a rounded extremity and a striated surface. The majority of the feces exhibited a coloration of a greenish-brown, accompanied by the presence of plant fragments on their surface (Camiolo, 2023). The measurements obtained to CMN1 feces ranged from 1.11 to 1.72 cm in length, from 0.97 to 1.21 cm in diameter, and their weight ranged from 0.25 to 0.39 g. In the case of SAL feces, the measurements ranged from 1.25 to 1.78 cm in length, from 0.96 to 1.58 cm in diameter, and their weight ranged from 0.42 to 0.61 g.

Phytoliths, pollen and plant fragments in feces

Phytoliths in feces

The analysis of phytoliths from feces samples showed a percentage of articulated phytoliths ranging from 48% to 71% in CMN1 and 27% to 58% in SAL, with many of them still associated with plant tissue fragments. The total number of phytolith morphotypes identified in the feces, used to determine the phytolith assemblage, ranged from 409 to 671. Among the morphotypes observed in CMN1, Rondels (RON) were dominant, ranging from 24% to 49%. The morphotypes Crenate (CRE), Elongate entire (ELO_ENT), Elongate arcuate (ELO_ARC), and Tabular (TAB) reached values between 12 and 15% in at least one of the samples. Within the Tabular (TAB) category, Tabular psilate (TAB_PSI) and Tabular granulate (TAB_GRA) were observed in all samples at less than 10%, except for sample CMN1_81, which did not present TAB_PSI. Meanwhile, Tabular entire to sinuate (TAB_ENT/SIN) was recorded in samples CMN1_81 and CMN1_82, with percentages of 0.7% and 3.2%, respectively. Notably, articulated Cones (CON), Polylobate (POL) and Papillate (PAP) phytoliths were identified in the samples CMN1_81, CMN1_82 and CMN1_83, respectively, with percentages below 3%. In the SAL feces, Rondels (RON), Bilobate Stipa type (BIL_Stipa type), Crenate (CRE), Elongate columnar/clavate (ELO_COL/CLA), Acute bulbosus (ACU_BUL), and Tracheary annulate/helical (TRA_ANN/HEL) morphotypes exhibited the highest percentages, each surpassing 10% but not exceeding 18% in any sample. Tabular psilate (TAB_PSI) and Tabular granulate (TAB_GRA) were observed in all SAL samples, ranging from 9% to 25%. Tabular entire to sinuate (TAB_ENT/SIN) was recorded in sample SAL_88 at 7%. Additionally, articulated Cone (CON) and Papillate (PAP) phytoliths were identified in sample SAL_92, and Saddle (SAD) phytoliths in sample SAL_97, all with percentages below 3%. Spheroid ornate and psilate (SPH PSI/ORN) morphotypes were observed in all SAL samples, with percentages ranging from 3% to 5%. The remaining of Non-diagnostic morphotypes accounted for less than 10% in both units, CMN1 and SAL. Unidentified phytoliths (<2%) were observed fragmented and/or showing signs of degradation (Figures 5 and 6a–k).

Figure 5.

Percentage diagram of phytolith morphotypes found in feces from CMN1 and SAL sites.

Figure 6.

Images obtained under an optical microscope showing phytoliths (a–k), pollen (l–p), and plant fragments (q–x) identified in Lama guanicoe feces from CMN1 and SAL.

Multivariate analysis (PCA) of feces phytolith

The principal component analysis (PCA) based on the phytolith assemblages of feces from CMN1 and SAL explained 96% of the total variance, with the first three dimensions capturing the majority of the variance (68.9% for Dim1, 22.1% for Dim2 and 5% for Dim 3; Figure 7). The analysis of variable contributions and their graphical representation showed a stronger association between the SAL feces and the Crenate, Trapezoid, and Bilobate morphotypes, which have been described in the phytolith assemblage of P. humilis. In contrast, the CMN1 feces showed a greater association with the Rondel morphotype, which has been observed in higher abundance in the phytolith assemblage of F. pallescens and the Tabular morphotype described for dicotyledonous species. There is also an association between the SAL feces and the Spheroid morphotype, which was identified in Lycium chilense leaves in this study. The partition generated by the analysis suggests the existence of two groups: one comprising the samples from CMN1 and the other containing the samples from SAL. The membership values of each sample to both groups is shown in Table 2.

Figure 7.

Biplot of the principal component analysis (PCA) of feces samples based on the relative frequencies of phytolith morphotypes: (a) dimension 1 versus 2 and (b) dimension 1 versus 3.

Table 2.

Membership values of each sample to groups G1 and G2 constructed using fuzzy c-means clustering.

Maximum membership value = 100.

Pollen in feces

A total pollen sum ranged between 247and 473 in the feces of CMN1, whereas in the analysis of SAL feces, the total pollen sum ranged between 85 and 239. No signs of degradation were observed in the pollen grains identified. The number of pollen types identified was higher in CMN1 (29) than in SAL feces (24). In CMN1 feces, pollen types of Poaceae were identified in lower abundance than SAL feces (19.45.% and 43.10%, respectively). Iridaceae (Monocotyledonous) pollen type reached a high percentage value in one sample (CMN1_82). Within the Dicotyledons the most abundant pollen types were Cerastium (6.8%), Mulinum (14.71%), Colliguaja (17%), Asteraceae subf. Cichoroideae (13.5%), and Fabaceae (8.2%). Pollen types of Colobanthus, Geraniaceae, Asteraceae, Rosaceae, Anacardiaceae, Solanaceae, Brassicaceae, Plantago, Nothofagus, Misodendrum and Podocarpus were identified in lower abundance. Indeterminate pollen was <8.4%. According to the criteria established to account for differences in pollen production between anemophilous and zoophilous taxa (see Methods), Poaceae reached concentrations higher than 10,000 grains/g (72,297 grains/g) in sample CMN1_82. Similarly, Iridaceae reached 11,002, 20, and 409 grains/g in samples CMN1_81 and CMN1_82, respectively, while Rumex reached 37,720 grains/g in CMN1_81. These taxa correspond to anemophilous taxa. In addition, zoophilous taxa such as Caryophyllaceae, Cerastium, Colobanthus, Geraniaceae, Mulinum, Azorella, Asteraceae subf. Asteroideae, Asteraceae subf. Cichorioideae, Nardophyllum, Nassauvia, Fabaceae, Colliguaja, Rosaceae, Anacardiaceae, Gomphrena, and Brassicaceae were recorded at high concentrations in at least one sample. (Table 3).

Table 3.

Concentrations of pollen types (grains/g) and total pollen concentration of feces from CMN1 and SAL.

Pollen types	CMN1			SAL
Pollen types	F81	F82	F83	F88	F92	F97
Poaceae	72297	4329	4304	6739	16926	2832
Cyperaceae	0	412	108	0	164	0
Iridaceae	11002	20409	430	0	329	508
Caryophyllaceae	0	618	3551	1078	2136	73
Cerastium	18074	4329	1829	0	0	73
Colobanthus	0	1443	0	0	0	0
Geraniaceae	4715	206	215	0	0	0
Mulinum	27504	10101	1722	0	0	436
Azorella	39292	0	0	0	0	145
Asteraceae	0	1237	108	0	0	0
Asteraceae subf. Asteroideae	3143	0	430	3504	164	290
Senecio	0	0	430	2156	822	436
Asteraceae subf. Cichoroideae	50294	1443	108	0	0	73
Nardophyllum	0	2061	1399	809	1643	145
Nassauvia	3929	0	215	2695	1315	1089
Verbenaceae	0	0	0	0	1972	0
Fabaceae	30648	2474	1506	0	329	0
Colliguaja	7858	5360	4519	1078	329	145
Rosaceae	6287	0	0	0	493	0
Anacardiaceae	3929	0	0	539	0	0
Schinus	786	0	0	270	0	0
Solanaceae	0	206	0	0	0	0
Empetrum	0	0	108	0	0	73
Chenopodiaceae	0	0	0	2156	11010	1017
Gomphrena	3929	618	0	270	0	1162
Brassicaceae	20432	0	0	0	0	0
Plantago	4715	825	0	0	657	145
Rumex	37720	618	108	270	164	73
Nothofagus	1572	1855	2690	539	822	508
Podocarpus	0	1443	1937	0	0	145
Misodendrum	0	2886	430	0	0	73
Unidentified	23575	5772	430	809	0	290
Total pollen concentration	371702	68647	26578	22911	39276	9731

In the SAL feces, the most abundant pollen types were Poaceae (43.10%), Chenopodiaceae (28%), Asteraceae subf. Asteroideae (15.29%), and Nassauvia (11.76%), all of which were found in higher abundance than in the CMN1 feces. In a low percentage were identified pollen types of Iridaceae, Caryophyllaceae, Cerastium, Asteraceae Senecio type, Colliguaja, Asteraceae subf. Cichoroideae, Rosaceae, Anacardiaceae, Schinus, Empetrum, Plantago, Rumex, Nothofagus, Podocarpus. Indeterminate pollen types were <3.5% (Figure 6l–p and Figure 8). In SAL feces, pollen concentration values higher than 10,000 grains/g for anemophilous taxa were recorded for Poaceae (16,926 grains/g) in at least one sample. In addition, Caryophyllaceae, Asteraceae subf. Asteroideae, Senecio, Nardophyllum, Nassauvia, Verbenaceae, Colliguaja and Gomphrena, zoophilous taxa, showed pollen concentrations values higher than 1000 grains/g in at least one sample (Table 3).

Figure 8.

Percentage diagram of pollen found in feces from CMN1 and SAL sites.

Multivariate analysis (PCA) of feces pollen

Pollen composition and abundance were also studied to determine whether the feces could be differentiated. The PCA based on the relative frequencies of pollen explained 84% of the total variance with the first three dimensions accounting for the majority of the variance (42.2% for Dim1, 24.8% for Dim2, and 17% for Dim3). The PCA results showed a differentiation between CMN1 and SAL feces compositions (Figure 9). Similarly, the fuzzy c-means clustering analysis identified two distinct groups: one composed of the SAL feces and the other consisting of the CMN1 feces (Table 2). The analysis of variable contributions and their graphical representation in the Dim1 versus Dim2 space showed that the most influential variables were Colliguaja, followed by Poaceae, Azorella, Mulinum, and Nardophyllum. In contrast, the Dim1 versus Dim3 representation indicated that Chenopodiaceae, Verbenaceae, Senecio, and Nassauvia contributed most to the variation, followed to a lesser extent by Mulinum, Cerastium, Poaceae, and Asteraceae subf. Asteroideae. In the graphical representation of three dimensions, feces from CMN1 were primarily associated with Colliguaja, Mulinum, Cerastium, Azorella, and Fabaceae as the main pollen types, while feces from SAL were mainly associated with Poaceae, Chenopodiaceae, Nassauvia, Senecio, Verbenaceae, and Asteraceae subf. Asteroideae.

Figure 9.

Biplot of the principal components analysis (PCA) of feces based on the relative frequencies of pollen: (a) dimension 1 versus 2, (b) dimension 1 versus 3.

Plant fragments in feces

A higher number of plant taxa was identified in the feces from the CMN1 site compared to those from SAL. The proportion of indeterminate plant fragments in CMN1 samples ranged from 7.1% to 23%, whereas in SAL samples it varied between 5.9% and 10.4%. At CMN1 site, the most representative taxa were Poaceae species, particularly Poa ligularis (51.9%), Pappostipa chrysophylla (7%), Festuca pallescens (6.1%), and unidentified Poaceae (17%). Among dicotyledons, Erodium cicutarium (11.2%) and Colliguaja integerrima (3.3%) were also present. Feces from SAL showed higher proportions of Pappostipa humilis (31.7%), Pappostipa sp. (16.6%), and Poa ligularis (45%). The most abundant dicots in SAL were Nardophyllum bryoides (9.3%) and Atriplex sagittifolia (6.6%; Figures 6q–x and 10).

Figure 10.

Percentage diagram of plant fragments found in feces from CMN1 and SAL sites.

Multivariate analysis (PCA) of feces plant fragments

The PCA based on plant fragments explained 89.8% of the total variance, with the first three dimensions accounting for the majority (41.7% for Dim1, 27.8% for Dim2, and 20.3% for Dim3). In the Dim1–Dim3 planes, the samples SAL_92 and SAL_97 from the SAL site are grouped together, while CMN1_82 and CMN1_83 form a separate cluster (Figure 11). The fuzzy c-means clustering analysis also identified two distinct groups corresponding to the SAL and CMN1 samples (Table 2). However, samples CMN1_81 and SAL_88 showed partial membership in both clusters (G1 and G2), suggesting intermediate or mixed characteristics. The analysis of variable contributions and their graphical representation of three dimensions showed that the most influential taxa were the monocotyledons Rytidosperma sp., Festuca pallescens, and Pappostipa sp., along with the dicotyledons Asteraceae, Cerastium arvense, Erodium cicutarium, Colobanthus lycopodioides, Atriplex sagittifolia, and Nardophyllum bryoides. These were followed by Deschampsia antarctica, Distichlis sp., Pappostipa humilis and Poa ligularis within Poaceae, as well as Colliguaja integerrima, Clinopodium darwinii, among dicotyledons and Ephedra sp. As a result, feces from the CMN1 site were primarily associated with Rytidosperma sp., Festuca pallescens, and the dicotyledons Asteraceae, Cerastium arvense, Erodium cicutarium, and Colobanthus lycopodioides, while feces from SAL were mainly associated with Pappostipa sp., Pappostipa crisophylla, Atriplex sagittifolia, and Nardophyllum bryoides.

Figure 11.

Biplot of the principal components analysis (PCA) of feces based on the relative frequencies of plant fragments: (a) dimension 1 versus 2, (b) dimension 1 versus 3.

Discussion

Vegetation characterization

The results of vegetation surveys conducted at sites CMN1 and SAL are consistent with the taxa previously reported by Cabrera (1976) and Oliva et al. (2001) for their respective phytogeographic regions and vegetation units. CMN1 site corresponds to a shrub-grass steppe, whereas SAL is characterized as a sub-shrub steppe. This distinction highlights the heterogeneity of plant communities within the study area, which is relevant for identifying foraging areas in two contrasting environments. At CMN1 site, the dominance of Azorella prolifera (=Mulinum spinosum), followed by Festuca pallescens, and the presence of accompanying species such as Colliguaja integerrima, align with the plant communities described by Oliva et al. (2001) for this study region. These authors characterize these steppes as dense cushion plant communities with high floristic richness. Additionally, the occurrence of Berberis buxifolia, Chiliotrichum sp., Senecio filaginoides, and Pappostipa speciosa in our surveys supports this characterization. In contrast, the SAL site is dominated by species typical of semiarid environments with high salinity and limited water availability. The abundance of Nassauvia glomerulosa and Pappostipa humilis, along with Lycium chilense, Azorella trifurcata, and Atriplex sagittifolia, reflects the presence of plant communities adapted to restrictive edaphic conditions. According to Oyarzabal et al. (2018), low shrub steppes are characterized by the presence of Atriplex sagittifolia. Although the number of vegetation surveys carried out so far is limited, the data obtained are consistent with previous descriptions of the regional vegetation (Cabrera, 1976; Oliva et al., 2001). No systematic quantitative vegetation surveys have been previously reported for the immediate study area. In this context, the present results provide new micro-scale quantitative information on vegetation composition, allowing the identification and comparison of dominant taxa within each vegetation unit. This quantitative framework is particularly relevant for interpreting plant assemblages recovered from fresh feces samples using multiple proxies, as it allows a comparison between local vegetation availability and dietary items selected by guanaco. Overall, the results suggest the dominance and presence of taxa with well-defined ecological characteristics specific to each site, reinforcing the environmental differentiation between CMN1 and SAL.

Plant phytolith analysis

The phytolith assemblages of the taxa included in the reference collection examined in this study partially agree with previous findings for plant species of the same families in other regions, although their phytolith contents were variable (Ball et al., 2007; Benvenuto et al., 2013; Brown, 1984; De Rito et al., 2018; Fernández Honaine et al., 2006; Mulholland, 1989; Piperno, 2006; Twiss, 1992). The dominant morphotypes found in the analyzed species of Poaceae include phytoliths characteristic of the leaf epidermal tissue of species in the Pooideae subfamily. Particularly, the presence of the Bilobate morphotype in Pappostipa humilis allows it to be distinguished from other Pooideae. Benvenuto (2011) studied the species Azorella trifurcata (Apiaceae) and identified it as a non-phytolith producer, whereas in this study, Azorella prolifera (=Mulinum spinosum) showed silicifications, mainly associated with xylem tissue. Benvenuto et al. (2013) and De Rito et al. (2018) have identified morphotypes derived from epidermal tissues of Asteraceae which is consistent with the phytolith assemblages described for Nardophyllum bryoides and Nassauvia glomerulosa. However, these studies did not describe the silica skeletons composed of Oblong to Ovate granulate phytoliths observed in Nardophyllum bryoides, nor the Elongate entire to Oblong entire phytoliths recorded in Nassauvia glomerulosa in this study. In the study conducted by Piperno (2006) and Ball et al. (2007), phytoliths were observed in the leaf tissues of Euphorbiaceae, specifically in the epidermis, mesophyll, and vascular tissues. In the present study, phytolith morphotypes were primarily associated with Colliguaja integerrima epidermal tissue, such as tabular, Acute bulbous, and Stomatal complexes, were observed, along with xylem structures such as Tracheary annulate/helical phytoliths. In relation to the Solanaceae family, Mercader et al. (2009) identified phytoliths in association with both epidermal and xylem tissues. In this study, the same phytoliths were observed in Lycium chilense, characteristic of the epidermal tissue, and Spheroid psilate and Spheroid ornate phytoliths, found in mesophyll tissue.

The separation between the relative frequencies of leaf phytoliths from grasses and dicotyledons showed in PCA analysis, even between the two grass species analyzed (Festuca pallescens and Pappostipa humilis), suggest that the leaf phytolith assemblages of these species could be key for taxonomic differentiation. On the other hand, among the analyzed dicotyledonous species, Lycium chilense was the only producer of Spheroid phytoliths, which allows us to consider the possibility that this morphotype is present within the studied assemblage. Although the phytolith reference collection does not encompass the full plant diversity of the study area, it includes the dominant taxa and those most likely consumed by guanaco, contributing to the development of an initial interpretative framework.

Feces proxies analysis

The different floristic associations identified through the vegetation survey at each site are reflected in the integrated multiproxy approach applied to guanaco feces from each sampling site. Differences between the two studied sites, as revealed by PCA and clustering analyses of phytolith assemblages, pollen types, and plant fragments, suggest that feces analyses may reflect locally available vegetation within this Patagonian landscape. These results highlight the potential of combining multiple biological proxies from modern feces to explore the identification of relative foraging areas of Lama guanicoe. At the same time, the correspondence between feces contents and surrounding vegetation may be associated with the territorial behavior of the guanaco that produced the analyzed samples, as sampling took place during the reproductive season, when guanaco tend to organize into relatively stable social groups with a certain degree of territoriality (Franklin, 1982). Increasing the number of vegetation surveys and modern feces samples across different seasons would help to strengthen this actualistic framework.

The good state of preservation of phytoliths, pollen, and plant fragments in the feces samples, together with the low proportion of indeterminate morphotypes and unidentified plant fragments, supports the retrieval of reliable information for identifying dietary items and foraging areas of Lama guanicoe. Nevertheless, in some cases, taxonomic identification of plant fragments was not possible due to the absence of diagnostic features, likely related to a high degree of fragmentation associated with the digestive physiology of Lama guanicoe, which represents a limitation in terms of taxonomic resolution.

Plant fragments are direct indicators of plant consumption. The plant fragments results showed a high prevalence of Poaceae in guanaco feces from both environments analyzed. This is consistent with findings reported for feces from Tierra del Fuego (e.g. Bonino and Pelliza Sbriller, 1991; Muñoz and Simonetti, 2013). Likewise, the phytolith and pollen analyses of feces conducted in this study support this high prevalence of Poaceae, providing an important signal of grasses as a food resource for guanaco in these environments. In addition, the high pollen concentration values in samples CMN1_81 and SAL_92 suggest that the Poaceae pollen type represents a food item (Table 3). According to this, Pelliza-Sbriller et al. (1997) report a marked predominance of grass-based food items, accompanied by variable proportions of woody and graminoid species, whose relative importance varies according to the availability of other forage taxa. In particular, the phytolith assemblages of CMN1 feces showed diagnostic grass phytoliths, with a higher percentage of types attributable to Festuca pallescens than in SAL feces. Previous studies have reported phytolith assemblages similar to those observed for Festuca pallescens and other Pooideae species, such as Deschampsia antarctica and Poa ligularis (Benvenuto et al., 2013; Gallego and Distel, 2004). The latter two species were also identified in the feces of this study through the analysis of plant fragments. This overlap suggests a possible correspondence between the evidence provided by phytoliths and plant fragments regarding the greater representation of F. pallescens, D. antarctica, and P. ligularis in the feces from CMN1. In addition, both proxies also indicate a higher proportion of phytoliths and plant fragments attributable to species of the genus Pappostipa at the SAL feces. These results indicate the complementarity of these proxies for identifying Poaceae (a dominant taxon in the diet of these herbivores) at different levels of taxonomic resolution.

In the feces from both sites, phytolith assemblages were identified that matched dicotyledonous species also recognized through plant fragments and pollen types, including Nardophyllum bryoides, Nassauvia, Mulinum, Colliguaja integerrima, and Lycium chilense. These findings may complement the information on plant consumption and support the validation of these proxies in the characterization and identification of foraging sites. However, in the phytolith analysis, since the Spheroid morphotype has been reported for various woody species, its exclusive origin from Lycium chilense cannot be confirmed (Mercader et al., 2009). For dicotyledonous taxa, this study recognizes the limitations of phytolith analysis for taxonomic identification, as certain morphotypes may be shared among multiple species or groups, reducing their diagnostic value (Rovner, 1971). Nevertheless, the identification of this woody taxon through pollen and plant fragments may support the signal inferred from phytoliths, highlighting the potential of the multiproxy approach.

In the feces from SAL, the exclusive presence of plant fragments of Atriplex sagittifolia and Chenopodiaceae pollen types indicates the consumption of these plants and supports the vegetation survey data recorded at this site. On the other hand, Azorella prolifera fragments were not identified in the CMN1 feces; however, the presence of the Mulinum (= Azorella prolifera) pollen type in samples CMN1_81, 82, and 83, together with its high concentration values (27,504 grains/g; 10,101 grains/g), supports its consumption and corroborates the vegetation survey results (Table 3). Romano (2012) identified fruit remains of Mulinum in guanaco feces from Perito Moreno National Park (site located about 70 km south of Pueyrredón lake), suggesting that it may be part of their diet. Moreover, plant fragments of Erodium cicutarium (Geraniaceae) were observed in sample CMN1_81, and in combination with the Geraniaceae pollen type, which showed a high pollen concentration (4715 grains/g), this may suggest its inclusion in the diet, although this taxon was not recorded in the vegetation survey (Table 3). Similarly, plant fragments of Clinopodium darwinii were detected in the CMN1 feces, although no pollen types associated with this genus were observed. In the case of Colobanthus and Cerastium, both were identified through plant fragments and pollen, mainly in CMN1, indicating complementarity between these proxies in corroborating the vegetation survey data.

Given that Lama guanicoe is characterized as a generalist herbivore with high trophic plasticity and is therefore expected to consume plants according to local availability, the high pollen concentration values and the presence of certain taxa characteristic of CMN1 but recorded in SAL feces suggest occasional foraging beyond the defined limits of each vegetation unit. In particular, the high pollen concentration values of Colliguaja and Nardophyllum, taxa typical of the shrub–grass steppe (CMN1 site), in SAL feces suggest that these plants were also part of the guanaco diet and may reflect a broader home range extending beyond the defined vegetation units of each sampling site (Table 3). To refine this interpretation, increasing the number of samples and incorporating vegetation surveys in intermediate units between the selected sites would help to better constrain the relative contribution of local versus non-local signals.

The presence of plant fragments, pollen types, and phytolith morphotypes affiliated with Asteraceae at both sites indicates that this family was present in the study area and may have formed part of the guanaco diet. Pollen types considered anthropogenic indicators (Rumex, Brassicaceae, Asteraceae subf. Cichorioideae, Plantago, and Gomphrena) were present in the feces at low concentrations and percentages, suggesting that they were likely not intentionally consumed by guanaco (Velázquez et al., 2019). On the other hand, it is important to note that the higher abundance of Nothofagus pollen in the modern feces from CMN1 compared to SAL suggests that guanaco may have foraged near Nothofagus forests or in locations where pollen rain was dominated by this taxon. Although some studies report that guanaco consumes Nothofagus shoots, and forest development begins only about 15 km northwest of the CMN1 site, no plant fragments of this taxon were identified in the present study. This result suggests that the presence of Nothofagus pollen in the feces samples likely resulted from secondary incorporation (via contaminated food, drinking water, or inhaled air), as westerly winds transport Nothofagus pollen from nearby forested areas (Horta et al., 2016). These observations underscore the need to consider the potential contribution of non-dietary pollen, including incidental ingestion and secondary deposition, when interpreting pollen assemblages from feces samples. In this context, the combined use of multiple proxies provides a more robust framework to distinguish between signals derived from direct consumption and those resulting from environmental inputs.

In this context, the contribution of this work lies not only in the identification of consumed taxa within each analyzed vegetation unit, but also in the multiproxy assessment of dietary signals and their spatial resolution, as well as in exploring the complementarities and limitations of analyzing these proxies together. Taken together, this approach provides a preliminary actualistic framework for paleoecological studies that may contribute to improving the interpretation of coprolite assemblages, particularly in terms of assessing spatial patterns of herbivore–environment interactions. However, such interpretations should be made with caution, as distinguishing foraging areas in Holocene samples requires considering additional limitations of the fossil record not addressed in the current model, including taphonomic processes that may bias the information derived from coprolites.

Conclusions

This research constitutes the first multiproxy analysis of Lama guanicoe feces from the Posadas-Pueyrredón-Salitroso Lacustrine System and the first development of a phytolith reference collection for native plant species from this region. The development of a local phytolith reference collection represents an essential step toward improving taxonomic accuracy and the reliability of ecological interpretations. However, phytolith analysis of dicotyledonous taxa still requires complementary approaches, such as morphometric analyses, to enhance diagnostic precision.

The study highlights the importance of assessing the potential of each proxy (phytoliths, pollen, and plant fragments) both independently and in combination, for reconstructing guanaco diet and habitat use across different environments. The analyses distinguished two discrete foraging areas. Feces from CMN1 were primarily associated with shrub–grass steppe taxa, whereas feces from SAL showed a significant association with sub-shrub steppe vegetation.

The interdisciplinary approach adopted here enabled the identification of dietary items with different levels of taxonomic resolution. For instance, the phytolith assemblage associated with Pappostipa humilis suggests that the leaf phytoliths of this species could serve as key taxonomic markers in semi-desert environments, especially when plant fragments are not preserved.

This framework provides a preliminary basis for interpreting fossil feces and offers insights into herbivore–environment interactions in Patagonia. Moreover, the study contributes to regional reference datasets and outlines a methodological approach that may support future research in the region.

Supplemental Material

sj-xlsx-1-hol-10.1177_09596836261458233 – Supplemental material for Differentiation of foraging areas of Lama guanicoe (Camelidae) in Patagonia using multiproxy analysis of feces: Modern reference dataset for paleoecological reconstruction

Supplemental material, sj-xlsx-1-hol-10.1177_09596836261458233 for Differentiation of foraging areas of Lama guanicoe (Camelidae) in Patagonia using multiproxy analysis of feces: Modern reference dataset for paleoecological reconstruction by María Laura Benvenuto, Nadia Jimena Velázquez, Ana Cecilia Martínez Tosto, Mauro Chaparro, Romina Sandra Petrigh, Ivana Silvia Camiolo and Lidia Susana Burry in The Holocene

Footnotes

Acknowledgements

We thank Carlos Otamendi Jr., Ranch Pueyrredón, and Ranch El Bagual for allowing access to their ranches and for their hospitality during fieldwork. We are also grateful to archeologists Teresa Civalero, Damián Bozzuto, and Carlos Aschero; to Ricardo Vázquez, Natalia Morrone, Cecilia Serpa, and Carolina Ávila, members of the Museo de Arqueología Carlos Gradín (Perito Moreno, Santa Cruz); to the Asociación Identidad Perito Moreno; and to the Municipalidad de Perito Moreno, Santa Cruz, for their involvement in the logistics of the 2019 field campaign in the Lake Pueyrredón area.

ORCID iDs

María Laura Benvenuto

Nadia Jimena Velázquez

Ana Cecilia Martínez Tosto

Mauro Chaparro

Romina Sandra Petrigh

Ivana Silvia Camiolo

Lidia Susana Burry

Ethical considerations

This study did not require approval from an ethics committee.

Consent to participate

Participant consent to take part in this work was obtained verbally.

Author contributions

María Laura Benvenuto: Conceptualization; Methodology; Investigation; Formal analysis; Writing – original draft; Writing – review and editing; Project administration; Funding acquisition.

Nadia Jimena Velázquez: Conceptualization; Methodology; Investigation; Formal analysis; Writing – review and editing; Project administration; Funding acquisition.

Ana Cecilia Martínez Tosto: Conceptualization; Methodology; Investigation; Formal analysis; Writing – review and editing; Project administration; Funding acquisition.

Mauro Chaparro: Conceptualization; Methodology; Investigation; Formal analysis.

Romina Petrigh: Conceptualization; Methodology; Investigation; Writing – review and editing; Project administration; Funding acquisition.

Ivana Silvia Camiolo: Conceptualization; Methodology; Investigation; Formal analysis.

Lidia Susana Burry: Conceptualization; Supervision; Investigation; Writing – review and editing; Project administration; Funding acquisition.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by the Agencia Nacional de Promoción Científica y Tecnológica, Ministerio de Ciencia y Técnica (PICT 0455/17; PICT 02815/20; PICT GRF-T1-0748/21), Universidad Nacional de Mar del Plata (EXA 1109/22; EXA 1223/24), and the Consejo Nacional de Investigaciones Científicas y Técnicas (PIP 0942/2022).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The data supporting the findings of this study are available in the supplementary material and upon reasonable request from the corresponding author. The research data associated with this work will be archived in the official institutional repository of CONICET (Argentina)*.

Supplemental material

Supplemental material for this article is available online.

References

Agliano

(2022) Análisis microhistológico y palinológico de heces y coprolitos de camélidos del Parque Nacional Perito Moreno, Santa Cruz. Tesis de grado, Licenciatura en Ciencias Biológicas, Facultad de Ciencias Exactas y Naturales, Universidad Nacional de Mar del Plata.

Agliano

Velázquez

Martínez Tosto

, et al. (2024) Camelid diet through micro histological and palynological analyses of feces and coprolites from Parque Nacional Perito Moreno, Patagonia, Argentina. Journal of Archaeological Science Reports 58: 104713.

Aschero

Bozzuto

Civalero

, et al. (2019) Se nos viene la noche. El volcán Hudson y su influencia en el NO de Santa Cruz: Integrando perspectivas arqueológicas. In: Gómez Otero

Svoboda

Banegas

(eds) Arqueología de la Patagonia: el pasado en las Arenas. Instituto de Diversidad y Evolución Austral, pp.239–249.

Ball

Baird

Al-Farsi

, et al. (2007) A survey of phytoliths produced by the vegetation of Dhofar, Oman. In: Madella

Zurro

(eds) Plants, People and Places: Recent Studies in Phytolith Analysis. Oxbow Books, pp. 29–40.

Benvenuto

(2011) Análisis fitolíticos en turberas, niveles de turba fósil y comunidades vegetales asociadas de Tierra del Fuego, Argentina. Tesis de grado, Facultad de Ciencias Exactas y Naturales, Universidad Nacional de Mar del Plata.

Benvenuto

Fernández Honaine

Osterrieth

, et al. (2013) Silicophytoliths in holocene peatlands and fossil peat layers from Tierra del Fuego, Argentina, southernmost South America. Quaternary International 287: 20–33.

Benvenuto

Martínez Tosto

Petrigh

, et al. (2022) Estudios paleoecológicos de camélidos a partir del análisis multiproxy de coprolitos del sitio CMN1, Santa Cruz. Revista del Museo de La Plata, Suplemento de resúmenes 7: 88.

Bonino

Pelliza Sbriller

(1991) Composición botánica de la dieta del guanaco en dos ambientes contrastantes de Tierra del Fuego, Argentina. Ecología Austral 1(2): 97–102.

Brown

(1984) Prospects and limits of a phytolith key for grasses in the central United States. Journal of Archaeological Science 11(4): 345–368.

10.

Burgi

(2007) Radio de acción del guanaco en el NE de Chubut. Mastozoología Neotropical 14(2): 285–291.

11.

Cabrera

(1976) Regiones Fitogeográficas Argentinas. Enciclopedia Argentina de Agricultura y Jardinería II, Acmé, Buenos Aires.

12.

Camiolo

(2023) Estudio palinológico e isotópico de heces de Lama guanicoe del sistema lacustre Pueyrredón–Posadas–Salitroso. Tesis de grado, Universidad Nacional de Mar del Plata.

13.

Campos

Labouriau

(1969) Corpos silicosos da gramineas dos Cerrados II. Pesquisa Agropecuaria Brasileira 4: 143–151.

14.

Caruso Fermé

Civalero

(2014) Holocene landscape changes and wood use in Patagonia: Plant macroremains from Cerro Casa de Piedra 7. Holocene 24(2): 188–197.

15.

Casamiquela

(1983) La significación del guanaco en el ámbito pampeano-patagónico. Mundo Ameghiniano 4: 20–46.

16.

De Rito

Fernández Honaine

Osterrieth

, et al. (2018) Silicophytoliths from a Pampean native tree community and their soil representation. Review of Palaeobotany and Palynology 257: 19–34.

17.

D’Antoni

(2008) Arqueoecología: Sistémica y Caótica. CSIC.

18.

Faegri

Iversen

(eds) (1989) Textbook of Pollen Analysis, 4th edn. Wiley and Sons.

19.

Fernández Honaine

Benvenuto

Osterrieth

(2019) An easy technique for silicophytolith visualization. Boletín de la Sociedad Argentina de Botánica 54: 353–365.

20.

Fernández Honaine

Zucol

Osterrieth

(2006) Phytolith assemblages and systematic associations in grassland species of the south-eastern Pampean plains, Argentina. Annals of Botany 98: 1155–1165.

21.

Franklin

(1982) Biology and ecology of south American camelids. In: Mares

Genoways

(eds) Mammalian Biology in South America. Pymatuning Laboratory of Ecology, University of Pittsburgh, pp. 457–488.

22.

Fredlund

Tieszen

(1994) Modern phytolith assemblages from the north American great plains. Journal of Biogeography 21: 321–335.

23.

Gallego

Distel

(2004) Phytolith assemblages in grasses native to central Argentina. Annals of Botany 94: 865–874.

24.

Garreaud

Lopez

Minvielle

, et al. (2013) Large-scale control on the Patagonian climate. Journal of Climate 26: 215–230.

25.

Grimm

(2011) Tilia software v1.7.14. Illinois State Museum.

26.

Heusser

(1971) Pollen and Spores From Chile. University of Arizona Press.

27.

Horrocks

Jones

Beever

, et al. (2002) Analysis of plant microfossils in prehistoric coprolites from Harataonga bay, Great Barrier Island, New Zealand Journal of the Royal Society of New Zealand 32: 617–628.

28.

Horta

Georgieff

Console Gonella

, et al. (2011) Registros paleobatimétricos del sistema Pueyrredón–Posadas. Serie Correlación Geológica 27: 100–109.

29.

Horta

Marcos

Bozzuto

, et al. (2016) Holocene paleoenvironmental variations in lakes Pueyrredón–Posadas–Salitroso. Palaeogeography Palaeoclimatology Palaeoecology 449: 541–552.

30.

Jouy-Avantin

Debenath

Moigne

, et al. (2003) A standardized method for the description and the study of Coprolites. Journal of Archaeological Science 30: 367–372.

31.

Markgraf

D’Antoni

(1978) Pollen Flora of Argentina. University of Arizona Press.

32.

Martínez Tosto

Burry

Arriaga

, et al. (2016) Archaeobotanical study of Patagonian Holocene coprolites, indicators of diet, cultural practices and space use. Journal of Archaeological Science Reports 10: 204–211.

33.

Mercader

Bennett

Esselmont

, et al. (2009) Phytoliths in woody plants from the Miombo woodlands of Mozambique. Annals of Botany 104: 91–113.

34.

Miotti

Salemme

(1999) Biodiversity and taxonomy in late pleistocene–Holocene Patagonia. Quaternary International 53–54: 53–68.

35.

Montes

De Lamo

Zavatti

(2000) Distribución de abundancias de guanacos en Tierra del Fuego. Mastozoología Neotropical 7: 23–33.

36.

Mosca Torres

de Porras

Aschero

(2022) Early-to-Middle Holocene environmental reconstruction in the southern Puna (26° S, Argentina) from camelid coprolite records. Journal of South American Earth Sciences 116: 103805.

37.

Mosca Torres

Urquiza

Aschero

(2018) Microhistological analysis of ancient camelid dung. Journal of Archaeological Science Reports 20: 347–354.

38.

Mueller-Dombois

Ellenberg

(1974) Aims and Methods of Vegetation Ecology. Wiley & Sons.

39.

Mulholland

(1989) Phytolith shape frequencies in North Dakota grasses: A comparison to general patterns. Journal of Archaeological Science 16: 489–511.

40.

Muñoz

Simonetti

(2013) Diet of guanaco in sheep-free rangeland in Tierra del Fuego, Chile. Ciencia e Investigación Agraria 40: 185–191.

41.

Neumann

Strömberg

Ball

, et al. (2019) International code for phytolith nomenclature (ICPN) 2.0. Annals of Botany 124: 189–199.

42.

Oliva

González

Rial

(2001) El ambiente en Patagonia Austral. In: Borrelli

Oliva

(eds) Ganadería sustentable en la Patagonia Austral. INTA, pp. 17–81.

43.

Oyarzabal

Clavijo

Oakley

, et al. (2018) Unidades de vegetación de la Argentina. Ecología Austral 28: 040–063.

44.

Pelliza-Sbriller

Willems

Nakamatsu

, et al. (1997) Atlas dietario de herbívoros patagónicos. INTA-GTZ.

45.

Piperno

(2006) Phytoliths: A Comprehensive Guide for Archaeologists and Paleoecologists. Alta Mira Press.

46.

Puig

Videla

Cona

(1997) Diet and abundance of guanacos. Journal of Arid Environments 36: 343–357.

47.

Raedeke

(1980) Food habits of the guanaco. Turrialba 30: 177–181.

48.

R Core Team (2024) R: A language and environment for statistical computing. R Foundation.

49.

Romano

(ed.) (2012) Identificación de restos vegetales en fecas de Lama guanicoe, Tesis de grado edn. Universidad Nacional de Mar del Plata.

50.

Rovner

(1971) Potential of opal phytoliths for use in paleoecological reconstruction. Quaternary Research 1: 343–359.

51.

Sacchi

Bozzuto

Horta

, et al. (2016) Dataciones y circulación humana. Revista Andes 27(2): 1–16.

52.

Shillito

Bull

Matthews

, et al. (2013) Geochemical and microscopic analysis of coprolites. In: Hodder

(ed.) Çatalhöyük Excavations. McDonald Institute for Archaeological Research, pp. 141–156.

53.

Sobolik

(1988) The importance of pollen concentration values from coprolites: An analysis of Southwest Texas samples. Palynology 12: 201–214.

54.

Twiss

(1992) Predicted distribution of C3/C4 grass phytoliths. In: : Rapp

Mulholland

(eds) Phytolith Systematics: Emerging Issues. Plenum Press, pp.113–128.

55.

Twiss

Suess

Smith

(1969) Morphological classification of grass phytoliths. Soil Science Society of America Journal 33: 109–115.

56.

Velázquez

Burry

(2012) Palynological analysis of Lama guanicoe modern feces. Review of Palaeobotany and Palynology 184: 14–23.

57.

Velázquez

Burry

(2019) Análisis palinológico de ítems dietarios de L. guanicoe. Boletín de la Sociedad Argentina de Botánica 54: 67–78.

58.

Velázquez

Burry

(2022) New insights into the palaeodiet and paleoenvironment of the camelids (Lama guanicoe) based on a palynological analysis of Mid-Holocene coprolites from Cueva Milodon Norte 1, Pueyrredón Lake, Patagonia, Argentina. Review of Palaeobotany and Palynology 300: 104626.

59.

Velázquez

Burry

Fugassa

, et al. (2014) Palynological analysis of camelid coprolites. Quaternary Science Reviews 83: 143–156.

60.

Velázquez

Burry

Martínez Tosto

, et al. (2021) Multiproxy analysis of omnivore and herbivore coprolites. Quaternary International 601: 130–142.

61.

Velázquez

Petrigh

Benvenuto

, et al. (2019) Diseño y evaluación de un protocolo de extracción múltiple. Anales de Arqueología y Etnología 74(2): 127–145.

62.

Viale

Bianchi

Cara

, et al. (2019) Contrasting climates at both sides of the Andes in Argentina and Chile. Frontiers in Environmental Science 7: 1–15.

63.

Wood

Vermeulen

Bolstridge

, et al. (2021) Mid-Holocene coprolites from New Zealand. Quaternary Science Reviews 263: 106992.

64.

Zuloaga

Belgrano

Anton

(2023) Flora Argentina: Flora vascular de la República Argentina. Available at: http://www.floraargentina.edu.ar

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.01 MB