Introduction

Working plants with lithic tools, either to process food or to manufacture architectural features, tools and other artefacts, has been a fundamental human activity since the earliest times (Alfaro, 1980; Shick, 1989; Stordeur, 1989; Zohary et al., 2012). However, the archaeological record offers little evidence serving to identify the processes of obtaining and transforming these materials. One of the main reasons for this is the lack of preservation of organic materials. The exceptions are wetland or extremely dry contexts. Yet even waterlogged and desiccated objects are not devoid of issues such as deformation, bacterial activity and pests (Abdel-Azeem et al., 2019; Blanchette, 2000; Piqué, 2006). Wood, especially its most resistant species, survives better than non-woody plants. This explains why these objects have benefited from more scientific contributions (e.g. Bosch et al., 2005; Caruso-Fermé et al., 2023; López-Bultó et al., 2020; Piqué, 2000; Vidal-Matutano et al., 2020) compared to non-wood plants (e.g. Herrero-Otal et al., 2023; Mineo et al., 2023; Palomo et al., 2013; Wendrich & Holdaway, 2018). Therefore, delving in depth into the actions affecting wood through the study of the lithic tools serving to work it can yield relevant data as to the conditioning of the workspaces, craftwork technological processes and storage strategies (Gutiérrez-Cuenca et al., 2017; Martín-Seijo & Panagiotakopulu, 2022; Vidal-Matutano et al., 2021).

Traces of these ‘hidden technologies’ (Hurcombe, 2014) can be identified, in part, through use-wear analyses applied to lithic (and at times even bone) tools, as these tend to better withstand the passage of time. This discipline has traditionally focused on siliceous rocks and their applications in working plants (Semenov, 1964; Keeley, 1980; Vaughan, 1985; González & Ibáñez, 1994; Hardy & Garufi, 1998; Bencomo & Jardón, 2023). Yet certain Palaeolithic research has confirmed the use of other types of volcanic rock tools for these tasks (Bello-Alonso et al., 2020; Panera et al., 2019; Plisson, 1982; Rios-Garaizar et al., 2018). Obsidian, for example, has been the subject of very compelling analyses in various chronological and cultural contexts, especially concerning its role in the origins of agriculture (Anderson-Gerfaud, 1990; Anderson-Gerfaud & Formenti, 1994; Astruc, 2022). However, much more research has focused on the role of flint in harvesting wild and domestic species in the Middle East (Anderson-Gerfaud, 1981, 1988; Ibáñez et al., 2007, 2008; Pichon et al., 2021).

In this sense, the Canary Islands, particularly the Indigenous Period (1st–fifteenth centuries AD), offers ample archaeological evidence (due to desiccation) of the exploitation of plants with lithic tools. The numerous botanical objects serving as models to reproduce through experimental work are made of wood, cultivated plants and wild plant fibres (Galván Santos, 1980; Morales et al., 2014; Vidal-Matutano et al., 2020). The Canarian archipelago was colonised, according to genetic, epigraphic, and material culture data, by groups of North African Berber/Amazigh origin (Hagenblad & Morales, 2020; Maca-Meyer et al., 2004; Mora Aguiar, 2021; Springer Bunk, 2019). Radiocarbon dating’s coupled with material evidence bear witness also to the presence of Romanised settlers on the Islet of Lobos at around the turn of the Era (Del Arco-Aguilar et al., 2020). Genetic evidence points to an East–West dynamic of occupation marked by subsequent inter-island contacts for an undetermined period (Fregel et al., 2019, 2021; Serrano et al., 2023; Santana Cabrera et al., in press). This was followed by an extended period of isolation from the African continent and between the islands themselves until the arrival of Europeans in the fourteenth century AD (Alberto Barroso et al., 2019, 2020; Morales, 2006; Pardo-Gordó et al., 2022; Velasco-Vázquez et al., 2020, 2021). These groups survived their isolation by adopting self-sufficient agro-pastoral economies based on cultivating cereals and legumes as well as on collecting endemic species (Henríquez-Valido, 2022; Morales et al., 2023) and raising livestock (goats, sheep and pigs). Their diet was complemented by ichthyo and malacofauna (Castellano-Alonso et al., 2018; Rodríguez Santana, 1994; Rodríguez-Rodríguez et al., 2021).

The way of life of the different islanders was greatly conditioned by the absence of metal ores in the archipelago. They adapted to this shortfall by knapping tools from obsidian and especially from volcanic rocks such as basalt, trachyte and phonolite (Lacave Hernández et al., 2023; Rodríguez-Rodríguez, 1993, 1998). The study of their material culture has traditionally been oriented towards identifying the technological processes to procure and transform various raw materials into products. These include investigations on plants whose findings have paved the way for experimental programmes designed according to different specific contexts. These have led to initial advances in identifying the technology linked to woodworking (Diego Cuscoy, 1968; Rivas Martínez et al., 1993; Vidal-Matutano et al., 2020, 2021) and, to a lesser extent, that of non-woody plants and the processes of their procurement, manufacture and use (Galván Santos, 1980; Morales et al., 2014).

Functional lithic industry research in the Canary Islands has until recently been limited, as noted above, to tools serving to obtain and transform plants. Initial research highlights their use in tasks of cereal grinding and woodworking specifically on the islands of La Palma, Tenerife and Gran Canaria (Naranjo-Mayor & Rodríguez-Rodríguez, 2015; Rodríguez-Rodríguez, 1993, 1998). However, there is a need to develop more robust experimental frameworks and extend the focus to a wider range of archaeological contexts. The development of archaeobotany and the documentation of working marks on wooden tools and certain plant fibres has given rise to new hypotheses regarding the role of lithic tools in harvesting cereals, woodworking and plant craftwork (Morales et al., 2014; Vidal-Matutano et al., 2020, 2021). A response to new questions requires improving experimental collections resorting to obsidian and especially in the other volcanic rocks.

This study has specifically attempted to offer a complete record of the transformation of plants based on experimentation, ethnography and archaeological research. The experiments were carried out with lithic tools similar in typology and raw material to those collected on archaeological sites. The use-wear results are presented according to the different types of volcanic raw materials, obsidian on the one hand, and on the other hand other eruptive rocks, in particular those most often used in the Canary Islands (basalts, trachytes and phonolites). Although the group are characterised by different compositions, texture and hardness, they all share a similar structure, notably an amorphous matrix with phenocrysts, which conditions the development of wear and the nature of traces. The experimentation described here therefore pursues two main objectives: (1) identify and characterise the attributes of the lithic raw materials according to use-wear traces deriving from working vegetal materials and (2) delve into the technological aspects linked to the procurement, preservation and manufacture of wood and plant fibres.

Materials and Methods

The experimental programme was designed to identify the use-wear on volcanic tools (obsidian and other volcanic rocks) after working vegetal materials. The tasks include harvesting cereals, working other non-woody plants (e.g. rushes) and carrying out actions to process palm leaves and wooden branches. The work likewise undertook steps linked to manufacture artefacts of wood and plant fibre. It was inspired by observations of the exceptionally well-preserved Canarian archaeological finds (Fig. 1a, d, g) often bearing traces of different technological actions (Fig. 2a, d, g, j). The study likewise borrowed from ethnographic research and, at the methodological level, it adopted the registration protocols established by previous use-wear analyses, notably those by Hurcombe (1992), Clemente-Conte (1997), Rodríguez-Rodríguez (1998), Kononenko (2011), Huidobro (2018) and Walton (2019).

Fig. 1
figure 1

Archaeological plant remains and different experimental tasks. a Archaeological woodworking, transversal actions on wood (Pinus canariensis) discovered at the funerary site Cruz de la Esquina, Gran Canaria, (no. 287); b experimental scraping of dry wood (Pinus canariensis) with a basalt flake (exp: 232); c experimental scraping of fresh wood (Pinus canariensis) with a basalt flake (exp: 216); d archaeological woodworking, chopping action of wood (Pinus canariensis) from the funerary site Cruz de la Esquina, Gran Canaria, (no. 285); e experimental splitting of dry wood (Ilex canariensis) by indirect percussion using a basalt flake (exp: 219); f experimental splitting of fresh wood (Ilex canariensis) by indirect percussion with a basalt flake (exp: 153); g archaeological remains of cut grass stems from the funerary site Cruz de la Esquina, Gran Canaria (no. 213); h experimental harvesting of ripe cereals (Hordeum vulgare) with a basalt sickle; i experimental reaping of fresh rush (Juncus acutus) with a basalt sickle

Fig. 2
figure 2

Archaeological plant remains linked to the craft and experimentation carried out. a Archaeological remains of a cut palm leaf (Phoenix canariensis) at the petiole from the granary of Cruz de la Esquina, Gran Canaria (no. 1075); b experimental sawing of the petiole of fresh palm leaves (Phoenix canariensis) with an obsidian flake (exp: 281); c experimental cutting of dry palm leaflets (Phoenix canariensis) with a basalt flake (exp: 275); d archaeological fragment of a rush (Juncus sp.) textile of type 1 from the funerary site of Cruz de la Esquina, Gran Canaria (no. 243); e experimental splitting of a remoistened rush (Juncus acutus) with a basalt flake (exp: 293); f experimental scraping of the pulp of a remoistened rush (Juncus acutus) with an obsidian flake (exp: 300); g archaeological twisted cord (Juncus sp.) from the funerary site of Cruz de la Esquina, Gran Canaria (no. 17–6); h experimental splitting of remoistened palm leaflets (Phoenix canariensis) with a basalt flake (no. 307); i experimental twisted cord of palm leaflets (Phoenix canariensis); j archaeological rush mat (Juncus sp.) of type 1 from the funerary site of Cruz de la Esquina, Gran Canaria (no. 1086); k-l experimental reconstruction of a funerary rush (Juncus acutus) bundle of type 1 from Gran Canaria

The Experimental Programme

The experimental tools, when possible, were manufactured with the same raw materials and following the technological processes and typologies identified at sites elsewhere in the Canary Islands. The knapping techniques consisted mainly of direct percussion with hard hammers (stone) and to a much lesser extent soft tools (wood). Indirect percussion has been only identified to date for the manufacture of querns. It consists of combining picks as intermediate tools with wood mallets. Obsidian, especially the small fragments, were also knapped using the bipolar technique on an anvil (Galván Santos & Hernández Gómez, 1996; Galván Santos et al., 1987; Lacave Hernández et al., 2023; Naranjo-Mayor & Rodríguez-Rodríguez, 2015; Rodríguez-Rodríguez, 1993, 1998). The knapping methods were intended mainly to produce flakes. Retouched flakes are not abundant and there are also few uni- or bifacial tools and picks.

Obsidian is only located on three islands of the Canary Islands. Here, we resorted to three varieties. The first is black phonolite marked at times by brown bands from the El Tabonal de los Guanches lava flow, located on the northwestern area of the Island of Tenerife. It is semi-translucent revealing under light honey-green colours. Its knapped fractures vary from very fine to coarse (Hernández Gómez, 2006) (Fig. 3a). The second and third types of obsidians are from two areas of Gran Canaria. Ignimbrite obsidian comes from the Hogarzales-El Cedro mine complex located to the west of the island. It is a trachytic glass which is opaque and coloured with different tones from green to blue-grey. Its aspect after knapping varies also from very fine to a slightly coarse (Fig. 3b). The third is an ignimbritic, phonolitic obsidian from the south of the island. This type, collected in open areas such as ravines and slopes, is black, semi-translucent, green under light and when knapped reveals a very fine fracture (Martín Rodríguez et al., 2001) (Fig. 3c, d).

Fig. 3
figure 3

Photographs (100x) illustrating the variation of the topography of the ventral surface of unused obsidian tools. a Phonolite obsidian from Tenerife; b trachyte obsidian from the mines of Hogarzales-El Cedro of Gran Canaria; c, d phonolite obsidian from Gran Canaria

The volcanic rocks serving for the experimentation, present on every island of the archipelago, are of the type most often found in Canarian archaeological sites: basalt, trachyte and phonolite. All share an amorphous matrix containing phenocrysts (plagioclase, olivines, pyroxenes…) which vary in size to naked eye from large to cryptocrystallines. Consequently, to the naked eye they reveal different knapped surfaces ranging from very coarse to fine (Fig. 4 a, b). Basalts are often black, phonolites grey and trachytes green to brown. The experimental tools were knapped mainly from cobbles of different size from beaches and ravines of La Palma, Tenerife and Gran Canaria. A few exceptions consist of fragments of trachyte from the Hogarzales-El Cedro mines, which bear finer knapped surfaces than the others (Fig. 4c, d).

Fig. 4
figure 4

Photographs (100X) illustrating the variation of topography of the ventral surface of unused phenocryst volcanic rocks. a Basalt; b phonolite; c, d trachyte with phenocryst of varying size

Most of the experimental tools (88,82%) consisted of flakes. The few remaining consisted of blades (9,32%), picks (1,24%) and knapped pebbles (0,62%) (Fig. 5). Experimental work, except for the harvesting of cereals and rushes, was carried out with unhafted tools as the morphology of the archaeological examples do not lend themselves to hafting and no hafts have been identified in archaeological record. There are nonetheless references amongst written narrative sources dating to the period of contact between the native and European populations suggesting the use of goat horns and wooden handles for certain tools (Gómez Escudero [1629–1694] in Morales Padrón, 2008: 374).

Fig. 5
figure 5

Lithic tools used in the experimental programme: 1. Basalt pick to chop wood (Fig. 31); 2. Basalt flake to saw wood (a: Fig. 23, b: Fig. 24); 3. Basalt flake to saw palm petioles (a: Fig. 28, b: Fig. 29); 4. Basalt flake to saw wood (a: Fig. 25, b: Fig. 24); 5. Obsidian flake to saw wood (Fig. 7); 6. Obsidian flake to saw wood (a: Fig. 8, b: Fig. 9); 7. Obsidian retouched flake to saw wood (Fig. 6); 8. Obsidian lame to split palm leaves (Fig. 21); 9. Basalt flake to scrape wood (Fig. 30); 10. Obsidian retouched flake to scrape wood (a: Fig. 15, b: Fig. 14); 11. Obsidian flake to split wood (a: Fig. 18, b: Fig. 19); 12. Obsidian blade to harvest cereal (Fig. 11); 13. Trachyte flake to harvest cereal (Fig. 26); 14. Basalt flake to harvest cereal (Fig. 27); 15. Obsidian flake to harvest rushes (Fig. 11); 16. Obsidian blade to split/scrape rushes (Fig. 20); 17. Trachyte flake to scrape rushes (Fig. 32); 18. Obsidian retouched flake to saw palm petioles (a: Fig. 12, b: Fig. 13); 19. Obsidian flake to split rushes (a: Fig. 16, b: Fig. 17); 20. Trachyte flake to split palm leaves (Fig. 34)

The experimentation comprised a total of 161 tools (82 obsidian; 79 different volcanic rocks) as listed in the spreadsheets (Supplementary file 1).

The stone tools serving for the wood experimental work consisted of 38 obsidian and 28 volcanic rocks. They are for the most part unretouched flakes followed by a few retouched cases, a pick and a bifacial pebble. The wood used for the experimentation included both endemic and foreign species, mostly known from the archaeological record (Machado Yanes, 1996; Vidal-Matutano et al., 2020) (Table 1). Following ethnographic examples, the wood consisted of both dry and fresh pieces. As the experimentation aimed at fashioning both roughouts and finished artefacts, the tasks included sawing, scraping and chopping and, to a lesser extent, splitting and drilling (Fig. 1b, c, e, f). These tasks required different timeframes ranging from 2 to 120 min (see Supplementary data 1).

Table 1 Plant species serving for the experimental work (X = abundance of silica)

The second set of experiments intended to shed light on working non-woody plants. The harvesting of ripe cereals involved 17 obsidian and 18 volcanic experimental tools. Fresh rushes in turn, were worked with 9 obsidian and 9 volcanic rocks (Table 1). Straight or curved three-piece hafted sickles served to reap the cereals and rushes, an action carried out at ground level (Fig. 1h). Rushes, in turn, were cut in bundles about 10 cm from the ground (Fig. 1i). The time involved in this work ranged from 4 to 180 min.

The third set of experiments involving 18 obsidian and 24 volcanic rock tools was oriented towards preparing fibres to manufacture artefacts (cordage, baskets and textiles). Rushes and palm leaves were collected whilst fresh. Then they were dried and finally remoistened for further work (Table 1). Different types of unhafted flakes of several sizes served to cut the bases of the palm leaves (Fig. 2b, c), split the rush stems, scrape their interior to remove the pulp and render the fibre more pliable (Fig. 2e, f) and split the palm leaves to make twisted rope (Fig. 2h, i). A funerary mat made of type 1 fabric (according to the typology of Galván, 1980) required the plant fibres to be braided and joined manually prior to cutting off the elements protruding from the cords and the ends (Fig. 2k, l). In these cases the time of use varied from 5 to 60 min.

The method also included a systematic recording of the weather conditions and the type of soil as they (e.g. dry and hard) could have an influence on the use-wear. Another key element recorded was of the time required by each action. All variables were then registered in specific forms and accompanied, especially during the final stages of the tasks, by photographs and videos.

Cleaning

The experimental tools were cleaned with water and neutral soap before placing them in an EMAG Emmi-H30 ultrasonic tank containing demineralised water, soap and pure acetone [C₃H₆O] during intervals of 20 and 25 min. In the cases where this process did not suffice to remove the residues, the procedure was repeated by adding pure alcohol (96%).

Tools and Laboratory Protocols

The stone tools were examined with three different types of microscopes, notably a Nikon SMZ1000 binocular magnifying glass (henceforth BMG) with a magnification of 8 × to 80 × to observe the macro traces. A Nikon Labophot-2 metallographic microscope (henceforth MM) with a magnification of 100 × to 400 × to view the micro traces and a Nikon Eclipse MA100 inverted microscope with a magnification of 50 × to 400 × to view the microtraces on the larger pieces. Photographs were taken with a DS-Fi2 camera at various heights controlled manually with NIS-Elements software (version 4.30). The images were compiled with the Helicon Focus 6 focus stacking programme prior to adding scales with Photoshop CS4.

Criteria of Observation

The comprehensive study of the traces of use-wear on obsidian and volcanic rocks resorted to criteria defined by prior research (Mansur-Franchomme, 1986; Cotterell & Kamminga, 1987; Hurcombe, 1992; González & Ibáñez, 1994; Rodríguez-Rodríguez, 1993, 1998; Kononenko, 2011; Clemente-Conte et al., 2015; Huidobro, 2018). A descriptive protocol was likewise designed to highlight the responses of each type of lithic raw material to identical tasks. As noted, the other volcanic rocks (basalt, trachyte and phonolite) of the Canary Islands are characterised by coarse surfaces due to the presence of varying degrees of phenocrysts (some can be classified as cryptocrystallines). Canarian obsidian is also characterised by a variety of surfaces, with that of Tenerife slightly coarser than that of Gran Canaria (Figs. 3 and 4). These different factors lead to uneven wear and development of the traces of use influencing on their distribution and development. This study resorted to the descriptive conventions of both glass and phenocrysts. Previous research has ratified these conventions in the case of other volcanic rocks (Bello-Alonso et al., 2020; Clemente-Conte, 1997; Clemente-Conte et al., 2015; Huidobro, 2018). This is a particularly important aspect when comparing the similarities and differences in the surfaces of the different raw materials when working identical plants in a similar way. It is nonetheless possible to define the following series of general characteristics for each of the two main lithic materials.

The traces of use-wear on the stone tools observed with a binocular magnifying glass (8–80x) are the following:

  1. a.

    Scars. Location (ventral or dorsal, unifacial, bifacial or alternate); distribution (continuous or discontinuous); disposition (isolated, in chain or overlapping); orientation (perpendicular or oblique); morphology (crescent, trapezoidal, semicircular or irregular); termination (feather, step, hinge or multiple) (Kamminga et al., 1979; González & Ibáñez, 1994; Huidobro, 2018).

  2. b.

    Rounding of the edge. Location, distribution and intensity (low, medium or high) (Huidobro, 2018).

  3. c.

    Abrasion. Location (edge, surface or both), distribution and intensity.

  4. d.

    Striae. Location, distribution, disposition (grouped or isolated); orientation (parallel, perpendicular, oblique or random); size (short 1 mm, medium 3 mm or large > 4 mm).

The traces of use-wear observed through a metallographic microscope (50-400X) are the following:

  1. a.

    Abrasion. Location, distribution, disposition (marginal, (patches); moderate (abrasion zones), orientation (parallel, perpendicular/oblique or indeterminate); intensity (low, medium, high) (Clemente-Conte et al., 2015).

  2. b.

    Striations. Location, distribution, disposition, orientation, depth (shallow or deep), length (short (< 2 μm), medium or long (> 2 μm); type of edge (straight or irregular); type of striations (rough-bottomed; sleek; flaked; intermittent; comet-shaped; fern-like) (Mansur-Franchomme, 1982; Hurcombe, 1992; Huidobro, 2018).

  3. c.

    Polish. Location, microtopographic distribution: poor development (smooth only along the edge, lancets or depressions), medium development (when the polish almost totally modifies the surface); high development (when the polished surface is almost totally modified); texture (very smooth, smooth or rough); morphology (smooth, bumpy or undulated). Variables such as the degree of the linkage of the polish (trame) were not considered due to the difficulty in detecting them on each of the two rock types.

Results

The results of the study are presented first according to the type of raw material (obsidian or other volcanic rocks) followed by the nature of the worked material (wood and non-wood plants) and the types of actions. The study considers that the physical and mechanical characteristics of each type of tool can react differently to the same type of action. These results focus on the more diagnostic criteria of the materials. The remaining attributes of use-wear are shown in the following spreadsheets (see Supplementary files 2 and 3). It is essential to bear in mind that the use-wear descriptions from woodworking were in its fresh and dry state and are highlighted only if they reveal differences.

Obsidian

Longitudinal Actions with Obsidian Tools

Wood

The experimental sawing of dry and fresh woody plants involved 13 obsidian tools. BMG observations reveal continuous, alternating and overlapping scars, disposed along the edge in a chain. Most are semicircular with feather terminations. The edge reveals abrasion with a discontinuous distribution. MM observations, in turn, reveal some bifacial, discontinuous and marginal traces along the edge, notably abrasion of low intensity restricted to the edge. The striations are also bifacial, parallel to the edge, and continuous in the case of working fresh wood but discontinuous when the wood is dry. The most common type reveals irregular edges and is rough-bottomed and fern-like (Figs. 6 and 7). The distribution of the polish linked to working dry wood is medium developed, appearing only along the ridges or on the protruding areas. The polish visible from working fresh wood, although not continuous, extends over a wide surface occupying areas adjacent to the most prominent relief. Its texture is smooth and the morphology is bumpy (Figs. 8 and 9).

Fig. 6
figure 6

Exp 132. Obsidian: sawing dry wood (Viburnum rugosum) for 20′ (200x): edge scarring, sleek and rough-bottomed striations, discontinuous smooth and bumpy polish mainly along the ridge

Fig. 7
figure 7

Exp 104. Obsidian: sawing dry wood (Ilex canariensis) for 60′ (200x): edge scarring, discontinuous abrasion, fern-like striations, discontinuous smooth and bumpy polish

Fig. 8
figure 8

Exp 105. Obsidian: sawing fresh wood (Ilex canariensis) for 20′ (200x): abrasion along the edge, intermittent, sleek and rough-bottomed striations, continuous smooth and bumpy polish

Fig. 9
figure 9

Exp. 105. Obsidian: sawing fresh wood (Ilex canariensis) for 20′ (200x): abrasion along the edge, sleek and fern-like striations, continuous smooth polish

Ripe Cereals

The harvesting of ripe cereals was carried out with 17 obsidian pieces. BMG analyses reveal bifacial, continuously distributed scars in chain and an overlapping disposition. Although most often semicircular, there are a few irregular cases with feather terminations. The rounding is bifacial and continuously distributed. Furthermore, there are cases marked by discontinuously distributed bifacial, dense and parallel striae along the edge. MM observations reveal discontinuously distributed bifacial areas of abrasion. Their disposition is moderate and characterised by a medium to high intensity. The striations are bifacial, continuously distributed and parallel to the edge, and characterised by straight and irregular edges, flaked and comet-shaped types. There arealso intermittent and comet-shaped cases. The polish is distributed throughout the surface’s microtopography adjacent to the edge, even slightly to the interior, and reveals a very smooth texture and flat morphology (Fig. 10).

Fig. 10
figure 10

Exp. 211. Obsidian: cereal harvesting (Hordeum vulgare) for 120′ (200x): continuous rounding along the edge, sleek, flaked and comet-shaped striations, continuous smooth and flat polish

Fresh Rushes

Nine obsidian tools served to reap green rushes. Their scars under BMG magnification are bifacial, distributed continuously in chain and an overlapping disposition. Their semicircular shape is marked mostly by step terminations and crescent morphology. The abrasion is on the edge and surface with discontinuous distribution. Also visible are bifacial and continuous isolated striae, oriented parallel to the edge. MM observations, in turn, reveal discontinuously distributed marginal bifacial abrasion of medium intensity. The striations are very abundant, bifacial and parallel. The most common have irregular edges and are rough-bottomed and sleek. The polish covers most of the edge, although it is denser along the protruding areas and in the interior of certain scars. Its texture is smooth and its morphology undulated (Fig. 11).

Fig. 11
figure 11

Exp. 202. Obsidian: reaping fresh rush (Juncus acutus) for 40 (200x): abrasion along the edge, rough-bottomed and sleek striations, continuous smooth and undulated polish

Fresh Palm

Working fresh palm leaves required sawing the base of each leaflet to separate it from the rachis. This involved four pieces of obsidian. Under BMG magnification, the scars are alternating with a discontinuous distribution and an isolated or overlapping disposition. Although most often crescent-shaped, they also appear at times semicircular with step terminations. The rounding is distributed discontinuously on the edge. Abrasion under the MM is bifacial, continuously distributed, arranged marginally and marked by a low intensity. Striae are abundant, bifacial and parallel to the edge and continuously distributed both far and near the edge. Although most often intermittent, they also reveal sleek with straight edges. The polish is bifacial, continuously distributed along the edge, practically modifying the obsidian's microtopography. Its texture is smooth and its morphology is bumpy (Figs. 12 and 13).

Fig. 12
figure 12

Exp. 280. Obsidian: sawing the petiole of fresh palm leaves (Phoenix canariensis) for 20′ (200x): continuous rounding along the edge, sleek and intermittent striations, continuous smooth and bumpy polish

Fig. 13
figure 13

Exp. 280. Obsidian: sawing the petiole of fresh palm leaves (Phoenix canariensis) for 20′ (200x): continuous rounding along the edge, sleek and intermittent striations, continuous smooth and bumpy polish

Transversal Actions with Obsidian Tools

Wood

Scraping both fresh and dry wood was undertaken with 14 obsidian flakes. The location and distribution of the traces when observed both by BMG and MM follow the angle of contact of the tool (positive, negative or perpendicular). Amongst the 14 pieces, 9 worked coupe positive. In this case, most of the traces are on the ventral face. The rest worked coupe negative and no traces were observed on the ventral face. The scars thus, on the whole, respond to the obsidian’s working angle. Most are unifacial, ventral or dorsal depending on the position of the attack face. They are distributed discontinuously along the edge and disposed in chain with semicircular and irregular morphology and feather or step terminations. The abrasion is continuous, on the edge and surface on the attack face. Abrasion under MM is bifacial with a marginal distribution and arranged discontinuously, although revealing a greater development along the face of contact. Its intensity is mainly medium. The location of the striations also depends on the conductive face, although their distribution in both dry and wet conditions is discontinuous. Although most commonly presenting a transversal orientation, there are also cases oblique with respect to the edge. The most representative type bears an irregular edge and is rough-bottomed, whereas a lesser number bear sleek striations. The polish is poorly distributed, appearing only on the upper areas of the microtopography. Its texture is smooth, and its morphology is slightly flattish and bumpy (Figs. 14 and 15).

Fig. 14
figure 14

Exp. 120. Obsidian: scraping dry wood (Morella faya) for 60′ (200x): abrasion along the edge, sleek and rough-bottomed striations, discontinuous smooth and flat polish

Fig. 15
figure 15

Exp. 120. Obsidian: scraping dry wood (Morella faya) for 60′ (200x): discontinuous abrasion along the edge, sleek striations, discontinuous smooth and bumpy polish

Remoistened Rushes

Scraping the interior of remoistened rush stems (subsequent to their splitting after drying) was carried out with three obsidian tools. All pieces have an acute angle. Therefore, the attack and contact faces were interchangeable. The BMG reveals a chain of discontinuously distributed bifacial scars. Although most are crescent-shaped, there are also semicircular cases with feather and step terminations. The abrasion is on the edge with a continuous distribution. Striae under MM magnification are bifacial with an irregular location and distribution. The most common type has straight edges and is sleek even if there are a few cases with irregular rough-bottomed. The polish is medium developed, especially along the upper areas of the microtopography such as the ridges. Its texture is very smooth and its morphology slightly flat (Figs. 16 and 17).

Fig. 16
figure 16

Exp. 292. Obsidian: scraping and splitting remoistened rush (Juncus acutus) for 10 (200x): continuous abrasion along the edge, sleek striations, discontinuous smooth and flat polish

Fig. 17
figure 17

Exp. 292. Obsidian: scraping and splitting remoistened rush (Juncus acutus) for 10 (200x): continuous abrasion along the edge, sleek striations, discontinuous smooth and flat polish

Splitting with Obsidian Tools

Wood

The experiments of splitting both fresh and dried wood were carried out with 11 obsidian tools. BMG observations reveal abundant bifacial overlapping scars distributed continuously along the edge. Most are crescent and semicircular morphology with hinge terminations. The abrasion is discontinuously distributed along the edge and the surface. Abrasion under MM is bifacial although it can reveal a greater development on one edge rather than on the other (depending on the working angle). The distribution is marginal and arranged discontinuously, although revealing a greater development along the face of contact. Its intensity is mainly medium. The striae are discontinuously distributed, bifacial with transversal and oblique orientations. Although most types of striations present irregular edges and are rough-bottomed, a few, by contrast, have straight edges and are sleek. No polish was observed (Figs. 18 and 19).

Fig. 18
figure 18

Exp. 44. Obsidian: splitting dry wood (Laurus novocanariensis) for 15′ (200x): discontinuous abrasion on the surface and rough-bottomed striations

Fig. 19
figure 19

Exp. 44. Obsidian: splitting dry wood (Laurus novocanariensis) for 22′ (200x): discontinuous abrasion along the edge and surface, sleek and rough-bottomed striations

Remoistened Rushes

The action of splitting remoistened rushes involved four obsidian tools. The scars visible through BMG magnification are bifacial, continuously distributed, overlapping and disposed in chain. Most are of irregularly morphology with step and hinge terminations. The abrasion is discontinuously distributed along the edge and the surface. MM views are more difficult to interpret as there is no evidence of clear use-wear. The scarce striae are bifacial, discontinuously distributed along the edge with both transversal and oblique orientations. Most have straight edges and are sleek. Otherwise, it is not possible to observe any polish (Fig. 20).

Fig. 20
figure 20

Exp. 297. Obsidian: splitting remoistened rushes (Juncus acutus) for 40′ (200x): scarring, discontinuous abrasion along the edge, sleek striations

Remoistened Palm

Splitting remoistened palm was carried out with eight obsidian fragments. The scars under the BMG are discontinuously distributed along the edge. Their disposition is in chain and they are most often of crescent morphology (in spite of some semicircular cases with feather terminations). The abrasion is discontinuously distributed along the edge. MM magnification reveals alternating discontinuously distributed striae very near the edge with both parallel and transversal orientations. They most commonly are straight edged and sleek. The polish, distributed alternately in the upper areas of the microtopography, is smooth with a bumpy morphology (Fig. 21).

Fig. 21
figure 21

Exp. 324. Obsidian: splitting remoistened palm leaves (Phoenix canariensis) for 20′ (200x): discontinuous abrasion along the edge, sleek striations, discontinuous smooth and bumpy polish

Other Volcanic Rocks

Longitudinal Actions with Other Volcanic Rock Tools

Wood

Experimental sawing of both fresh and dry wood involved 12 tools. These pieces under BMG magnification reveal continuously distributed, alternating scars, both in chain and overlapping (in the case of dry wood). They are mostly of crescent or trapezoidal morphology with feather and hinge terminations. The abrasion is discontinuously distributed along the edge and the surface. MM observations reveal discontinuously distributed bifacial, marginal abrasion. Striations are along the amorphous matrix along the edge with parallel orientations. The most common type has an irregular edge and is rough-bottomed (Figs. 22 and 23). The microtopography of the bifacial polish is smooth in texture and flat in morphology. It only occupies the upper areas of the amorphous matrix. Sleek striations and flat polished surfaces can be observed on the phenocrysts (Figs. 24 and 25).

Fig. 22
figure 22

Exp. 144. Basalt: sawing dry wood (Juniperus turbinata) for 25′ (200x): discontinuous abrasion on the surface, rough-bottomed striations, discontinuous smooth and flat polish on the amorphous matrix

Fig. 23
figure 23

Exp. 144. Basalt: sawing dry wood (Juniperus turbinata) for 25′ (200x): discontinuous abrasion on the surface and discontinuous smooth and flat polish on the amorphous matrix

Fig. 24
figure 24

Exp. 155. Basalt: sawing fresh wood (Dracaena draco) for 15′ (200x): discontinuous abrasion on the surface and discontinuous smooth and flat polish on both the amorphous matrix and the phenocrysts

Fig. 25
figure 25

Exp. 155. Basalt: sawing fresh wood (Dracaena draco) for 15′ (200x): discontinuous abrasion on the surface, sleek striations and discontinuous smooth and flat polish on the phenocrysts

Ripe Cereals

Harvesting with the other volcanic rocks involved 18 tools. Under the BMG the scars are bifacial, distributed continuously and disposed in chain. Although most are crescent-shaped, there are also cases of semicircular morphology with feather terminations. The rounding is bifacial and continuous along the edge. Striae, at times visible, are bifacial, continuously distributed and arranged following a dense parallel orientation. Under the MM the abundant striae are also bifacial, continuously distributed and oriented parallel to the edge. Most are characterised by straight edges and sleek or irregular rough-bottoms. There are likewise some intermittent cases. Polish is visible on the microtopography along most of the edge. Its distribution, when the experimentation endured less than 60 min, is partially developed along both the low and upper areas. In the experiments exceeding 180 min it is greatly distributed along almost the entire edge. Its texture is very smooth with a flat morphology and a few micropits. Striae and flat polished surfaces can be observed along most of the phenocrysts of the tools having worked for several hours (Fig. 26).

Fig. 26
figure 26

Exp. 354. Trachyte: harvesting ripe cereals (Hordeum vulgare) for 180′ (200x): continuous rounding along the edge, sleek and rough-bottomed striations, continuous smooth and flat polish on the amorphous matrix and on the phenocrysts

Fresh Rushes

Nine volcanic rock tools served to reap green rushes. BMG observations reveal alternating scars, mostly of crescent morphology, distributed continuously and disposed either in chain or overlapping. Abrasion is continuously distributed along the edge and the surfaces. Striae are visible in certain cases. They are bifacial, dense and oriented parallel to the edge. MM magnification reveals a moderate continuously distributed bifacial abrasion. Striations are very abundant along the amorphous matrix. They are bifacial and continuously distributed parallel to the edge. Most reveal an irregular edge and are rough-bottomed or intermittent (although there are some cases with straight edges and sleek). The polish is bifacial and extensive, modifying part of the amorphous matrix. It is located on the edge and on the surfaces. Its smooth texture and morphology are bumpy marked by micropits and abrasion. The phenocrysts of these tools only feature polish (Fig. 27).

Fig. 27
figure 27

Exp. 153. Basalt: harvesting fresh rushes (Juncus acutus) for 60′ (200x): discontinuous abrasion along the edge, rough-bottomed and intermittent striations, discontinuous smooth and bumpy polish

Fresh Palm

The sawing of fresh palm leaflets was undertaken by four tools. BMG observations evidence continuously distributed, alternating scars, most often of crescent morphology, disposed either in chain or isolated in other areas of the edge. Abrasion is discontinuously distribution along the edge and surface. MM observations, in turn, reveal a marginal discontinuous and alternating abrasion. The striae are bifacial, located along the amorphous matrix. Their distribution is discontinuous and, although mostly parallel, can be oblique. The most common type has an irregular edge and are rough-bottomed and sleek. Their phenocrysts are marked by very few striae bearing the same characteristics. The polish is bifacial and very slightly distributed amongst certain areas of the microtopography. It is usually of smooth texture, of bumpy morphology and devoid of micropits. It is very difficult to observe traces on the phenocrysts as their surfaces are completely worn by abrasion (Figs. 28 and 29).

Fig. 28
figure 28

Exp. 145. Basalt: sawing fresh palm (Phoenix canariensis) for 20′, discontinuous abrasion on the surface, sleek striations, discontinuous smooth and bumpy polish

Fig. 29
figure 29

Exp. 145. Basalt: sawing fresh palm (Phoenix canariensis) for 20′, discontinuous abrasion on the surface and discontinuous smooth and bumpy polish

Transversal and Chopping Actions with Volcanic Rock Tools

Wood

The scrapping of both dry and fresh wood was undertaken with 12 pieces. Here, the BMG observations evidence scars in a variety of locations depending on the working angle and the edge of the tool. Although they are at times bifacial, in other cases with positive angles they are on the face of attack. Most are distributed discontinuously and are overlapping. Although predominantly semicircular with step termination, a lesser number are crescent-shaped. Abrasion is distributed discontinuously along the edge and surfaces. MM observations indicate that abrasion is bifacial, marginal and discontinuously distributed. Striae are scarce in both states of the wood. Their location, away from the edge, depends on the face of attack. They are distributed discontinuously with a varied orientation (predominantly transversal). The most common bear irregular edges and have rough-bottom. Striae on phenocrysts are only visible at 400x. The orientation varies depending on the working angle. The most common have straight edges and are sleek. The polish is poorly developed and only found along the upper areas of the microtopography of the amorphous matrix. Their texture is smooth with a flat morphology in the case of fresh wood and only smooth when the wood is dry. Areas of polish are only observed on the edges of the phenocrysts with bumpy morphology (Fig. 30).

Fig. 30
figure 30

Exp. 149. Basalt: scraping dry wood (Juniperus turbinata) for 45′, discontinuous striations and discontinuous smooth and bumpy polish on the crystal

The chopping of fresh wood was carried out with a bifacial knapped pebble and two picks. The traces on the picks are located on the active apexes and on the pebble along the entire knapped ridge. BMG observations of the knapped pebble reveal bifacial scars mostly along the upper face. Their distribution along the edge is discontinuous, disposed in chain and slightly overlapping. Although most often semicircular, there are also cases of trapezoidal and crescent morphology with feather and hinge terminations. A slight discontinuous abrasion is visible along the protruding areas. In the case of the picks, the traces visible by BMG, mainly bifacial scars, are specific to the apexes. Their distribution is continuous, clustered and overlapping. Their morphology is irregular, with step termination. The abrasion is distributed discontinuously along the edge and the surfaces. MM reveals bifacial abrasion that is marginal and discontinuous. Randomly oriented striae are visible along the surface of some restricted polished areas far from the edge. The most common bear irregular edges and rough-bottoms. Striae are not visible on the phenocrysts. Polish is poorly developed in patches. Its texture is smooth with a bumpy morphology (Fig. 31).

Fig. 31
figure 31

Exp. 375. Basalt: chopping fresh wood (Pinus canariensis) for 120′, discontinuous abrasion on surface, rough-bottomed striations, discontinuous smooth and bumpy polish

Remoistened Rushes

The inner part of the remoistened rushes was scraped with four volcanic rock pieces. The scars deriving from these actions under BMG magnification are bifacial, predominantly crescent-shaped, discontinuously distributed, overlapping and isolated in certain areas. Abrasion is discontinuously distributed along the edge. The development of use-wear as viewed under the MM is difficult to ascertain. Few continuously distributed, bifacial, highly variable oriented striae are far from the edge. The most common striae have irregular edges and rough-bottoms. Otherwise, it is not possible to observe any polish (Fig. 32).

Fig. 32
figure 32

Exp. 290. Trachyte: scraping remoistened rushes (Juncus acutus) for 20′, rough-bottomed striations on the phenocrysts

Splitting and Drilling with Volcanic Rock Tools

Wood

The single drilling experiment carried out with a basalt tool on dry wood yielded highly visible wear. BMG observations reveal no scars as the apex is completely rounded. There are striae, predominantly transversal, densely arranged over the entire surface. The striae under MM magnification are highly visible on the amorphous matrix. They are arranged along the distal area, with a continuous distribution and random orientation. The most common type has irregular edges, sleek with rough-bottoms. Polish, visible on the surface of the amorphous matrix, is developed on the upper areas of the microtopography. Its texture is smooth and its morphology slightly bumpy. No traces are visible on the phenocrysts (Fig. 33).

Fig. 33
figure 33

Exp. 160. Basalt: drilling dry wood (Ilex canariensis) for 15′, discontinuous sleek striations, discontinuous smooth and bumpy polish of the microtopography

The wedging of fresh wood was carried out with two volcanic rock flakes. Both bear complete bifacial scars disposed in chain with very irregular shapes and step terminations. The numerous chipping scars render it impossible to view the other use-wear, which explains why we did not include photographs.

Remoistened Rushes

Splitting rushes were undertaken with five volcanic rocks. Under BMG, they display discontinuously distributed and overlapping bifacial scars. Although most are crescent-shaped, some are semicircular with feather terminations. MM observations do not evidence any use-wear.

Remoistened Palm

The splitting of remoistened palm leaves involved 10 volcanic rocks. BMG reveals continuously distributed bifacial scars, most often crescent shaped, and disposed in chain. Abrasion is discontinuously distributed along the edge. MM observations reveal bifacial striae near the edge which are distributed discontinuously with random orientation. The predominant type has an irregular edge and is rough-bottomed. The bifacial polish is poorly distributed which explains why it can only be observed near the edge along the upper areas of the microtopography. Its texture is smooth and its morphology is bumpy. The pattern of polish on the phenocrysts is very similar to that visible on the amorphous matrix (Fig. 34).

Fig. 34
figure 34

Exp. 306. Trachyte: splitting remoistened palm leaves (Phoenix canariensis) for 20′, discontinuous striations and bumpy polish on the crystals

Discussion

This article is focused on wear visible on two main categories of volcanic effusive elements: obsidians and other volcanic rocks. Within each of the two categories it is possible to observe variability of use-wear. Firstly, there is the relationship between the topography of the knapped surfaces, the distribution of features and the time of use for the action. As mentioned previously, the surfaces of obsidians from Tenerife and Gran Canaria vary from the very fine to slightly rough. For example, the phonolite obsidian from Tenerife is the roughest which in terms of traces of identical actions yield less rounding, less scarring and less striae. Although their polish is usually more irregularly distributed, their characteristics resemble those of the phonolite and trachyte obsidians from Gran Canaria. The experimentation with obsidian from Tenerife was in general more resistant, even in the case of woodworking. It has been stressed that obsidian has physical properties which differ from other fine knapped fracture rocks such as flint (Astruc, 2022; Hurcombe, 1992; Kononenko, 2011). This is a key notion when studying use-wear. Obsidian is in fact more brittle that flint and its edges are more easily abraded. It is also particularly prone to surface alteration by chemical and, to a lesser extent, mechanical transformation (Hurcombe, 1992). As wear has a greater effect on the upper features of the microtopography, one must keep in mind that the rougher the surface, the more discontinuous the attributes. Polish is less visible as the natural surface of obsidian reflects light. In addition, attributes such as extension or morphology are less accessible and others as the degree of linkage (trame) of the polish is not recognisable.

In the case of the other volcanic rocks, the heterogeneous composition of their amorphous matrix and the variability of their phenocrysts in terms of size and quantity renders is difficult to observe the use-wear. The greater amount of crystals the less chance of identifying traces such as striae or polish even at high magnification (400X). The development of these traces in the cases of rocks with an amorphous matrix is very uneven. Abrasion (and sometimes polish) together with striations can only be observed on the upper areas of the surfaces. As in other amorphous crystalline rocks, the difficulty in locating and describing them complicates observations (Clemente-Conte, 1997; Clemente-Conte et al., 2015; Huidobro, 2018). It is key to take into account that the completely irregular microtopography of this type of raw material yielding discontinuous wear, segmented striations and polish in patches or spots, reduces visibility and the possibility of diagnosis.

This discussion will thus delve into how to distinguish the different types of worked plants according associations of attributes. To do this, one must first describe those specific to obsidian before turning to those of the remaining volcanic rock types.

Obsidian Experimental Tools

Contact of this rock with woody and siliceous plants yields different types of wear and different distribution. Woodworking creates abrasion along the edge, fern-like striations and discontinuously distributed polish of bumpy morphology. Contact with siliceous plants leads to rounding, sleek striations, and a polish with smooth morphology. In spite of this, there are certain minute differences (that will be explained later) that allow to identify certain plants. It is important to mention that woody and non-woody plants can be differentiated by means of associations of attributes. Moreover, their distribution is also influenced by the condition (fresh, dry or intentionally re-moistened) of the plant and its hardness. Hardness mainly influences not only the development of striae (distribution and morphology) but also other attributes such as polish (Table 2). In general, it is possible to distinguish between hard and soft plant material, but a more detailed interpretation such as whether it is in a dry or fresh state is more difficult to advance. The mode of action in general also yields different types of traces such as striations or polish. It is possible to observe, in the case of woody plants, a greater number of longitudinal than transversal actions. This is potentially due to the working angle as in scraping the angle of the cutting edge and the angle of work significantly influence the distribution of the traces (notably striae and polish). Siliceous plants also yield more attributes in the case of longitudinal tasks such as harvesting or cutting. Another influence in these cases is the silica content and the time of use.

Table 2 Summary of the traces of use-wear visible on obsidian experimental tools

Woodworking with obsidian yields characteristic use-wear. Scarring and abrasion are concentrated along the edge. Striations can be near the edge attached to the areas of abrasion, on the surface and inside certain scars. They are mostly fern-like and rough-bottomed. Striations are linked to polished areas. Polish is discontinuously distributed along the microtopography of the surface and reveals a smooth texture and bumpy morphology. Diagnosing woodworking is possible subsequent to at least 20 min of work. Distinguishing the state of the wood (hardness and degree of humidity) was not possible in the current experimental programme due to the fact that scarring cannot always be linked to these conditions, and striae and polish are similar in all cases.

Cereal harvesting yields specific types of use-wear on obsidian. Soft plants rich in silica yield a high degree of rounding of the edge. Striations are located near the edge, on whole surfaces and inside scars. The most common are comet-shaped or sleek. Polish is invasive along the edge and adjacent surfaces and completely modifies the microtopography of the surface of the tool. It has a smooth texture and flat morphology. The invasiveness of the wear is a criterion serving to identify cereals as opposed to other silica-rich plants. Their interpretation is possible even when the experimentation was relatively short (e.g. 10 min).

Rushes although siliceous are harder than cereals yielding more developed scarring and abrasion on the edge and surface. Striations are limited to the surfaces situated near the edge. The rough-bottomed type is the most common. Polish also develops along the edge with a smooth texture and an undulated morphology. It is not possible to distinguish this plant when in a dry state as these traces do not develop before 60 min of use. Hence, working time greatly influences on the level of interpretation.

Palms are also siliceous plants whose differences in hardness can be observed when processing the stem and the leaves. The clearest traces on the obsidian tools are those stemming from the cutting the stem. Scarring and rounding are also common along the edge. Striations are near the edge and on the surface, although not often inside the scars. The most common are intermittent. Polish near the edge is not as invasive as in the case of cereals. It has a smooth texture and bumpy morphology. These types of striations and the polish located the near the edge are characteristic of the cutting of palm stems. These criteria can be identified subsequent to working lasting at least 20 min. Splitting palm leaves, in turn, does not yield developed attributes meaning it is much more difficult to identify the matter of contact. It is also not possible to differentiate whether the plant was in a fresh or remoistened state.

A broad comparative framework is available to contrast the findings of the experimentation we carried out with obsidian tools. Yet it is necessary to note that it is not possible in the case of woodworking to offer specific comparisons because most obsidian study variables correspond to either hard or soft states whilst in our case the observations correspond to fresh/dry states. Hence the state of the wood from different contexts is an important factor to take into consideration. In addition, this is the first experiment conducted on fresh palm and remoistened palm and rushes.

Similar patterns to those observed here have been described in the case of woodworking. For example, Rodríguez-Rodríguez (1998) observed scarring while Kononenko (2011) has cited rough-bottomed striations. Huidobro (2018) also reported a poorly developed distribution of polish and Walton (2019) has described what is labelled as bumpy morphology. However, there are certain discrepancies such as the observation of rounding by Hurcombe (1992) and the presence of intermittent striae by Kononenko (2011) and Huidobro (2018). Although these characteristics were not identified in our experimentation, we did observe abrasion along the edge and fern-like striations. These may be due to the difference of the species of wood and their humidity or degree of hardness.

The traces stemming from harvesting cereal is also the subject of previous research. Rodriguez-Rodriguez (1998) and Astruc (2012) observed moderate rounding. Hurcombe (1992) identified long striae both near and far from the edge, most often sleek, rough-bottomed or intermittent. Smooth and flat-shaped polish, in turn, is regular and very invasive along the edge, attributes likewise have observed in the experimentation. Moreover, cereals with a high silica content will yield similarly distributed attributes on all types of obsidian.

The results of our experiments in harvesting green rushes coincided with the microscopic observations by Hurcombe (1992) in that they reveal well-developed striae both along the tool's edge and the interior of the scars that are predominately rough-bottomed (in spite of intermittent or comet-shaped cases). Differences with this earlier study relate to the distribution of the regular polish and the bumpy morphology which in this case are undulated. It is possible that this may be due to the characteristics of the knapped surface of the obsidian.

Other Experimental Volcanic Rock Tools

Basalt, trachyte and phonolite, the other volcanic rocks serving for this experimentation, reveal use-wear features subsequent to working wood and other siliceous plants that differ from those on the obsidian tools. These differences affect in particular the amorphous matrix and the phenocrysts. Woodworking yields surface abrasion, rough-bottomed striations and discontinuously distributed polish of flat morphology. Contact with siliceous plants, in turn, gives rise to rounding or abrasion, depending on the plant. Sleek or intermittent striations and polish with bumpy morphology are more common. As these traces require more time to appear on the surfaces of these rocks than obsidian, their diagnosis is more arduous (Table 3). It must also be stressed that these rocks yield a lesser variety of striations and polish than obsidian which likewise renders it more difficult to identify the worked plant. Therefore, identifying the type of plant is complicated in the case of scraping, even amongst those that are silica-rich. Although the state of the plants and their hardness also have an influence on these rocks, these conditions are not easy to discern.

Table 3 Summary of the traces of use-wear visible on experimental tools of other volcanic rock

Woodworking yields scarring along the edge. Abrasion, for example, is visible along both the edge and the amorphous matrix surface. Striations, most commonly rough-bottomed, are near the edge, linked to the polished areas. Polish is discontinuously distributed through the microtopography of the surface. Its texture is smooth and its morphology flat. A diagnosis of wood is possible only for the most part after a lapse of at least 20 min. However, distinguishing the state of the wood (hardness and humidity) was not possible in our experimental programme.

Cereal harvesting also has specific traces due to their high silica content. The actions yield rounded edges, and striations are visible over all the polished surface. The most common are sleek or rough-bottomed. Polish appears close to the edge and adjacent surfaces. It modifies part of the microtopography (amorphous matrix) yielding a smooth texture and a flat morphology. In this case, interpretation is only possible after 120 min of use, a timeframe which is much longer time than the case of obsidian tools.

Rushes, also siliceous plants, are harder than cereals which produces more developed scarring. Abrasion appears along the edge and the surface of the amorphous matrix. Striations, rough-bottomed and intermittent, are limited to surfaces near the edge. Polish marked by a smooth texture and bumpy morphology develops near the edge. These characteristics are common to green and remoistened rushes when worked with a transversal motion. These traces only appear after 60 min, as in the case of obsidian.

It is possible, in the case of palm, to observe differences in the processing of stem and leaves. However, the most readable traces are those linked to cutting green stems. Scars appear along the edge, and abrasion is on the edge and surface. Striae, usually rough-bottomed, are located near the edge of the amorphous matrix. Polish, in turn, develops far from the edge and is less invasive than in the case of cereals. Its texture is smooth and a bumpy morphology. These traces only appear on the pieces after at least 20 min of use.

There are relatively few use-wear studies related to other volcanic rocks (compared to research on obsidians). Worth highlighting is the research on woodworking by Clemente-Conte (1994) and Bello-Alonso et al. (2020) that also observed scarring as well as short striations on or near the polished areas. Huidobro (2018) also reported poorly developed polish whilst Richards (1988) described smooth pitted polish in patches. Plisson (1985) also cited short striations and marginal polish distributed in spots. Although our observations do at time concur with their findings, our experiments did yield different results. For example, we did not observe rounding along the edge (Richards, 1988) or a polish of bumpy morphology (Huidobro, 2018) but abrasion along the edge and surface, and a flat polish. It remains to be seen if these differences are due to the particularities of the raw material or the nature of wood.

The harvesting of cereals and rushes has been described by Richards (1988) and Clemente-Conte (1994). The descriptions of Richards (1988) coincide with our observation in that rounding and striations are associated with a smooth polish. Clemente-Conte (1994) did not discriminate between rushes and cereals. His study notes intermittent scars and striations associated with a polish of irregular morphology, attributes that coincide with our experimentation with rushes. They can be distinguished traces linked to mature cereals by the presence of smooth grooves and a polished of flat morphology. These differences could be due to the characteristics of the volcanic raw material and the composition of their phenocrysts.

Conclusion

This systematic experimental study is an attempt to interpret the use-wear stemming from working plants with stone tools from the Indigenous Period of the Canary Islands. Although the characteristics of the raw materials serving to manufacture volcanic tools may differ slightly from site to site, there remain, as in the case of other lithic raw materials, enough similarities to compare the use-wear produced by the different contact materials. Because this work is experimental, some results may not match those of the archaeological record. This is not only due to the great variability of working conditions but also to post-depositional alterations. One of the issues of this study has been the reproduction of certain craft techniques that to date have not received sufficient attention. We have identified several problems linked to these tasks. It was not simple, for example, to discern use-wear from working fresh or remoistened plants. Furthermore, we experimented with two types of volcanic rocks yielding very different types of use-wear, notably glassy obsidian, where the traces are easier to observe, and rough volcanic rocks with phenocrysts where observations are arduous. In any case, the patterns of use-wear are essential to identify not only each type of worked plant but the motion or gesture used. This study has attempted to offer an overview of the different diagnostic features, as well as highlight the limits imposed by each type of volcanic material. This study thus offers new data as to the treatment of different plant fibres such as rushes and palms and in particular those related to working plants with coarse-grained volcanic rocks.