Identifying archaeological parenchyma in three dimensions: Diagnostic assessment of five important food plant species in the Indo‐Pacific region

Archaeobotanical evidence for the exploitation of vegetatively propagated underground storage organs (USOs) in tropical regions Australia, Southeast Asia and Pacific is currently limited. Although there have been several key studies archaeological parenchyma published past two decades, systematic application identification methods crop species utilising charred, desiccated or waterlogged remains parenchymatous tissue not undertaken on a regular basis. Here, microCT imaging technology used to compile three-dimensional virtual reference collection tissues five USO known extensively cultivated by people these regions. The are Dioscorea alata, esculenta, Colocasia Alocasia macrorrhiza Ipomoea batatas. These samples illustrate character virtual, derived collection, they also capture inter-species differentiation intra-species morphological variation characteristic many tuberous root crops. Les indices archéobotaniques témoignant de l'exploitation d'organes réserve à propagation végétative dans les régions tropicales d'Australie, d'Asie du Sud-Est et Pacifique restent présent très limités. Malgré publications d’études sur parenchymes archéologiques au cours des deux dernières décennies, l'application systématique méthodes d'identification espèces cultivées est rarement suivie, qu'il s'agisse restes calcinés, desséchés ou gorgés d'eau tissus parenchymateux. Nous utilisons ici la technologie d'imagerie par microtomographie pour produire une référence virtuelle en trois dimensions cinq abondamment ces régions: Ces échantillons illustrent obtenue microtomographie, présentent différences inter-spécifiques variations morphologiques intra-spécifiques caractéristiques plusieurs tubercules. Parenchymatous long history humans dating back at least 170000 years ago (Wadley al., 2020) important staple crops billions globally today (FAO, 2021). include anatomically distinct bulbs, corms, rhizomes, roots tubers (Denham 2020; Hather, 2000). Within Indo-Pacific region, historical importance carbohydrate-rich plants has demonstrated through archaeobotanical significant sites such as Niah Cave (Barton 2016; Barton & Paz, 2007), Kuk Swamp 2003; Lewis Pritchard al. 2018) Madjedbebe (Clarkson 2017; Florin 2020). Comparable significance USOs diets documented Europe (e.g. Kubiak-Martens, 2002, 2006; Mason 2000), mainland Castillo 2018; Denham Americas Ugent 1981, 1982, 1984, 1986, Peterson, 1988). USO-bearing can be studied varying degrees using starch granules (Torrence Barton, 2006) phytoliths (Piperno, 2006), still an under-utilised technique (although exceptions 1991, 1994, 2000; Kubiak Martens, Oliveira, 2008, 2012; 2001, 2005; Ussher, 2015). difficulty identifying asexually plant sites, particularly regions, largely due forms which preserved. do ordinarily, especially under cultivation, produce taphonomically-robust macroscopic pollen, seeds, spikelet bases, charcoal nutshells therefore, it extraction analysis alternate grains, that identified contexts 2009, Hather 1994; Pearsall, While phytolith analyses become more common over last from remain comparatively rare. plastic often homogenous nature parenchyma, well fragmentary most material hinder taxonomic genus level, discrimination wild phenotypes. Consequently, study these, often, remained relatively limited comparison other macroremains 2005). Without knowledge how were managed exploited we only partial view likely complex mixed subsistence strategies employed different groups people, Pacific. Paz (1999) lamented persistent lack histories some world's cultivars esculenta spp. He highlighted inferred linguistics (Blust, 1976, 1995) rather than direct evidence. In recent genetic provided information regarding natural human-mediated (Chaïr Gichuki Lebot, 1999; Lebot 1998; Malapa Roullier, Kambouo, 2013; Scarcelli 2019). However, hypotheses scenarios domestication dispersal genetics await confirmation. this paper, make initial step confront research gap region. Over term, paper's aim encourage practice collecting analysing parenchyma. As starting point, though, present first steps compilation development comprehensive database archaeologically-significant species, will available on-line open access near future. Here criteria cultivar presented order demonstrate: substantively, diagnostic features enable species; and, methodologically, advantages benefits applying Parenchyma cells undifferentiated, ground occur all comprise majority total mass. They continuous mass up internal bulk leaves, fruits, seeds flowers. living, meristematic vary widely morphology, but generally thin, primary cell walls perform number functions dependent location within plant, photosynthesis, storage, transport, secretion repair. plants, large numbers store abundant quantities sugars, leading formation fleshy bulbous organs, provide food source herbivores omnivores, including humans. terms bulbs Allium spp.), corms esculenta), rhizomes Zingiber officinale), (Ipomoea batatas) (Dioscorea spp.). interchangeably, terminology actually specific represents organography. avoid confusion, collectively referred USOs, although sometimes aerial bulbils (i.e. bulbifera) caudex base stem macrorrhiza, Cycas spp., Zamia Jon discussed depth non-standardised fields both botany archaeobotany (2000: 5–6). For field archaeobotany, concluded anatomical was little except determination major developmental functional translating differences, namely, whether produced tissues. components cellular arrangements structurally morphologically distinctive other, then broken down into smaller categories specificity. literature refers “archaeological parenchyma”, note alone, cases, cannot determine differentiation. cells, archaeologically charred samples, fairly indistinct greatly shape size, throughout same organ, therefore their own distinguish even family (Hather, 2000: 73). Therefore, involves multiple characteristics across range scales robust characterisation discrimination. Diagnostic characteristics, (2000), arrangement vascular bundles; surface morphology; patterns decay/preservation; types, schlerenchyma; distribution abundance idioblasts, containing calcium oxalate crystals (see Table 1). ancient acknowledged since late nineteenth century (de Candolle, 1886). Historically, archaeologists extracting contexts. Since made non-lignified texture preserve when deposited fresh aerobic conditions. Like macrobotanical materials, tend desiccated, waterlogged. Furthermore, usually small pieces activities peeling, cooking incidental discard. A handful Canna edulis, Manihot Solanum tuberosum, dry, conditions South American 1950s, 1960s 1970s Engel 1970; Towle, 1961). Subsequently, (1981, 1988) batatas edulis All because found nearly whole dried tuber fragments enabling macro level. SEM batatas, Ullucus tuberosus Tropaeolum tuberosum Peru (Martins, 1976) sp. Brazil (Scheel-Ybert, 2001). It until 1980s undertook task developing method based micro-level structures, skin features, specialised size patterning charring damage, sort possibly identify Kirch, 1991). Hather's process involved extensive modern historically he subjected firing replicate effects visible His covered continents, with focus European, Asian plants. (outlined detail 2000) follows stages takes consideration wide variety accurate highest confidence levels. stressed no single could fragment combination must taken conservative taxonomy (Table species-level would possible state preservation, overlap morphologies, certain size. adopted part suite his unpublished PhD thesis (2001), sought cover utilisation practices Wallacean during twenty-first century. Ladefoged (2005) plant's introduction Hawaii. (2007) confidently alata cf. hispida Cave, Sarawak, dated c.40000 ago, aroids 23850–23020 calBP (also see Reynolds 2013). limitations current however, evident 17 42 Ussher (2015) created region her agricultural Tonga, species-wise overlaps extends upon collections compiled Hather. Archaeological domestic Papua New Guinea (Lewis 2016) Northern Australia early c.65000 accepted, perceived too labour-intensive standard analytical technique. applied attributed reasons, including: need exhaustive tissues; detailed species-specific structures; low-rates without our understanding use necessarily biased towards fruits nuts. MicroCT scanning taking series 2D X-ray images 360° around sample digital reconstruction those slices create 3D representation attenuation. X-rays penetrative waves, non-destructively external details imaged light electron microscopy. resultant datasets digitally segmented, rendered manipulated commercially open-source software packages, each visualisation tools, automated segmentation filters. programs slicing viewing infinite angles orientations meaning coverage scanned much greater microscopy, slice. commonly analyse morphologies palaeopathologies human Lynnerup 1997; Zollikofer 1998), contents sealed objects, jars, urns scrolls (Allegra Harvig Minozzi 2010). types steadily increased previous date years. Murphy Fuller (2017) Zong synchrotron changes associated legumes, horse gram (Macrotyloma uniflorum) soybean (Glycine max) respectively. Barron (2018) developed protocol segment out discriminate domesticated organic inclusions pottery sherds. investigation status rice (Oryza spp.) bases sherds Vietnam (Barron 2017) Borneo (Barron, Datan, 2020), contribute sorghum (Sorghum bicolor) Sudan Fuller, pearl millet (Pennisetum glaucum) Mali (Fuller Whitau (2016) scans Australian wooden artefacts image structure identification. To date, visualise taxonomically able sweet potato batatas), sugarcane (Saccharum officinarum), recovered Guinea. easily accessible needed routine point building collection. broadly: greater/water yam alata), lesser taro (Colocasia esculenta) alocasia (Alocasia macrorrhiza). volume data investigate interspecies intraspecies variation, enables specimens scan, individual sample. Methodologically, should protocols collections, relied and/or viewed reflected microscopy (RLM) (SEM) (discussed 2018). chosen factors. (Li, Yen, 1973). eventually term cultivation practices, domesticatory relationships derive three families (Araceae, Dioscoreaceae Convolvulaceae) genera Alocasia) (Araceae) included demonstrate inter-genera Two D. tease intra-genera differences. just its crop; example dicot compare four monocots. Additionally, representing plant. case originating consumed (Esau, 1977; Onwueme, 1978), whereas (Lebot, Plucknett, 1984; Strauss, 1983). Yams (Dioscoreceae), possess making them traditional examples either (Onwueme, 1978). expected broad variations. genetically diverse highly polymorphic produces enlarged, starchy diet half billion 2009). Modern C. geographical spanning India, Asia, China, Melanesia northern Australia. unclear if natural, result translocation. Matthews suggests confined Northeast India may initially evolved (Matthews, 2014), dispersed before became spread (Ahmed Taro, alongside yams, among earliest wet tropics (Lebot 1999). At Guinea, grains extracted stone tools residues show processing site 10000 BP integrating 6950–6440 (Fullagar 2006). uncertainty about times locations communities Aradhya, 1991; Li, Matthews, 1990; 1973, 1995; Yen Wheeler, 1968). Plausibly, engaged various relationship populations cultiwild (namely, wild-types) proto-domesticated forms, few diploid triploid Denham, 2018, Prior European exploration extended globe Japan north Zealand south, Southern west Polynesia east (Leon, 2014; Grimaldi 2022). succulent perennial herb thickened, woody-like stems. thickened stems peeled being baked boiled. Very consumption rarely definitively Loy (1992) crystal raphides thought Alocasia-type Solomon Islands 28000 ago. Starch longiloba Pleistocene-aged similar taro, distributed Pacific; consequently, local domestications Island 2011) coastal inhabitants Cultivated A. grown intensively Pacific, parts America Africa annual yams. considered traits scale buds (“eyes”) terminal buds. yams millions countries Africa, undergoing independent Fullagar (2006) c.10000 sp.). globally, wild-type, does ordinarily sexually reproduce 2009; Sharif suggested true cultigen, bred selection precursor (Hahn, related nummalaria transversa, form group Wallace line (Malapa after arrival Sahul Plate 60000 present-day 2009), centre diversity Martin Rhodes, 1977). along probable Madai-1 Sabah, ca.2200–1500 (Paz, Claims alata-type pre-30000 year-old (Summerhayes 2010) poss. context 2013), Holocene rockshelter Philippines (Pawlik tentative basis identifications. native Mainland outside (Oliveira Purseglove 1972). date. Horrocks Nunn deposits Fiji 3000–2500 calBP. hundred highlands (Bayliss-Smith 2017: fig. 16.13), subsequently batatas; vine-like trailing twining system. grow nodes cutting system producing between 5 10 adventitious roots. world (Roullier, Benoit, 1974). Its relatives Yucatan Peninsula Mexico Orinoco River Venezuela (Austin, 1977, 1987). Central seen likeliest (Gichuki 2003). Chilca Canyon c.8000 uncalBP (Engel, Woolfe, 1992). (1981) similarly Casma Valley 3735 3070 uncalBP. food, reached (contrast Muñoz-Rodríguez 2018 already Cook's voyages pre-European age Eastern (Hather Rosendahl 1971). finds years, suggesting rapid yet adoption 2022; Multiple selected stored Institute Archaeology, UCL London. Each includes one had muffle furnace 250 400°. Charring designed transformations analogous preserved pre-charring state, geographic origin listed 2. Subsamples transported ANU CTLab. <9 54 3.9 21.8 <7 51 3.1 20.6 47 2.8 18.9 56 22.5 3.8 19.0 USO, largest fragment, scanning. placed aluminium tube stabilised packing foam. Heliscan high-resolution optimised space-filling trajectory (Kingston yield sharp (Latham 2008). Due low-attenuating 0.5 mm filter energy 85 kV 60 mA durations 13 hours 15 diameter, resulting voxel (digital measure per pi