Phytoliths, parasites, ﬁ bers, and feathers from dental calculus and sediment from Iron Age Luistari cemetery, Finland Quaternary Science

Our understanding of subsistence strategies, resources and lifeways of Finnish Iron Age populations remains incomplete despite archaeological, osteological, macrobotanical, and palynological investigations. This is due in part to poor preservation of organic macroremains in the acidic boreal sed- iments. To address this problem, here we present the ﬁ rst data from microscopic remains preserved in prehistoric dental calculus from Finland. We extracted and analysed both plant and animal microremains from human calculus and burial site sediment samples, originating from Luistari cemetery in southwestern Finland (samples from c. 600 e 1200 calAD). We recovered phytoliths, parasites, ﬁ bers and feathers. While in Finland few previous archaeological studies have investigated phytoliths, our study con ﬁ rms the importance of these microremains for interpretating dietary patterns. It is also the ﬁ rst time that intestinal parasites have been reported in Finland. Our study demonstrates that, especially when working with acidic sediments typical for boreal en- vironments, microremain studies can considerably increase the information value of archaeological samples, and that dental calculus and phytolith analysis are important new methods in the research of prehistorical lifestyles. This combined microremain analysis should be more broadly applied in contexts where other dietary records do not remain. © 2019


Introduction
In boreal environments, the archaeological analyses of plant and animal remains are often challenged by acidic sediment conditions, which promote rapid decomposition of calcium-containing organic matter (Lempi€ ainen, 2002;Riikonen, 2011). This limits the reconstruction of local behaviors of time periods for which we have no written documents. One such example is the Luistari Iron Age cemetery in Western Finland. While sites elsewhere in Scandinavia, which are located on limestone bedrock, have abundant macroremain records (Larsson, 2015), in Finland there is only limited information on the diet, subsistence, and resources of Iron Age populations. Although graves from this period have been intensively studied, most analyses report limited finds (Aalto, 1997;Lempi€ ainen, 2002Lempi€ ainen, , 20052009;Lempi€ ainen-Avci et al., 2017;Vanhanen, 2012). Studies of the Luistari graves using traditional macrobotanical approaches have reported sparse finds of weeds and cereals, despite the antimicrobial and preservative effect of metal oxides originating from the numerous pieces of bronze jewellery (Lehtosalo-Hilander, 1982aLempi€ ainen, 2002;Riikonen, 2011). Only a small number of animal bones have been studied and reported, and these are mainly teeth due to their better preservation (Lehtosalo-Hilander, 1982a:309e310). Microremains have almost never been studied from Finnish Iron Age sites, the only exception being pollen analysis of a handful of grave sediments (Lempi€ ainen- Avci et al., 2017:132;Tranberg, 2018;U. Moilanen, personal communication 8 March 2019). Despite their rarity, these pollen studies have provided new information on burial practices and plants used for grave furnishing or decorations (Lempi€ ainen-Avci et al., 2017).
Many studies show that a multiproxy approach that combines microfossil analysis with traditional approaches should become a norm for archaeological analyses (García-Granero et al., 2015;Namdar et al., 2011). This approach would improve the quality of grave studies and yield a more comprehensive overview of past environments, cultures, health, and behavioral practices, particularly for acid-sediment sites, such as described below.
The eggs of intestinal parasites, (e.g. whipworms, pinworms, tapeworms, and roundworms) have been found in a variety of archaeological contexts, reflecting diet, health, and density of human and livestock populations (Bouchet et al., 2003;Cleeland et al., 2013;Dittmar and Teegen, 2003;Fugassa et al., 2008;Hald et al., 2018;Pichler et al., 2014;Søe et al., 2015Søe et al., , 2018. Parasites are passed from an individual to another under unhygienic conditions, but humans can also become hosts of animal parasites (Reinhard, 1992). Parasite remains have never been reported from Finnish sites.
Fibers, including animal hairs and plant materials, reflect the production and use of cloth and furs. Woollen textiles have been reported from the Luistari graves (Lehtosalo-Hilander, 2001). However, textiles made of plant fibers usually decompose in boreal acidic sediments, and the few fibers from Luistari that may be of plant origin, remain ambiguous (Riikonen, 2011).

Study site
Luistari cemetery is situated in Eura, southwestern Finland, and was in operation between 500 and 1500 calAD, and excavated in 1968e1992 (Lehtosalo-Hilander, 1982a, 1982b1982c. Most of the 1300 inhumation, i.e. non-cremated, burials date to the Late Iron Age (800e1200 calAD). Because of its size and richness, the cemetery is one of the most significant Iron Age sites in Finland. The Luistari archaeological collection can be considered as historically and culturally very valuable and unique, therefore we used multiproxy microremain methods to maximize the recovered information.
We analysed Iron Age inhumation burials by applying an extensive set of microfossil analyses on human dental calculus samples, supplemented by microfossil analyses of sediment samples derived from the graves, and residues on grave items. We discovered multiple types of microremains: phytoliths, intestinal parasites, animal and plant fibers, and feathers. Our data confirms that grains from grasses were eaten in the Late Iron Age, and suggests that the community suffered from poor hygiene, which enabled the spread of intestinal parasites. Moreover, our results suggest that bast fibers were eaten and processed, and that domesticated and wild animal furs and feathers were processed, and support a previously published suggestion that feather-filled pillows were used in the graves.

Collection of samples
The Luistari specimen collections are archived by the Finnish National Board of Antiquities, and we selected samples from fifteen of the graves. Our samples consisted of 32 dental calculus samples, eight small sediment samples from the graves, five carbonized residue samples, and seven sediment samples that had been classified as unidentified organic matter (here called organic residue samples). In addition, possible residue particles (here called "dirt") on the surface of 14 items were sampled (See Tables 1 and 2). Two pieces of birch bark were collected from graves 56 and 345 for radiocarbon dating (See Table A1).

Sampling and preparation procedures
Microscopic particles are readily transported even long distances, and a risk that archaeological samples might be contaminated by modern particles has to be acknowledged (Crowther et al., 2014). Therefore, to prevent contamination, we followed the cleaning procedures used in the HARVEST laboratories, at Leiden University (as described in Power et al. (2018)). We prepared the calculus samples following the recommendations by Warinner et al. (2014) and Tromp et al. (2017), and sampled items with sterile water, following Li et al. (2013), Pearsall et al. (2004), and Perry (2004). The carbonized material was prepared modifying Zarrillo et al. (2008). The sediment samples and samples from organic substances were prepared using sediment preparation protocol of the HARVEST laboratories.
See full process descriptions in Appendix A.

Analysis
The analysis was performed with transmitted light and polarised microscopy, using a Leica DM 2000 LED microscope with 400X magnification.
The intestinal parasites were identified by their morphology, such as the shape of the eggs, the type of outer membrane, the structures within the oocysts, as well as size, following Cruz et al. (2012), Kreier and Baker (1987), and Fugassa et al. (2008). Table 1 Dental calculus samples. Tooth identifications Salo (2005). Sex and age according to Lehtosalo-Hilander (1982a, 1982b. (*) ¼ Female according to Salo (2005 The morphological identification of animal hairs was based on the diameter of the hair and on the structures of medulla and cuticular scales (e.g. Chernova, 2002;Goodway, 1987;Tridico, 2005). The classifications followed Teerink (2003) and Rast-Eicher (2016), and the identification keys on Appleyard (1978), Teerink (2003), and Rast-Eicher (2016) were applied. The classification and identification of bird feathers was based on Brom (1991) and Dove and Koch (2010). In addition, samples were compared with a reference collection of Fennoscandian wild animal species and North European domestic breeds.

Phytoliths
In sediment sample 27177:24b of grave 1260 we observed a multicell phytolith, composed of dendritic long cell phytoliths and one short cell rondel phytolith in anatomical connection (See Fig. 1 and Table A2); confirmed by Dan Cabanes, Rutgers University (personal communication 5 Dec. 2018). These two phytolith types are diagnostic of the inflorescences of C 3 grasses (Ball et al., 1999(Ball et al., , 2009Out et al., 2016;Portillo et al., 2006;Rosen, 1992), including wheat (Triticum sp. L.) and barley (Hordeum sp. L.) (Ball et al., 2009). There is little information on phytoliths from Finnish native grasses, which limits our ability to identify the taxa that our phytoliths represent. We did compare the shape of the dentritic long cells to those from Triticum and Hordeum, which have species-specific morphologies. Following the measurement protocols by Ball et al. (1999Ball et al. ( , 2016 and Out et al. (2016:39e40), we were able to measure the widths of six dendritic long cells from the multicellular structure. In Table A2 our measurements are compared with Triticum and Hordeum measurements reported in Ball et al. (1999) and Rosen (1992), and our phytolith widths seem to be closer to the mean values of Hordeum vulgare L. than to those of Triticum species. We acknowledge six measurements do not enable statistically confident identification. A more extensive set of measurements and comparable morphometric measurements for rye (Secale cereale L.) and native wild grasses should be produced to enable more reliable identification.
Nonetheless, this multicell phytolith represents an important find in the context of the Luistari site and Finnish archaeology in general. This is the first time in Finnish archaeology that these types of species-indicative multicellular phytolith structures were found. Furthermore, this sediment sample was collected from around a bronze bracelet, which was located on an arm that was bent over the middle part of the body. Thus it is possible that the phytolith structure actually originates from the alimentary canal. A single grass seed from a domesticated cereal, unidentified to species, has been reported from a clay pot excavated from this same grave, supporting the dietary use of cereals at this site (Lehtosalo-Hilander, 2000).

Intestinal parasites
In addition to the phytolith (in section 3.1), two probable parasite life cycle forms were identified from the sediment sample 27177:24b of grave 1260. The first closely resembled an egg of either roundworm Ascaris lumbricoides L, which may infect humans, or A. suum Goeze, the type found in swine; these are morphologically indistinguishable (Betson et al., 2014;Søe et al., 2015). The microremain had the undulating membrane (mammilated outer surface, Cruz et al., 2012), thick middle layer, and granular content, typical for Ascaris sp. L. (see Fig. 1 and Table 3).
Although microscopic examination seldom enables species-level identification, it gives information on the parasite life stage (Cleeland et al., 2013). Ascaris sp. L. is a common parasite in humans, and has frequently been identified from ancient settlements, for instance in Viking Age Denmark (1018e1030 AD) (Søe et al., 2015).
The second probable parasite remain resembled an oocyst of a coccidia (see Fig. 1 and Table 3). It is difficult to make an exact identification because the oocyst remains were poorly preserved. The number and morphological characteristics of sporocysts and sporozoites within the oocyst are important characteristics used to distinguish coccidia (Kreier and Baker, 1987). Oocyst size may also aid identification but a reduction in oocyst size over time has been documented in, for example, eimeriid cysts from archaeological samples (Fugassa et al., 2008).
These probable intestinal parasite remains are the first reported from Finnish archaeological samples. Because these were extracted from the intestinal position of a body, it is likely that this population suffered from parasites. The effect of any endoparasite species depends on the nutritional status of the host, but also on possible primary infections with microparasites, bacteria and viruses. Because the parasites found in this study have direct life cycles, i.e. are not dependent on intermediate hosts, they do not directly indicate, for example, dietary preferences, but they may imply poor sanitary conditions.

Plant fibers
We discovered bast fibers from dental calculus samples of graves 323 and 1260 (See Table 3 and Appendix B). Bast fibers can originate from flax (Linum usitatissimum L.) or hemp (Cannabis sativa L.), the seeds of which are known from other Iron Age sites (Aalto, 1997;Lempi€ ainen, 1999Lempi€ ainen, , 2011Núñez and Lempi€ ainen, 1992), or from nettle (Urtica dioica L.), a native species, which can be both eaten or used as textile fiber (Suomela et al., 2017;Vahter, 1953). The fiber in grave 323 was blue in color, indicating a textile source. The other fiber was colorless and can originate either from textile fibers or, if it is nettle, also from food.
Bast fibers are extremely rare and interesting finds, because they were recovered from dental calculus samples, indicating that the calculus helped in preserving them. Some previous studies report plant fibers in calculus, for instance cotton from Danbury, Ohio (900e1000 AD) (Blatt et al., 2011), bast fibers from the Mediterranean Mesolithic (Cristiani et al., 2016(Cristiani et al., , 2018, hemp fibers from Eneolithic or Bronze Age Italy (Sperduti et al., 2018), and plant fibers from Early Medieval Colonna (Gismondi et al., 2018).

Animal hairs
We extracted a total of 20 animal hair fragments from graves 56, 320, 323, 345, 346, 352, 390 and 1260. These fragments were extracted from dental calculus, from sediment samples that were in contact with metal items, and from the surfaces of ceramic sherds and the bronze cauldron.
The hairs were very short, most were only 0.2e0.6 mm long, and for this reason only some of them could be identified to family or species level. Four fragments were coarse or fine sheep (Ovis aries L.) wool, and it is likely that the other mammal hairs were also from sheep. The hairs were white or brown and did not show any signs of dyeing, which might indicate that they were not originally from garments but from the dust that is created during the processing of sheep skin and the production of woolen yarns and textiles.
Additionally, several possible mountain hare (Lepus timidus L.) hairs were identified in dental calculus from grave 352, and weasel family (Mustelidae) hairs and four deer family (Cervidae) coarse hair fragments were found in the sediment sample from grave 1260. Cervidae hairs are common finds from Late Iron Age inhumations, where elk and deer skins were used for covering the deceased (Kirkinen, 2015).. (See Appendix B.)

Feathers
We identified nine definite bird feather fragments from graves 320, 325, and 390, and two additional possible feather fragments from graves 56 and 323. Three of the feather fragments were extracted from dental calculus samples (graves 320, 323, and 325), one from the surface of a piece of flint (grave 56), and five from an organic residue sample (grave 390). The fragments were 0.14e0.95mm-long barbules, with hardly any diagnostic features. However, a fragment of a plumulaceous (downy) barbule, in grave 320 calculus sample, was tentatively identified as originating from waterfowl (Anseriformes).
The feather remains from grave 390 may come from a featherfilled pillow, as previously suggested (Kirkinen, 2015).
Three feather fragments in dental calculus samples might have been layered on teeth surfaces for instance through chewing or by inhaling the dusty air when plucking birds. (See Appendix B). Feather fragments have previously been reported from dental calculus samples from Mesolithic Balkans and Early Medieval Italy (Cristiani et al., 2016;Gismondi et al., 2018).

Conclusions
The acidic sediment conditions typical for boreal environments reduce the preservation of organic archaeological remains. Consequently, traditional excavation methods, and osteological and botanical analysis cannot yield a comprehensive overview of the livelihood, culture, and resources of prehistoric people.
We examined the microremains preserved in boreal sediment samples and dental calculus remains from the Luistari cemetery collection. Sixteen of the 66 samples, and nine of the fifteen graves, contained microremains. Microremains were found from both male and female burials, and from all types of samples: dental calculus, organic remains, carbonized remains, item surfaces, and sediments.
Our study demonstrates that systematic microremain sampling of grave sites produced information that may not have been recovered otherwise. We recovered new types of finds, such as intestinal parasites, hare hairs and waterfowl feather remains, that were not represented in previous reports from this site.
Microremains preserved within dental calculus and other samples can give an extraordinary insight to the lifeways of the Iron Age people, their environment, behaviors, or even the presence of particles in the air they breathed (Radini et al., 2017). In cold weather woolen clothes may be considered as a necessity but textiles of plant fiber as luxury goods (Riikonen, 2011), and the microremains in our study record the presence of both. There is some debate whether the textiles were produced locally or acquired through trade (Riikonen, 2011), but bast fibers and animal hair remains in dental calculus suggest fibers and furs were processed by the people themselves.
It is important to recognize that the contents of the alimentary canal can be studied using microremain analysis, even when the body has already decomposed. In this study we demonstrated the opportunities that phytoliths can provide in studying digested food. The parasite egg finds addressed health and hygiene issues.
Although previous work with ancient dental calculus has discovered starch granules (Cristiani et al., 2016;Hardy et al., 2012;Henry and Piperno, 2008;Henry et al., 2011Henry et al., , 2014Power et al., 2018), our samples did not yield any. We believe that one of the reasons can be the minuscule sizes of the calculus samples. The teeth carried minor deposits of calculus, and we sampled only a fraction of this from each tooth, leaving much for future research. Moreover, acidic sediments of the site may have destroyed starch e the effect of pH on starch preservation has not been fully explored.
In this study we did not have proper control samples from sediments outside the cemetery area, to which the microremains could be compared. In the future, grave research projects should include sampling both onsite materials and offsite controls. In addition, because samples from intestinal and stomach areas have proven to have much potential, it is important to take samples from these areas from all graves, as well as other locations of the burials.
It is likely that graves contain many other plant and animal residues in addition to the types observed here. More proxies should be investigated, such as bedding and decoration material such as mosses, herbs, tree leaves, branches, and animal remains such as fish scales, insect remains, and other microfauna (Cristiani et al., 2016(Cristiani et al., , 2018Lempi€ ainen, 2009;Radini et al., 2017;Tranberg, 2018). The potential in microremain studies is endless. Research projects combining macroremain with multiproxy microremain studies can be very successful in obtaining new data on the environment, cultures, and practices of prehistorical people. Table 3 List of results.

Declarations of interest
None.