Iron Age agriculture at Trogsta, North Sweden Wennberg, Barbro Fornvännen 80, 254-262 http://kulturarvsdata.se/raa/fornvannen/html/1985_254 Ingår i: samla.raa.se

Iron Age agriculture at Trogsta, North Sweden Wennberg, Barbro

Fornvännen 80, 254-262

http://kulturarvsdata.se/raa/fornvannen/html/1985_254

Ingår i: samla.raa.se

Iron Age agriculture at Trogsta, North Sweden

By Barbro Wennberg

Wennberg, B. 1986. Iron Age Agriculture at Trogsta, North Sweden (Järnål- dersjordbruk i Trogsta, Hälsingland). Fornvännen 81. Stockholm.

Soil samples were collected for macrofossil flotation from post holes and a hearth at an Iron Age house at Trogsta, in the province of Hälsingland, N.

Sweden. Seeds of wetland plants and arable weeds dominated. While hulled barley formed the main cereal crop, oats and rye were also cultivated. Flax seeds suggest cultivation for linen production. The composition of the weed community indicates that the crops were spring-sown on well-manured fields.

The large number of seeds of wetland plants found suggests that the vegeta- tion of nearby lakeshores and marshes was either grazed or used for hay. The distribution of the different kinds of seeds implies the existence of different functional areas within the house.

Barbro Wennberg, Department of Archaeology, Umeå University, S-90187 Umeå, Sweden.

Excavations have been carried out since 1976 at T r o g s t a in the province of Hälsingland by the Dept. of Archaeology of Umeå Univer- sity. Burial m o u n d s , stone-settings and house terraces existed within the excavated area.

T h e settlement dates from the Roman Iron Age—Migration period according to the ar- chaeology finds which came to light there (Liedgren 1981, p . 57ff.). Soil samples were collected for macrofossil flotation from the post holes and a hearth in one of the houses.

I h e macrofossil analyses were primarily in- tended to answer the following questions:

Could the macrolössils provide adequate information about the kind of agricultural economy practised at Trogsta during the settlement period?

C o u l d they provide information about the type a n d size of the areas used for födder production?

Does the distribution of the different kinds of macrofossil suggest the existence of sepa- rale functional areas within the house?

Methods

A b o u t 25 % of the soil filling from each post hole was collected. This was put into plastic bags, which were taken back to the laboratory for flotation. T h r e e sieves of different mesh- sizes were used, the smallest with a mesh-size of 0.5 m m . Charcoal, d a u b and fragments of bones remained behind on the largest mesh together with most of the cereal grains. T h e smaller seeds were found in the other two sieves. T h e flotated material was sorted under a stereo-microscope Wild M 5 ( 1 2 x ) . All seeds a n d fruits were then examined at a higher magnification. Reference literature a n d a reference collection of recent material were used for the identilications.

T h e soil filling of post holes is suitable for macrofossil analyses for the following reasons:

post holes are usually evenly distributed t h r o u g h o u t the house, i.e. they yield a repre- sentative picture of the residual material in different parts of the house. Seeds easily fall d o w n into the space between a post and its stone-packing and are then protected against

Forn, annen 80 11985)

Iron Age agriculture at Trogsta 255

3.8 4,2 4,6 50 5.4 5.8 6.2 6.6 7S> 7.4 " " "

Fig. 1. F r e q u e n c y histogram for the lengths of the Horde- um vulgäre g r a i n s . — Frekvenshistogram lor längden på k o r n av Hordeum vulgäre.

T W UO 160 180 MO 250 540 L/B«100

Fig. 2. F r e q u e n c y histogram for the L e n g t h / B r e a d t h ra- tios ol the Hordeum vulgäre grains. — Frekvenshistogram liir förhållandet mellan längd och bredd på korn av Hor- deum vulgäre.

fierce fires. T h e consequences of subsequent c o n t a m i n a t i o n by recent seeds is reduced, since the post holes are sealed by the burnt remains of the posts after such fires. In north- ern Sweden the original floors are situated on the bare ground and since the earth-forming processes here are very slow, the original floor m a y låter be worn away, e.g. by trampling by grazing animals, and any subfossil finds on them destroyed. If, as commonly happens, the site is intentionally burnt off to improve

the grazing, any carbonized material found d u r i n g excavations in the upper layers of the soil may well be derived from these periods a n d not from the occupation.

Presentation of data

T h e botanical material is wholly carbonized a n d often in a poor state of preservation.

A b o u t 1400 seeds were recovered after soil flotation (Table 1). T h e largest numbers of seeds, about 500 of each, were those oiCheno- podium album and Carex spp. Those of the culti- vated plants, i.e. cereals and flax, comprise a b o u t 200 seeds altogether. Ninety-two per cent of the cereal finds were barley, Hordeum vulgäre. T w o types of barley exist, hulled and naked. Hulled barley in carbonized condition can be distinguished from naked barley by the more or less angular shape of its grain, seen in cross-section. Longitudinal striations, representing the position of the former veins on the palea, are often visible on the dorsal side. T h e ventral furrow is broader apically.

Naked barley has grains with a faint trans- verse rippling on the surface. One-third of the Hordeum grains found at Trogsta have distinct striations on the dorsal side and are angular in cross-section. M a n y grains have a twisted ventral furrow, which indicates that they are Hordeum vulgäre ssp. hexastichum or Hordeum vul- gäre ssp. tetrastichum. T h e two lateral grains on

the spikes of both these subspecies possess such furrows but no determination to subspe- cies level is possible since the internodes are missing.

T h e lengths of 85 Hordeum grains, excluding the p r o t r u d i n g part of the embryo, were measured (Feindt 1977, pp. 79ff.) together with breadth and thickness.

Max.

Mean Min.

L 7.30 5.54 3.98

B 4.07 3.14 2.08

T

3.74 2.54 1.66

L/BxlOOT/BxlOO 232 97

178 81 118 65

T h e length values show a normal distribution with the mode lying in the 5.6-6.0 m m size- range (Fig. 1). T h e distribution of the indices is shown in Figs. 2 and 3.

Fiim: linnen 80 11965)

6 5 7 0 75 80 85 90 95 T / B » 100

Fig. 3 Frequency histogram for the Breadth/Thickncss ratios of the Hordeum vulgäre grains. — Frckvenshistogram lör förhållandet mellan bredd och tjocklek på korn av Hordeum vulgäre.

The following dimensions were recorded for the oats grains found.

Max.

Mean Min.

L 5.73 5.04 3.98

B 2.82 2.12 1.66

T 2.16

1.74 0.91

L/BxlOOT/BxlOO 286

239 203

KK) 84 55

Both the two grains of rye found were rather damaged.

One grain could be measured, length 6.16.

mm, breadth 2.74 mm and thickness 2.49 mm, but the other was broken.

Altogether 16 seeds of flax were found. All but two were well preserved, showing their original form with only minor damage to the seed surface. The characteristic beak was well preserved. The flax seeds were found in post holes nos 3a, 2c and 5b, 1, 2 and 13 seeds respectivdy. The lengths and breadths of all seeds were measured.

Max.

Mean Min.

L 3.74 3.43 2.41

B 1.99 1.75 1.41

L/BX100 72 52 40

Carbonization leads to a reduction in both the length and the breadth whereas the thick- ness increases. The degree of reduction de- pends on the råte of carbonization. The oil content of the seeds changes with degree of

N

/

13b 13

a

14a

I l a I 2 b . l 2 a

9» WblOc

• • • • 'Od

iiiiiiiiima

5m

I C f l r s a l i S S a r a b l « w e e d a

graaaland p l a n t a w s t l a n d p l a n t a

Fig. 4. The distribution of the different plant groups within the house, expressed as percentages ol all the seeds found in adjacent groups of postholes. — Utbred- ningen av de olika växtgrupperna inom huset, uttryckt som procent av alla frön som finns i närliggande grupper av stolphål.

ripening and is also of significance regarding the extent of shrinkage during carbonization because the oil exerts pressure on the seed walls when vaporising. This causes the change in seed shape to a more rounded form (Helbaek 1959, p. 107f). According to Hel- baek (Helbaek 1959, p. 107 f.) the length is reduced by 25 % and the breadth by 30 %, although van Zdst (1972, p. 12) cites values of 13 % and 19 % respectivdy. On the basis of Hdbaek's values, the original size of the Trogsta flax would be as follows:

Förmännen 80 11985)

Iron Age agriculture at Trogsta 257

Max Mean Min.

4.68 4.29 3.00

2.59 2.28 1.83

a n d the size correction using van Zeist's valu- es:

Max.

Mean Min.

L 4.23 3.88 2.72

B 2.41 2.12 1.71

A large n u m b e r of Fat Hen seeds were found, most of them well preserved. Even in those cases in which only seed-halves or smaller fragments were recovered, determination to species level was still possible. T h e sizes of 100 seeds were measured. They varied in dia- meter from 1.52 m m to 0.92 m m , with a mean value of 1.17 m m .

In the absence of the perigyneum, carboni- zed Carex nutlets can r a r d y be determined to species level so that the identifications made a r e provisional. T h e predominant type be- longed within the Carex acute-group. T h e dif- ferences in form and size lie within the natu- ral limits of variability of two of the species in this g r o u p , viz. Carex acuta and Carex aquatilis.

Discussion

Hulled barley was the predominant cereal cultivated at Trogsta, compared with oats a n d rye. T h i s finding is further confirmed by a pollen d i a g r a m from a nearby marsh at T r o g s t a ( E n g d m a r k , pers. comm.). Hulled barley also formed the main cereal crop in other parts of Sweden during the Early Iron Age (Hjelmqvist 1960, p . 154, 1979, p. 51, E n g d m a r k 1981, p . 40).

Cultivation of oats became more common d u r i n g the Iron Age (Hjelmqvist 1979, p. 55).

According to numerous macrofossil finds in Seandinavia, the cultivation of rye first start- ed d u r i n g the initial part of the Iron Age, whereas on the European mainland the oldest finds only date back to the Roman Iron Age

Fig. 5. Barley grain from posthole 2c. Magnification 13 X. — Frö av korn från stolphål 2 c. Förstoring I3x.

(Willerding 1970, p . 344). In Skåne, S. Swe- den, the first finds of carbonized grains of rye also d a t e from the Roman Iron Age (Hjelm- qvist 1955, p . 86). An impression of a rye grain has also been found on a sherd from the P r e - R o m a n Iron Age in Västergötland (Hjelmqvist 1955, p. 142). Because the finds of rye m a d e at sites north of Skåne and Dan- m a r k a r e older, the condusion has been d r a w n that rye was introduced into Sweden from the east (Hjelmqvist 1955, p. 143). Early Iron Age finds of rye were also made at both V a l l h a g a r (Helbaek 1955, p . 668) and Eke- t o r p (Helbaek 1966, p. 216ff).

T h e earliest finds of flax in Europé are from Neolithie settlements in Switzerland, Germa- ny a n d northern Italy. Survey maps of the continental finds of Linum usitatissimum seeds at different periods are given in Willerding (1970). Flax seeds were found in Iron Age deposits at Eketorp (Helbaek 1966, p. 220), a n d pottery impressions of flax seeds are known from U p p l a n d (Hjelmqvist 1955, p.

145). Flax has probably been cultivated in Sweden since the Bronze Age, viz. a find in U p p l a n d of flax capsules and a flax sheaf d a t i n g from about 1000 B.C. (Lundholm

Formännen 80 (1985}

Table 1. The identified plant material from the different parts within the house.

från de skilda delarna av huset.

Det identifierade växtmaterialet

Posthole nos. la Ib 2a 2b 2c 2d 3a 4b 5b 5c 6a 6b 6c Hearth

site 7b 8a 9a 10b

Cultivated plants

Hordeum vulgäre 4 Avena sativa 1 Secale cereale

Ceralia indet.

Linum usitatissimum Arable weeds

Polygonum lapathifolium Polygonum persicaria Polygonum aviculare Chenopodium album Spergula arvensis Stellaria cf. media Fumaria sp.

Galeopsis sp.

Lamium sp.

Grassland plants Graminae Rumex acetosella Rumex acetosa Caryophyllaceae Ranunculus sp.

Cruciferac Potentilla sp.

cf. Melilotus cf. Trifolium pratense Trifolium sp.

cl. Veronica cf. Melampyrum Berry-bearing plants

Juniperus communis 1 Rubus idaeus

Rubussp.

cf. Fragaria vesca cf. Vaccinium myrtillus

Wetland plants Carex cf. acuta Carex cf. diandra Carex cf. lasiocarpa Carex Tristigmatica Scirpus cf.

Tabernaemontani Other plants Urtica dioica cf. Sagina Rosaceae Leguminosae Viola sp.

cf. Epilobium Primulaceae cl". Labiatae Galium sp.

Indet.

7 1 3 2

11

2 1

72 2 2 II 13

3 2 2

1 1 1 2

4 10 1 I 28

90 1 23 3 2 1 1

4 6

I 13 22 48

3 1 2

36 1 22 1 3 52 1

1 1

130 8 169 19 5 47 15 10 16 98 4

10c lOd Ila 12b 13a 13b 13c 13d 14a 16a S:a

4

1

5 1

1 3

6 1

1

1

6 2 1

1

1 2

2

33 1

1 2 1 2 1

10 3

1

36 7

1

1 2

1 2

5 1

1

1 1 21

2 1

1 1

2

1 6 1

1 70

1 2 5 40

1 1

1 1

5 1 29

1

1 1

1

1 3 1 1 129 1 1

9

1

1

138 10 2 21 16

8 26 8 507 10 3 1 3 1 12 6 2 1 7 1 5 1 6 8 1 1

3

136 55 45 59 17

3 2 1 1

I I I I

1 I

Fornvännen 80 (1985)

1 4 1 2

499 1 1 14

2 1 1 2 5 1 1 1 7 33 219 99 121 112 58 2 3 140 12 2 1391

1947, p . 21 f ) . T h e tradition of flax cultiva- tion in the province of Hälsingland in N. Swe- den is very old (Bennet 1738; Broman 1733).

Cultivation of Linum usitatissimum for oil is not c o m m o n in northern Sweden, because of the climate. Cultivation of flax for its fibre for textile production gives a satisfactory yield. It seems reasonable to assume that people living a t T r o g s t a cultivated flax for this purpose.

Even though the n u m b e r of seeds found was low, its cultivation may have been considcr- able. Finds of d a y loom weight also came to light in the house, indicating that weaving was carried on. T h e early cultivation of flax in n o r t h e r n Sweden is also evidenced by finds of flax pollen from the 5th century A.D, made at Rudetjärn, 12 km west of Sundsvall and a b o u t 90 km north of Trogsta ( E n g d m a r k 1978, p . 46). T h e biostratigraphy of a marsh n e a r the Trogsta exeavation site was also in- vestigated. According to the C 1 4 datings m a d e , pollen of flax and a flax seed found in these m a r s h deposits d a t e from the first few centuries A.D. ( E n g d m a r k unpubl.).

As mentioned above, the predominant ar- able weed seeds found were those of Chenopo- dium album. T h i s species is well documented in E u r o p é from numerous finds from the Neoli- thie period onwards. Villaret von Rochow (1957) considered that Fat Hen was utilised for food by m a n , whereas Muller-Schndder (1959) held the opinion that the seeds were derived from the excrement of grazing ani- m a l s . In Sweden considcrable quantities of subfossil Chenopodium seeds, both carbonized a n d uncarbonized, were found at the lake settlement at Alvastra in central Sweden (Berggren 1956, p . 101) as well as at Vallha- g a r (Helbaek 1955, p. 217). At Gene, near Örnsköldsvik on the northern Swedish coast, a similar predominance of Chenopodium album seeds was observed ( E n g d m a r k 1981, p. 40).

O n l y a few seeds of other species of arable weeds were found at Trogsta, all of which are a n n u a l s which germinate in spring. This suggests that the cereals were probably spring-sown crops. M a n y of these arable weeds require well-manured and wdl-worked soils, i.e. the fields at Trogsta were probably m a n u r e d .

Fornvännen 80 (1985)

260 B. Wennberg

Fig. 6. Oat grain from post- hole 5b. Dorsal and ventral view. Magnilication 17X. — Havrekorn från stolphål 5 b.

Dorsal och ventral sida. För- storing 17 x .

F"ig. 7. Flax seed from posthole 5 b. Magnilication 20X.

— Linfrö frän stolphål 5 b. Förstoring 20 x .

T h e large n u m b e r of Carex-nutlets found suggests that sedge hay may have been used as födder a n d / o r the wetlands such as the lakeshores a n d marshes were utilised for gra- z i n g ' Since most Carex species in northern Seandinavia grow in wet environments, even a determination of such seed finds to family level has an indicative value. In this area today Carex acuta is more common than Carex aquatilis. At Gene, the inhabitants appear to have used winter födder, although seeds of seashore plants were there more predomi- n a n t . ( E n g d m a r k 1981, p. 43).

T h e presence of only a few seeds of grass- land plants at Trogsta indicates that man- m a d e meadows were probably not important as fodder-producing areas.

T w o concentrations of seeds were found within the house (Fig. 4). Post holes 9-11 contained a large n u m b e r of the seeds of födder a n d / o r grazing plants. This part of the

Fornvännen 81111985)

h o u s e m a y have been used as a stable, or a hay b a r n , although in addition to födder stor- age within the house, there were probably o t h e r s e p a r a t e buildings for this purpose. In the s o u t h e r n p a r t of the house a concentration of finds of cereals and flax seeds occurred in t h e post holes nos. 3 a n d 5. T h i s part of the h o u s e m a y have been the kitchen, where ce- real grains were used in cooking and baking.

C l e a n i n g of flax may have taken place there too. It h a d been expected that large n u m b e r of cereal seeds would be recovered from the h e a r t h area, but in fact this did not prove to be the case. Large n u m b e r s of Chenopodium album seeds were found in both the southern a n d n o r t h e r n p a r t s of the house. In the former case the seeds may represent the residue from cleaning of the cereal crop, or they may have been used as food. T h e plants may have been b r o u g h t in as green födder for the livestock.

References

Bennet, S. 1738. Berättelse om Lins Planterande, Beredande, Spinning Wäjnadoch öfriga tilberedning. Slockholm.

Berggren, G. 1956. Växtmaterial från Iräskboplatsen i Dagsmosse. Svensk botanisk tidskrift 50:1, 97-112.

Broman, O. 1753. Underrättelse om lin. (Ursprungligen i And. Celsii. . . Almanach 1733-1735).

Engelmark, R. 1978. l h e comparative vegetalional histo- ry of inland and coastal sites in Medelpad, N Swe- den, during lhe Iron Age. Early Norrland 11:25-62.

Stockholm.

— 1981. Carbonized plant material from the Early Iron Age in N Sweden. Wahlenbergia 7:39-43. Umeå.

— Spänadsväxtodling i Norrlands förhistoria. (L'npub- lished).

Fcindt, F. 1977. Die Bedeutung des L/B Index als ein Bestimmungsmerkmal bei Getreidekörnern aus ar- chäologischen Ausgrabungen am Beispiel von Hor- deum vulgäre L. var. nudum. Mitt. Inst. Allg. Bot.

Hamburg 15:79-98.

Helbaek, H. 1955. The botany of the Vallhagar Iron Age field, pp 653—699, in Stenberger, M: Vallhagar, a Mig- ration period site on Gotland, Sweden. Storkholm.

— 1959. Notes on the evolution and history of Linum.

Kuml 103-120.

— 1966. Vendeltitnc farming produets at Eketorp on Öland, Sweden. Acta Archaeologica 37:216-221.

Hjelmqvist, H. 1955. Die älteste Geschichte der Kulturpjlan- zen in Sweden. Opera Botamca 1:3,1—186.

— 1960. Några växtlynd från Sveriges järnålder. Botanis- ka Notiser 113:141-160.

— 1979. Beiträge zur Kenntnis der prähistorischen Nutzpllanzen in Sweden. Opera Botanica 47:1—58.

Liedgren, L. 1981. Järnåldcrsgårdar i Hälsingland, pp 39—73, in Skoglund, C. Gårdar, borgar och hamnar i del äldsta Hälsingland. Umeå.

— 1984. Iron Age settlements in Hälsingland, northern Sweden. Archaeology and Environment 2, 93—112.

Lundholm, B. 1947. Abstammung und Domestikalion des Hauspferdes. Zoologiska bidrag jrån Uppsala 27:21-22. Uppsala.

Miiller-Schneider, P. 1959. Ist Chenopodium album eine prähistorische Nutzpllanze? Ber. Geobot. Forschungs- inst. "Rubel", ETH Zurich 1957, 96-114.

Villaret von Rochow, M. 1958. Die Pflanzenreste der bronzezeitlichen Plälbauhten von Valeggio am Min- chio. Ber. Geobot. Forschungsinst. "Rubel", ETH Ziirich

1957,96-114.

Willerding, U. 1970. Vor- und friihgeschichtliche Kul- turpflanzenfunde in Mitteleuropa. Neue Ausgrabungen und Forschungen in Niedersachsen 5, 287-375.

Zcist, VV. van, 1972. Palaeobolanical results ol the 1970 season at Cayönii, Turkey. Helmum 12, 1-19.

Förmännen 80 11985)

262 B. Wennberg

Sammanfattning

Sedan 1976 har arkeologiska institutionen vid U m e å Universitet utfört utgrävningar i Trog- sta, Forsa sn, Hälsingland. Bosättningen är d a t e r a d till romersk järnålder-folkvandrings- tid. Från en av husgrunderna har jordprover tagits från stolphålen samt från en härd och floterats för att utvinna makrolössil. Makro- fossil-analysen har utförts för att belysa bl.a.

följande frågeställningar: K a n dessa makro- fossil ge information om jordbruksekonomin i T r o g s t a u n d e r bosättningsperioden? Kan m a n fa reda på vilken typ och storlek på foderproducerande marker som använts?

K a n m a n av fördelningen av makrofossilen se en funktionell uppdelning av husgrunden?

Det botaniska materialet är helt och hållet förkolnat och delvis i dåligt skick. Ca 1400 frön finns i det floterade materialet. Det stör- sta antalet frön utgörs av svinmålla, Chenopo- dium album och nötter från starrväxter med ca 500 av vardera. K n a p p t 200 frön represente- rar odlade växter, sädesslag och lin.

Det d o m i n e r a n d e sädesslaget i Trogsta är skalkorn. Att d e n n a odling dominerat över de övriga sädesslagen havre och råg bekräftas även av ett pollendiagram från Trogsta stor- myr. Linodlingen har enligt litterära källor g a m l a a n o r i Hälsingland. Odling av oljelin

är ej vanlig i Norrland på grund av klimatet.

U t b y t e t av linodlingen är större vad gäller spånadslin, varför m a n kan utgå ifrån att det v a r d e t t a slags lin man odlade i Trogsta.

Fynd av vävtyngder av lera i husgrunden in- dikerar att det förekommit vävning.

Det d o m i n e r a n d e åkerogräset är svinmålla.

ö v r i g a åkerogräs finns endast i mindre antal och i samtliga fall är det ettåriga växter som i h u v u d s a k är vårgroende. Detta ger en anty- d a n om vårsådda grödor. Många av ogräsen kräver näringsrik jord varför åkrarna torde ha varit gödslade.

Det stora antalet nötter av starrväxter an- tyder att dessa växter använts som foder och a t t m a n som fodermarker använt sig av fukti- g a r e m a r k e r såsom sjöstränder och myrar.

Att m a n ej haft ett utvecklat ängsbruk fram- g å r av den ringa andelen ängsväxter i fröma- terialet.

I norra delen av husgrunden dominerar frön från foder- och betesmarker. Denna del av huset borde ha använts som stall- och foderdel. I södra delen återfinns huvuddelen av sädesslagen och samtliga linfrön. Den de- len av huset borde ha varit köksavdelning där säd a n v ä n t s vid matlagning och bakning. Där kan även linrensning ha förekommit.

Förmännen 80(1985)