SPRINGERLINK
Wood Science and Technology
Abstract
Aging of archeological wood in a salt environment was assessed in Hallstatt, Upper Austria, where Bronze Age wooden findings of the archeological salt mining site were compared with recent wood samples from the same site. Three species (Picea abies, Abies alba and Fagus sylvatica) were investigated. Recent samples covered all main geological units within the salt valley. FTIR spectroscopy was used to determine differences in wood chemistry. Principal component analysis was used to display the segregation of the data set according to the different factors and to give a measure of the strength of the effects. Results revealed that deacetylation processes took place at the wood material. The aging process affects wood chemistry significantly stronger than differences in earlywood and latewood. Together these two factors accounted for 80 % of data variability. As a third factor, the tree species segregated the data set. Geological differences in the sampling sites were not mirrored in the FTIR spectra.
Introduction
Recent and prehistoric wood samples were taken in Hallstatt, Upper Austria. Altogether, 120 wood cores were measured. The heterogeneity of the geological situation in the salt valley of Hallstatt was covered by wood samples of trees grown on all important geological zones which are limestone, depleted Haselgebirge and a mixture of both (Mandl et al. 2012). Haselgebirge is the predominant breccia rock of the salt deposit in Hallstatt, containing mainly chlorides, clay, gypsum and carbonates. Depleted Haselgebirge is the cover material at the top, where rainwater leached mainly chlorides and gypsum. Furthermore, trees from different orographic origin were sampled (650 and about 1000 m asl.).
All Samples
Spectral regions from 3485 to 2430 cm−1 and from 1838 to 400 cm−1 were included in the analysis as these regions potentially contribute to any separation. Two main effects separated the samples into four sets: age and whether only earlywood or a mixture with latewood was measured. The spectral region from 1118 to 895 cm−1 had a main impact on the separation. This spectral region contains prominent C–O bands of cellulose (Fengel and Ludwig 1991). The intensity of the cellulose band is influenced by the aging process (Fig. 2), but additionally by the tree species. In order to improve the assignment of the PCs to the different effects, the region was therefore excluded from PCA. Figure 1 displays the resulting PCA. PC 1 explains 57 % of data variability and separates the data set into two groups with a clear distance in between which indicates a strong distinguishing effect. These two groups represent the prehistoric samples on the one side and all recent samples on the other side with no exception (Fig. 1a). Therefore, the loadings plot of PC 1 explains differences in the spectra reasoned by age of the sample groups. The loadings plot displays four strong spectral regions influencing PC 1: the OH-stretching region from 3000 to 3300 cm−1 and two comparatively narrow maxima at 1730 and 1235 cm−1 (Fig. 1e). In Fig. 1b, the same scores plot is shown, but the data set is marked according to the type of measurement. The second PC can be assigned to differences in measurement (only earlywood vs. narrow tree rings with a mixture of earlywood and latewood). The PC explains 23 % of data variability. To display this effect more clearly, the prehistoric samples were removed in a second step (Fig. 3). Figure 1c again shows the same scores plot, but in this case arranged according to the tree species. Abies alba (AA) and Picea abies (PA) samples are well mixed. The separation of Fagus sylvatica (FS) samples can be reasoned at least partly by measuring only narrow rings at these samples. Further details are discussed in combination with Fig. 4. Finally, in Fig. 1d, data are arranged according to the sampling site (corresponding to different geological underground). For this factor, no obvious segregation was found.
Only Narrow Rings
Tree species result in different chemical wood composition. Differences in the ratio of lignin/cellulose/hemicelluloses, but also differences in extractives lead to different spectral patterns. To prove the effect of the factor “tree species,” the samples with only earlywood (Picea abies and Abies alba) and the samples with narrow rings (Picea abiesand Fagus sylvatica) were analyzed separately. The softwood species Picea abies and Abies alba were not separated (PCA not shown), but Picea abies and Fagus sylvaticadiffered strongly as displayed in Fig. 4. PC 1 explains 47 % of data variability and mainly segregates the two species. No specimen overlaps the two groups. As displayed in the loadings plot, maxima at 1730 and 1230 cm−1 and minima at 1510 and 1270 cm−1 are mainly responsible for the segregation. Rana et al. (2008) investigated beech spectra and found (in the region of Fig. 4b) typical peaks at 1730, 1596, 1505, 1462, 1425, 1375, 1330 and 1235 cm−1. Müller et al. (2009) investigated beech and fir samples and found in the softwood the corresponding peaks at 1738 (weaker than in beech), 1505, 1460, 1372, 1320 (as a shoulder from 1350 similar to Fig. 4b) and 1267 cm−1. These results correspond to the result given in Fig. 4c, especially the peaks at 1730, 1270 and 1230 cm−1.
Copyright Information
For further details log on website :
http://link.springer.com/article/10.1007/s00226-016-0830-4/fulltext.html
Wood Science and Technology
10.1007/s00226-016-0830-4
Aging of archeological wood in a salt environment was assessed in Hallstatt, Upper Austria, where Bronze Age wooden findings of the archeological salt mining site were compared with recent wood samples from the same site. Three species (Picea abies, Abies alba and Fagus sylvatica) were investigated. Recent samples covered all main geological units within the salt valley. FTIR spectroscopy was used to determine differences in wood chemistry. Principal component analysis was used to display the segregation of the data set according to the different factors and to give a measure of the strength of the effects. Results revealed that deacetylation processes took place at the wood material. The aging process affects wood chemistry significantly stronger than differences in earlywood and latewood. Together these two factors accounted for 80 % of data variability. As a third factor, the tree species segregated the data set. Geological differences in the sampling sites were not mirrored in the FTIR spectra.
Introduction
Organic matter usually undergoes a rapid transformation process in our environment. Depending on the recalcitrance of the molecules, mineralization and humification take place within days to few years after exposure. Several conditions can inhibit these processes and lead to subfossil materials. Dry (deserts), cold (ice, permafrost) and wet (groundwater, lake sediments, bogs and moors) conditions are favorable to keep good preservation, but also salt or some metal oxides can prevent aging processes (Jacomet and Kreuz 1999). Archeological wood samples found in marine sediments are, for example, the wrecks of the Swedish “Wasa” (Rowell and Barbour 1990) sank in 1628 and salvaged in the harbor of Stockholm 1961—the English “Mary Rose”—sank in 1545 and salvaged 1982—or the two Russian frigates St. Mikael and St. Nikolai—sank in 1747 and 1790 and salvaged in the Gulf of Finland (Reunanen et al. 1989, 1990).
The salt mine in Hallstatt, Upper Austria, housed prehistoric mining activities during the Bronze Age (1500–1100 BC) and the Early Iron Age (850–350 BC). Due to a geological disaster, both mines were filled with material. Huge amounts of soil, stones and whole trees liquidated the mining activities. Since that time, all organic material was covered by salt-rich clay. Since the nineteenth century, the prehistoric mines have been studied and therefore excavated. The high amount of organic findings gives reason for the high value of the archeological site of Hallstatt. Very rare insights into prehistoric life are possible due to textiles, wood handicrafts, leather, excrements and other findings, which usually are rotted and decomposed in other sites (Kern et al. 2009; Reschreiter 2013). In the salt environment, microbial activity can be seen as negligible (Sorokin et al. 2013). The good preservation of the findings regarding shape and even colors suggests that no aging process at all has taken place. Among the organic findings in Hallstatt, wood plays a special role due to the high number of wood pieces and the good status of preservation (Grabner et al. 2010). The latter allows dating by dendrochronology. Up to now, no study assessed the chemistry of these wood findings.
Materials and Methods
Recent and prehistoric wood samples were taken in Hallstatt, Upper Austria. Altogether, 120 wood cores were measured. The heterogeneity of the geological situation in the salt valley of Hallstatt was covered by wood samples of trees grown on all important geological zones which are limestone, depleted Haselgebirge and a mixture of both (Mandl et al. 2012). Haselgebirge is the predominant breccia rock of the salt deposit in Hallstatt, containing mainly chlorides, clay, gypsum and carbonates. Depleted Haselgebirge is the cover material at the top, where rainwater leached mainly chlorides and gypsum. Furthermore, trees from different orographic origin were sampled (650 and about 1000 m asl.).
Only earlywood was measured from Picea abies and Abies alba. Both prehistoric and recent samples of Fagus sylvatica had such narrow tree rings that it was not possible to measure earlywood-like tissue separately. Therefore, 18 samples of prehistoric and recent samples of Picea abies with narrow tree rings were chosen and additionally measured at the whole ring to quantify this effect.
Sixty-eight cores from living Picea abies, 32 from living Abies alba and 20 from living Fagus sylvatica as well as 39 cores from archeological findings were analyzed (Table 1). Recent core samples were taken by random sampling of living trees on the respective geological underground. Cores with 5.5 mm diameter were taken using usual forestry increment borers, dried at room temperature, and polished on one side. Prehistoric samples were taken from inside the archeological mining site with the help of a machine-driven hollow corer. All findings were introduced into the mine either by Bronze Age men as mine timber or during a mass movement together with debris. In both cases, one can be quite sure that wood came locally from the high valley of the salt mine as the transport of huge logs would have been very complicated. Drilling cores with 7 mm diameter were taken, dried at room temperature, and polished on one side. Storage conditions in the mine can be described as follows: The wood was embedded in mass movement material. This material consists of partly depleted Haselgebirge with water contents between 10 and 15 % DM. The wood was stored under wet conditions with a saturated salt solution in its pores. The salt consists of 95 % NaCl. The solution features a pH slightly above 7. Other inorganic compounds play only a subordinate role.
Results and DiscussionAll Samples
Spectral regions from 3485 to 2430 cm−1 and from 1838 to 400 cm−1 were included in the analysis as these regions potentially contribute to any separation. Two main effects separated the samples into four sets: age and whether only earlywood or a mixture with latewood was measured. The spectral region from 1118 to 895 cm−1 had a main impact on the separation. This spectral region contains prominent C–O bands of cellulose (Fengel and Ludwig 1991). The intensity of the cellulose band is influenced by the aging process (Fig. 2), but additionally by the tree species. In order to improve the assignment of the PCs to the different effects, the region was therefore excluded from PCA. Figure 1 displays the resulting PCA. PC 1 explains 57 % of data variability and separates the data set into two groups with a clear distance in between which indicates a strong distinguishing effect. These two groups represent the prehistoric samples on the one side and all recent samples on the other side with no exception (Fig. 1a). Therefore, the loadings plot of PC 1 explains differences in the spectra reasoned by age of the sample groups. The loadings plot displays four strong spectral regions influencing PC 1: the OH-stretching region from 3000 to 3300 cm−1 and two comparatively narrow maxima at 1730 and 1235 cm−1 (Fig. 1e). In Fig. 1b, the same scores plot is shown, but the data set is marked according to the type of measurement. The second PC can be assigned to differences in measurement (only earlywood vs. narrow tree rings with a mixture of earlywood and latewood). The PC explains 23 % of data variability. To display this effect more clearly, the prehistoric samples were removed in a second step (Fig. 3). Figure 1c again shows the same scores plot, but in this case arranged according to the tree species. Abies alba (AA) and Picea abies (PA) samples are well mixed. The separation of Fagus sylvatica (FS) samples can be reasoned at least partly by measuring only narrow rings at these samples. Further details are discussed in combination with Fig. 4. Finally, in Fig. 1d, data are arranged according to the sampling site (corresponding to different geological underground). For this factor, no obvious segregation was found.
Only Narrow Rings
Tree species result in different chemical wood composition. Differences in the ratio of lignin/cellulose/hemicelluloses, but also differences in extractives lead to different spectral patterns. To prove the effect of the factor “tree species,” the samples with only earlywood (Picea abies and Abies alba) and the samples with narrow rings (Picea abiesand Fagus sylvatica) were analyzed separately. The softwood species Picea abies and Abies alba were not separated (PCA not shown), but Picea abies and Fagus sylvaticadiffered strongly as displayed in Fig. 4. PC 1 explains 47 % of data variability and mainly segregates the two species. No specimen overlaps the two groups. As displayed in the loadings plot, maxima at 1730 and 1230 cm−1 and minima at 1510 and 1270 cm−1 are mainly responsible for the segregation. Rana et al. (2008) investigated beech spectra and found (in the region of Fig. 4b) typical peaks at 1730, 1596, 1505, 1462, 1425, 1375, 1330 and 1235 cm−1. Müller et al. (2009) investigated beech and fir samples and found in the softwood the corresponding peaks at 1738 (weaker than in beech), 1505, 1460, 1372, 1320 (as a shoulder from 1350 similar to Fig. 4b) and 1267 cm−1. These results correspond to the result given in Fig. 4c, especially the peaks at 1730, 1270 and 1230 cm−1.
Copyright Information
© The Author(s) 2016
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/) which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
http://link.springer.com/article/10.1007/s00226-016-0830-4/fulltext.html
No comments:
Post a Comment