Groningen University Library Handbook of plant palaeoecology 9491431072

Table of contents :
Cover
Table of contens
Preface
1 General Introduction
1.1 Plant taxonomy
1.1.1 Taxonomic ranks
1.1.2 Abbreviations
1.1.3 Synonyms and type identifications
1.1.4 Naming of cultivated plants
1.1.5 Plant names in written sources
1.1.6 Vernacular names of common crops
1.1.7 Genetic research
1.2 Plant ecology
1.2.1 Seed production and seed predation
1.2.2 Seed dispersal
1.2.3 Environmental conditions
1.2.4 Water stress
1.2.5 Agricultural practices
1.3 Flora and vegetation
1.3.1 Landscape, flora, and vegetation
1.3.2 Flora
1.3.3 Vegetation
1.4 Subfossil plant remains
1.4.1 Biomolecules
1.4.2 Phytoliths
1.4.3 Spores and pollen
1.4.4 Seeds and fruits
2 Palynology
2.1 The archaeobotanical archive
2.1.1 Dispersal of spores and pollen
2.1.2 Pollen precipitation
2.1.3 Peat Formation
2.1.4 Sampling and microscopic analysis
2.2 Vegetation reconstruction
2.2.1 Pollen diagram
2.2.2 Anthropogenic pollen indicators
2.2.3 Pollen precipitation and vegetation
2.3 Pollen morphology
2.3.1 Anatomy and morphology of pollen and spores
2.3.2 Glossary of terms
2.3.3 Atlas of spores, pollen, and algae
3 Non-woody macro-remains
3.1 The archaeobotanical archive
3.1.1 Origin and taphonomy
3.1.2 Sampling
3.2 Morphology of fruits and seeds
3.2.1 Morphology of fruits
3.2.2 Fruit types
3.2.3 Morphology of seeds
3.2.4 Subfossil seeds, fruits, and threshing remains
3.2.5 Cereals
3.2.6 Pulses
3.2.7 Oil and fibre crops
3.2.8 Seed atlas
4 Vegetation history of the Netherlands
4.1 The Late Glacial
4.1.1 General overview
4.1.2 The Older Dryas
4.1.3 The Allerød
4.1.4 The Younger Dryas
4.2 The Holocene
4.2.1 General overview
4.2.2 The Preboreal (10 300–8800 BP)
4.2.3 The Boreal (8800–7500 BP)
4.2.4 The Atlanticum (7500–5000 BP)
4.2.5 The Subboreal (5000–2800 BP)
4.2.6 The Subatlanticum (2800 BP to present)
4.2.7 Case study
5 Food economy
5.1 Transition to farming
5.1.1 Modelling the dawn of farming
5.1.2 Domestication
5.2 Reconstruction of the diet
5.2.1 The food spectrum
5.2.2 Cereals
5.2.3 Pulses
5.2.4 Oil crops
5.2.5 Vegetables and fruits
5.2.6 Case studies
6 Fuel
6.1 Woody plants
6.2 Non-woody plants
7 Appendixes
7.1 Chronology of the Near East
7.2 Chronology of ancient Egypt
8 Literature
8. 1 Taxonomy
8. 2 Ecology
8. 3 Flora and vegetation
8. 4 Identification
8. 5 Spores and pollen
8. 6 Fruits, seeds, and mosses
8. 7 Vegetation history
8. 8 Food economy
8. 9 Fuel
8. 10 Additional references cited
9 Indices
9.1 Taxonomic and syntaxonomic index
9.2 Subject index

Citation preview

Handbook of

PLANT PALAEOECOLOGY

GRONINGEN ARCHAEOLOGICAL STUDIES VOLUME 19

EDITORIAL BOARD Prof. dr. P.A.J. Aema Prof. dr. R.T.J. Cappers Prof. dr. L. Hacquebord Dr. W. Prummel Prof. dr. D.C.M. Raemaekers Prof. dr. H.R. Reinders Prof. dr. S. Voutsaki GRONINGEN INSTITUTE OF ARCHAEOLOGY Poststraat 6 9712 ER Groningen the Netherlands [email protected] WEBSITE www.gas.ub.rug.nl

IN COOPERATION WITH Deutsches Archäologisches Institut D.A.I. Zentrale Podbieldskiallee 69-71 D-14195 Berlin Germany [email protected] WEBSITE www.dainst.de

PUBLISHERS’ ADDRESS BARKHUIS Zuurstukken 37 9761 KP Eelde the Netherlands Tel. 0031 (0)50 3080936 Fax 0031 (0)87 7844285 [email protected] www.barkhuis.nl

Handbook of

PLANT PALAEOECOLOGY R.T.J. Cappers

Groningen Institute of Archaeology University of Groningen

R. Neef

Deutsches Archäologisches Institut Berlin

Barkhuis Groningen University Library Groningen 2012

Photomicrography: Judith Jans & René T.J. Cappers Photomacrography: René T.J. Cappers & Dirk Fennema, unless otherwise noted in the captions Book cover photographs: René T.J. Cappers Book cover design: Nynke Tiekstra, ColfsfootMedia – Noordwolde Book interior design and typeseing: Nynke Tiekstra, ColfsfootMedia – Noordwolde Copy editor: Suzanne Needs-Howarth Cover photographs: Top: Harvesting Bread wheat (Triticum aestivum) with a sickle (Ezbet Basili, Egypt; April 2002) Boom le: Goat browsing a dump area next to a roadhouse with concentrations of discarded vegetables and burned cans (Ras Banas, Egypt; March 1998) Boom right: Culms of Sorghum (Sorghum Sorghum bicolor bicolor) used as kiln fuel (El Nazla, Egypt; November 2009) Printed by: HooibergHaasbeek te Meppel

ISBN 9789491431074 Copyright © 2012 Groningen Institute of Archaeology (University of Groningen) and the Deutsches Archäologisches Institut (Berlin). All rights reserved. No part of this publication or the information herein may be reproduced, stored in a retrieval system, or transmied in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior permission from the Groningen Institute of Archaeology (University of Groningen) or the Deutsches Archäologisches Institut (Berlin). Although all care is taken to ensure the integrity and quality of this publication and the information herein, no responsibility is assumed by the publishers nor the authors for any damage to property or persons as a result of operation or use of this publication and/or the information herein.

Contents Preface 1 General Introduction 1.1 Plant taxonomy 1.1.1 Taxonomic ranks 1.1.2 Abbreviations 1.1.3 Synonyms and type identifications 1.1.4 Naming of cultivated plants 1.1.5 Plant names in written sources 1.1.6 Vernacular names of common crops 1.1.7 Genetic research 1.2 Plant ecology 1.2.1 Seed production and seed predation 1.2.2 Seed dispersal 1.2.3 Environmental conditions 1.2.4 Water stress 1.2.5 Agricultural practices 1.3 Flora and vegetation 1.3.1 Landscape, flora, and vegetation 1.3.2 Flora 1.3.3 Vegetation 1.4 Subfossil plant remains 1.4.1 Biomolecules 1.4.2 Phytoliths 1.4.3 Spores and pollen 1.4.4 Seeds and fruits

9 13 13 13 14 15 15 16 20 22 26 26 28 34 37 47 93 93 94 97 128 128 130 131 135

2 Palynology 2.1 The archaeobotanical archive 2.1.1 Dispersal of spores and pollen 2.1.2 Pollen precipitation 2.1.3 Peat Formation 2.1.4 Sampling and microscopic analysis 2.2 Vegetation reconstruction 2.2.1 Pollen diagram 2.2.2 Anthropogenic pollen indicators 2.2.3 Pollen precipitation and vegetation 2.3 Pollen morphology 2.3.1 Anatomy and morphology of pollen and spores 2.3.2 Glossary of terms 2.3.3 Atlas of spores, pollen, and algae

143 143 143 144 146 151 153 153 157 159 159 159 163 165

3 Non-woody macro-remains 3.1 The archaeobotanical archive 3.1.1 Origin and taphonomy 3.1.2 Sampling 3.2 Morphology of fruits and seeds 3.2.1 Morphology of fruits 3.2.2 Fruit types 3.2.3 Morphology of seeds 3.2.4 Subfossil seeds, fruits, and threshing remains 3.2.5 Cereals 3.2.6 Pulses 3.2.7 Oil and fibre crops 3.2.8 Seed atlas

173 173 173 199 239 239 244 246 247 248 322 333 339

4 Vegetation history of the Netherlands 4.1 The Late Glacial 4.1.1 General overview 4.1.2 The Older Dryas 4.1.3 The Allerød 4.1.4 The Younger Dryas 4.2 The Holocene 4.2.1 General overview 4.2.2 The Preboreal (10 300–8800 BP) 4.2.3 The Boreal (8800–7500 BP) 4.2.4 The Atlanticum (7500–5000 BP) 4.2.5 The Subboreal (5000–2800 BP) 4.2.6 The Subatlanticum (2800 BP to present) 4.2.7 Case study

351 351 351 353 356 356 357 357 358 359 361 364 366 370

5 Food economy 5.1 Transition to farming 5.1.1 Modelling the dawn of farming 5.1.2 Domestication 5.2 Reconstruction of the diet 5.2.1 The food spectrum 5.2.2 Cereals 5.2.3 Pulses 5.2.4 Oil crops 5.2.5 Vegetables and fruits 5.2.6 Case studies

375 375 375 380 387 387 388 397 402 403 405

6 Fuel 6.1 Woody plants 6.2 Non-woody plants

423 423 425

7 Appendixes 7.1 Chronology of the Near East 7.2 Chronology of ancient Egypt

435 435 436

8 Literature 8. 1 Taxonomy 8. 2 Ecology 8. 3 Flora and vegetation 8. 4 Identification 8. 5 Spores and pollen 8. 6 Fruits, seeds, and mosses 8. 7 Vegetation history 8. 8 Food economy 8. 9 Fuel 8. 10 Additional references cited

439 439 439 440 443 445 446 447 449 451 451

9 Indices 9.1 Taxonomic and syntaxonomic index 9.2 Subject index

457 457 467

Burning of kitchen trash outside the camp area (Berenike excavation, Egypt; 2000–2001 season).

Preface Plant palaeoecologists use data from plant fossils and plant subfossils to reconstruct ecosystems of the past. This book deals with the study of subfossil plant material retrieved from archaeological excavations and cores dated to the Late Glacial and Holocene. Subfossil plant remains offer us great opportunities for the reconstruction of former landscapes and the ways in which humans exploited and even transformed vegetations. However, to improve our knowledge of the past, we need to employ sampling procedures that will provide us with samples that contain the relevant kinds of plant material. This, in turn, means we need to have solid knowledge of all processes that act upon plant material. By modelling the transport of plant material from outside the settlement towards the settlement and its final destination in specific archaeological contexts, we can design an optimal sampling strategy to be used during the excavation. Most archaeological contexts contain plant remains that originate from more than one source and that entered that context by more than one pathway. Recognizing these pathways — which is a real challenge — may facilitate the identification of the different depositional origins within the archaeological context. It also may help to improve the typology of archaeological contexts. Fortunately, nowadays archaeobotanical research focuses much more than it did in the past on a clear link between archaeology, taphonomy, and biology. One of the main objectives of this book is to describe the processes that underlie the formation of the archaeobotanical archive and the ultimate composition of the archaeobotanical records, being the data that are sampled and identified from this immense archive. Our understanding of these processes benefits from a knowledge of plant ecology and traditional agricultural practices and food processing. This handbook summarizes the basic ecological principles that relate to the reconstruction of former vegetations and of agricultural practices in particular. Ethnoarchaeobotanical research offers the opportunity to document processes that assist us in interpreting subfossil records. We were fortunate to have many opportunities to observe traditional agricultural practices during our research stays in the Near East, the cradle of agriculture—especially in Turkey, Syria, and Egypt. Although there, too, globalization is resulting in the modernization of agriculture, some small-scale, traditional agriculture still exists. This offered us the possibility to research methods of crop growing, crop processing and storage, and food processing that in other parts of the world have long since been lost because of the transformation to large-scale modern agriculture. Although we recognize that, as with any ethnographic analogy, agricultural practices and food processing will have differed in the past because they were adapted to local conditions and traditions, our observations are nevertheless a valuable source of information that can be used to interpret

subfossil records. We grew up and were educated in the Netherlands, so because of this background, we have provided information on wild and crop plant ecology and history — and on vegetation ecology and history in general — from the Netherlands as well, as a means of verifying of our work in the Near East. Of course in the Netherlands, traditional agriculture has disappeared, and the definition of what is waste and garbage has changed enormously in an era of combine harvesters and waste disposal services. Botanical evidence relating to the past is extensive and includes both written sources and subfossil records. Our ability to interpret and use earlier written sources and identifications is, however, hampered by the ongoing changes in taxonomy, which recently have become tightly linked with genetic research. Because of these changes, the same plant can be presented under different names — which is especially true for crop plants. To facilitate the consultation of these sources, this handbook presents the valid scientific plant names together with their most important synonyms. In addition, we link the most commonly used plant names in classical Latin and Greek texts with modern plant taxonomy. A substantial part of this book gives a more detailed view of plant taxonomy and morphology, especially of the main crop plants in Europe, western Asia, and northern Africa. We hope this book will help palaeobotanists, environmental archaeologists, and colleagues from related disciplines optimize inferences based on what we could term “old-style” archaeobotany. And we hope that our observations will serve as an eye-opener and improve future research, not only as it is practised in our laboratories, but also as it is practised in the field. We would like to thank all those who supported our research and studies. In the first place, we want to express our gratitude to the people (too many to name) of the countries where we documented traditional agriculture — especially in Turkey, Syria, and Egypt — for their hospitality and sincerity. We also extend many thanks to the archaeologists (again, too many to name individually) with whom we collaborated, and of course to the institutions that have supported our work, the Groningen Institute of Archaeology (GIA) of the Rijksuniversiteit Groningen and the Deutsches Archäologisches Institut (DAI) in Berlin. We thank Suzanne Needs-Howarth (copy editor), Nynke Tiekstra (designer and typesetter), and Roelf Barkhuis (publisher) for giving a “face” to the book. We are grateful to Renée Bekker, Susanne Jahns, Benjamin Kilian, and Jan Frits Veldkamp for their critical remarks on some of the chapters. We thank our former professor Willem van Zeist, who introduced us to the field of palaeobotany. And, finally, we would like to express our gratitude to the late professor Sytze Bottema (1937–2005), who influenced both of us a lot by the way he incorporated field observations into his research.

René Cappers and Reinder Neef Groningen and Berlin, February 2012

Bark of Walnut ((Juglans regia) used for cleaning teeth and reddening lips, among other things (Cairo, Egypt; November 2007).

1

General Introduction 1.1

Plant taxonomy

1.1.1

Taxonomic ranks Plants are arranged according to their relationships within a hierarchical system. Major ranks are ordered from the general to the particular: family, genus, species, subspecies, and variety (Table 1). The International Code of Botanical Nomenclature (ICBN) governs the naming of plant names when a new classification is made. The term taxon (plural taxa) is used to indicate a body independent of the systematic level. The word taxa can thus be used in a list of plant names that represent different systematic levels. Table 1: Scientific names of some plant species, subspecies, and varieties arranged by family, and their common names in English.

Family

Genus

Species

Fabaceae

Vicia

ervilia

Vicia

faba

var. minuta

Celtic bean

Vicia

faba

var. equina

Horse bean

Vicia

faba

var. faba

Broad bean

Vicia

sativa

ssp. nigra

Black pod vetch

Vicia

sativa

ssp. sativa

Common vetch

Lens

culinaris

Lentil

Oryza

sativa

Asian rice

Oryza

glaberrima

African rice

Zea

mays

Maize

Poaceae

Subspecies/variety

Common name Bier vetch

Family names are generally derived from a genus name with many species. The family name Fabaceae constitutes an exception because the genus name Faba is used in a higher taxonomic level (the order Fabales). The name Faba vulgaris Moench. is a synonym of Vicia faba L. For some plant families, an alternative name is still used and accepted as a conserved name (Table 2).

1.1 Plant taxonomy 13

Table 2: Examples of plant families that can be labelled with the accepted name or a conserved name.

Accepted name

Genus

Conserved name

Apiaceae Arecaceae Asteraceae Brassicaceae Clusiaceae Fabaceae Lamiaceae Poaceae

Apium Areca Aster Brassica Clusia Faba Lamium Poa

Umbelliferae Palmae Compositae Cruciferae Guiferae Leguminosae Labiatae Gramineae

The basic name of a plant species consists of the genus name followed by a word designating the species (epithet). The term epithet means adjunct. Plant names for subspecies or varieties have an infra-specific epithet preceded by the abbreviation ssp. (or subsp.) and var., respectively. The abbreviation spp. after the name of the genus indicates that more than one species is involved. Family names and genus names are always capitalized. In running text, genus names, species names, and epithets names indicating a subspecies or variety are italicized. In tables, italicization is optional. Abbreviations such as ssp. and var. are not italicized. Compare the following examples: • Scrophulariaceae • Odontites spp. • Odontites vernus ssp. serotinus In a list of taxa, the genus name is abbreviated to the first letter after the first mention (e.g. Anchusa arvensis and A. officinalis).

1.1.2

Abbreviations The following abbreviations are used in relation to scientific plant names: cf. confer (compare) This abbreviation can be used before a family name or genus name (e.g. cf. Carex) or before an epithet (e.g. Carex cf. nigra). sp. species (singular) This designation following the genus name indicates that we are dealing with a taxon that is not identified beyond the genus level (e.g. Carex sp.). Alternatively, it is used when only the genus name is mentioned. spp. species (plural) This designation following the genus name indicates that we are dealing with multiple taxa, but that these are not identified to the species level (e.g. Carex spp.) ssp. subspecies (e.g. Silene latifolia ssp. alba) Alternative abbreviation: subsp. subg. subgenus (e.g. Ranunculus subg. Batrachium) var. variety (e.g. Atriplex patula var. bracteata)

14 Handbook of plant palaeoecology | Cappers & Neef

s.s.

s.l.

1.1.3

sensu stricto (in the strict sense) This is added after a taxon when its name could, in theory, include multiple taxa (e.g. Salicornia europaea s.s.); these other taxa are excluded by the use of s.s. sensu lato (in the broad sense) This is added after a taxon when no distinction is made among the multiple taxa that are included (e.g. Salicornia europaea s.l.).

Synonyms and type identifications Type identifications are often used in palaeobotanical reports to indicate a group of at least two plant species. To prevent misunderstandings about which combination of species is being referred to (for example, because plant names can change during taxonomic revisions), it is advisable to always spell out the combination of all the taxa concerned. Such enumerations should preferably be presented in alphabetical order to avoid the appearance of prioritization. Table 3 presents the traditional and updated names of a selection of type identifications from archaeobotanical publications. Table 3: Examples of traditional names and new names for combinations of plant taxa.

1.1.4

Traditional name

New name

Carex inflata/vesicaria Carex nigra type Carex otrubae type Carex serotina type Galeopsis speciosa/tetrahit Melandrium album/rubrum

Carex rostrata/vesicaria Carex acuta/nigra (elata/trinervis) Carex otrubae/vulpina Carex flava/lepidocarpa/oederi (hostiana) Galeopsis bifida/speciosa/tetrahit Silene latifolia ssp. alba/dioica

Naming of cultivated plants There are two sequential phases in the naming of plants: distinguishing taxa and naming taxa. The ICBN is used to ensure correct (re)use of scientific names. Because the ICBN only prescribes the naming of plant taxa after a taxonomic decision has been made, it cannot be used for the naming of domesticated plants and their progenitors once their relationship has been determined. However, the identification of a progenitor could be used as an argument to consider the progenitor and the domesticate to be varieties or subspecies of the same plant species. This is in accordance with the species concept, which prescribes that the progeny resulting from a crossbreeding should be fertile. This argument is, however, not always applicable. When the domestication process includes hybridization, as is the case in tetraploid and hexaploid wheat taxa, domesticated crops and their relatives remain separate species, partly representing different genera. Currently, it is only for a selection of domesticated taxa, such as Barley, Lentil, Pea, and Italian millet, that the same species name is used for both the progenitor and the domesticated plant taxon (Table 4).

1.1 Plant taxonomy 15

Table 4: Common names, old names, and new names of wild and domesticated barley. Common name

Old name

New name

Wild barley 2-row Barley 6-row Barley

Hordeum spontaneum Hordeum distichum Hordeum vulgare

Hordeum vulgare ssp. spontaneum Hordeum vulgare ssp. distichon Hordeum vulgare ssp. vulgare

Because our understanding of the relationship between progenitors and their cultivated counterparts is of recent date, many publications use the old plant names, not to mention their various synonyms. A complicating factor is that sometimes, even though a plant has been identified to subspecies, the subspecies is not mentioned when it is equal to the epithet-type. There may also be discrepancies between the names of plant taxa mentioned in tables and those mentioned in the running text. Scientific plant names can be reused in taxonomic revisions. Therefore, it is important to understand what the taxonomic status is of a particular plant name. This is achieved by adding the name of the author who was responsible for the description of the taxon in question. The author’s name is placed after the name that identifies the species within a genus (the epithet). If a taxon name includes a name for the subspecies or variety, its author’s name is also added after these names. If the taxonomic status of a subspecies or variety has to be elucidated at the level of subspecies or variety only, it is sufficient to mention the author’s name only for these lower taxonomic ranks. Thus, Triticum turgidum L. ssp. dicoccon (Schrank) Thell., being the scientific name of domesticated Emmer, can also be written as Triticum turgidum ssp. dicoccon (Schrank) Thell. The author’s name is abbreviated to a unique level. For example, L. stands for Carolus Linnaeus (1707–1778), who modernized taxonomy by introducing binomial nomenclature, and DC. stands for Augustinus Pyramus de Candolle (1778–1841), who introduced a natural plant classification system based on a proper subordination of plant characters. If a scientific name is reused, the name of the author responsible for the previous description is placed in parentheses (e.g. Vicia hirsuta (L.) Gray). Author’s names are not italicised. It is sufficient to present the scientific plant names without the authors’ names if a reference is made to a flora in which the scientific plant name is presented with the authors’ names. The scientific name of domesticated Emmer can then just be written as Triticum turgidum ssp. dicoccon.

1.1.5

Plant names in wrien sources The coupling of plant names in ancient texts with modern scientific names is often difficult. The linkage of plant names is based on a variety of evidence. Useful information may be provided by ancient lists of plants giving morphological and anatomical features or specific uses. A second source of information is linguistic evidence, in which both the ancient literature and the modern Latin nomenclature of plant names are considered. The identity of a plant can be inferred from the enumeration of plants; for example, a list of plants including only cereals or spices. The description of a plant can also present diagnostic features, such as those related to flowering or to chemical composition (e.g. toxicity and taste). Such information is, however, not always reliable. This is especially true for plants with a high

16 Handbook of plant palaeoecology | Cappers & Neef

commercial value. Traders had no interest in disseminating knowledge about plants that could threaten their trade monopoly. Compare, for example, the information that Herodotus mentions on Cassia. According to him, Cassia plants grew in shallow wetlands that were protected by fierce, winged animals. A similar description is devoted to Cinnamon: giant birds near Arabia took twigs of it to their nests and stuck them in mud, inaccessible to humans. By giving the birds large chunks of donkey meat, humans ensured the nests became heavy as lead, broke off, and fell down. Obscuring the real provenance allowed traders to both collect the Cinnamon and charge a high price at market. A good example of a clear botanical description of the Doam palm (Hyphaene thebaica) is given by Theophrastus (c. 370–285 BC) (EIP: 4.2.7; translated by A. Hort) (fig. 1): “The tree called the Doam palm is like the date palm; the resemblance is in the stem and the leaves, but it differs in that the stem and the date palm is a tree with a single undivided stem, while the other, as it increases, splits and becomes forked, and then each of the two branches forks again: moreover the twigs are very short and not numerous. They use the leaf, like the palm leaf, for plaiting. It has a peculiar fruit, very different from that of the date palm in size form and taste; for in size it is nearly big enough to fill the hand, but it is round rather than long; the colour is yellowish, the flavour sweet and palatable. It does not grow bunched together, like the fruit of the date palm, but each fruit grows separately; it has a large and very hard stone, out of which they turn the rings for embroidered bed-hangings.” Figure 1: Full-grown tree and fruit of the Doam palm (Hyphaene thebaica). The cross section shows the multi-layered fruit and the ivory-coloured hollow seed.

1.1 Plant taxonomy 17

A compilation of Latin and Greek plant names used in classical texts is presented in Table 5. Latin names are obtained from Marcus Porcius Cato (234–149 BC; De Agri Cultura; On agriculture); Marcus Terentius Varro (116–26 BC; Res Rusticae; On agriculture); Lucius Junius Moderatus Columella (c. 4–70 AD; Res Rustica; On agriculture); and Gaius Plinius Secundus (23–79 AD; Naturalis Historia; Natural history). The Greek names are taken from two books by Theophrastus (Пερί Φυτών Іστορίας [Enquiry into plants]; Тα Περί Φυτών Αιτιών [De Causis Plantarum]). In this table, only those names that can be related to their valid scientific names are mentioned. The absence of a Latin or Greek name for a specific taxon that is currently accepted does not necessarily imply that such taxa were not recognized in the past. In describing the variability of barley, for example, Theophrastus distinguishes between barley bearing two, three, four, five, and six rows of grain kernels (EIP: 8.4.3). This spectrum of barley ears most probably represents lax and dense forms of 2-row Barley (Hordeum vulgare ssp. distichon) and 6-row Barley (H. vulgare ssp. vulgare). Unfortunately, the interpretation of the Latin and Greek names in classical texts is not unequivocal. Their use can change through time, and sometimes an author designates the same name to plants that are currently considered separate taxa. The name Triticum, for example, in its broadest sense refers to both hulled and naked wheat. For example, dehusked grain kernels of Triticum turgidum ssp. dicoccon can be labelled with the name Triticum. In a broad sense this label means only naked wheat and in a narrow sense it refers only to naked wheats with hard grain kernels, including Triticum turgidum ssp. durum and T. turgidum ssp. turgidum. In the most strict sense it relates only to the latter, having the softer grain among the harder types of naked wheat (Jasny, 1944). Until recently, the classical sources mentioned above were the main sources for reconstructing ancient farming practices (e.g. Spurr, 1986). Jasny (1944) gives examples of mistranslations in these classical sources. Dalby (2003) has published an updated compilation of food plants mentioned in classical texts, including more obscure sources such as poems and recipes. Table 5: Modern scientific plant names and Latin and Greek names of some cereals and pulses mentioned in classical texts (aer Estienne [1816-1828], Berkowitz, Squitier and Johnson [1990] and Jasny [1944] for wheat, and Hondelmann [2002]).

Scientific plant name

Latin name

Greek name

Avena sativa

• avena

• βρόμος

Cicer arietinum

• cicer

• ἐρέβινθος

Hordeum vulgare ssp. distichon

• hordeum distichum • hordeum galaticum

• κριθή

Hordeum vulgare ssp. vulgare

• hordeum hexastichum • hordeum cantherinum

• κριθή

Lathyrus sativus

• cicercula

• λάθυρος

Lens culinaris

• lens

• φακός

Lupinus albus

• lupinum

• θέρμος

Oryza sativa

• oryza

• ὅρυζον

18 Handbook of plant palaeoecology | Cappers & Neef

Table 5: Modern scientific plant names and Latin and Greek names of some cereals and pulses mentioned in classical texts (aer Estienne [1816-1828], Berkowitz, Squitier and Johnson [1990] and Jasny [1944] for wheat, and Hondelmann [2002]).

Scientific plant name

Latin name

Greek name

Panicum miliaceum

• milium

• ἔλυμος • μελίνη • κένγχος

Pisum sativum

• pisum

• πισός

Secale cereale

• secale

• βρίζα

Setaria italica

• panicum

• ἔλυμος • μελίνη • κέγχος

Triticum aestivum ssp. aestivum

• siligo

• σιτανίας

Triticum aestivum ssp. compactum • siligo

• σιτανίας

Triticum aestivum ssp. spelta

• siligo (soft wheat)

• σιτανίας (soft wheat)

• scandulae sive speltae (hulled tetra-/hexaploid wheat)

• σκανδοῦλης (Frank et al., V, p. 318 s.)

• speltae mundae (dehusked tetra-/hexaploid wheat)

• πιστίκιον (Boak 1940, pp. 49 ff.)

Triticum monococcum

• tiphe

• τίφη (Gal., De Al. Fac., I, 13, 26) • τίφη ἀπλῆ

Triticum turgidum ssp. dicoccon

• far • far adoreum • ador; adoreum

• όλυρα (hulled tetraploid wheat; soer kernel) • τίφη δίκοκκον • ζειά (hulled tetraploid wheat; harder kernel) • ζειά δίκκοκος

• triticum • similago

• σεμιδαλίτης (hard kernel; Rome)

• scandulae sive speltae (hulled tetra-/hexaploid wheat) • speltae mundae (dehusked tetra-/hexaploid wheat)

• σκανδοῦλης (Frank et al., V, p. 318 s.) • πιστίκιον (Boak 1940, pp. 49 ff.)

• alica (Cato, LXXV-LXXXII) • χόνδρος (dehusked grain from ζειά) • tragum (Pl., Nat. His., XVIII, 7, 16, 76) (Gal., De Al. Fac., I, 6, 1) • arinca (Pl., Nat. His., XVIII, 8, 19, 81) • τράγος (tetra-/hexaploid dehusked grain from όλυρα) (Geoponica, ed. Beckli, III, 3, 8) • σεμιδαλίτης (Greece)

Triticum turgidum ssp. durum

• triticum

Triticum turgidum ssp. turgidum

• triticum ramosum (branched rachis) • κριθανίας (branched rachis) • triticum robus (Columella, II, 6, 1)

Vicia ervilia

• ervum

• ὅροβος

Vicia faba var. equina

• faba

• κύαμος

Vicia faba var. faba

• faba

• κύαμος

Vicia faba var. minuta

• faba

• κύαμος

Vicia sativa

• vicia • aphaka

• ἀφάκη

1.1 Plant taxonomy 19

The linguistic affinity of old names with modern ones can also be used, but with some reservations. In principle, any taxon may be subject to some remodelling, and as a result, new names may be introduced, with the assignment determined by the International Code of Botanical Nomenclature (ICBN). Each scientific plant name is followed by the author citation, referring to the Latin description or diagnosis of the taxon in question. This enables researchers to judge the status of a specific scientific plant name and also provides a foundation for comparing synonyms. Because the ICBN rules may result in the reuse of existing names for newly defined taxa, linking old trade names with analogous modern scientific plant names may lead to incorrect identifications. Cassia, for example, being an ancient commodity of which the uncurled bark was traded, should not be confused with cultigens of the genus Cassia, including senna (Senna alexandrina Mill.) and purging cassia (C. fistula L.), both cultivated for their fruits. Sarcocolla may serve as a second example. Though this brittle resin is still traded in the spice markets of the Near East, its real source is not yet clear. Species belonging to three genera have been put forward. Some of them are known as sarcocolla, yet it has to be realized that Sarcocolla Kunth is a synonym of the currently accepted name Saltera Bullock (Cappers, 2006).

1.1.6

Vernacular names of common crops In each language, vernacular or common names are used for crop plants that are part of the diet (Table 6). For some plants, such as Camelina sativa and Linum usitatissimum, more than one common name is in use, sometimes relating to a specific use. Whereas scientific plant names are unique for a particular plant taxon and indicate the phylogenetic relationship, common plant names may be confusing because they are not related to the classification in genera and families. The enumeration of commodities in classical texts, for example, shows that rice was not considered a grain. And even today, common names can be misleading, as is the case with the trade name Pearl barley, which is used for polished grain kernels of Coix lacrymajobi. Another example are the trade names used for the flattened fruits of Prunus persica var. compressa. Both the English name (Wild peach) and the Dutch name (Bos perzik) erroneously suggest a kind of ancestry for the more common Peach (P. persica var. persica). It also happens that a common name is used in different languages for different plants with a similar use. An example of such a name is cress (Dutch: kers; German: Kresse; French: cresson), which is used for different plants. As a common name for plants eaten as salad, it is applied to several species of the Brassicaceae in particular, including species of Arabis, Nasturtium, and Lepidium. This vernacular name can even be traced back to cuneiform texts, in which it is used for edible and medicinal plants. The identity of the plant labelled with this name in cuneiform tablets unearthed in Bronze Age tell Sabi Abyad, located in northern Syria, could be established as Lepidium sativum because a small pottery container filled with seeds of this plant, most probably used as seeds for sowing, was found as well.

20 Handbook of plant palaeoecology | Cappers & Neef

Table 6: Selection of Dutch, English, German, and French names for some economic plants. Varieties and subspecies are only mentioned if domesticated species are differentiated at these levels.

Latin

Dutch

English

German

French

Avena abyssinica

-

Abyssinian oat

-

-

Avena brevis

Korte haver

Short oat

Kurzhafer

-

Avena nuda

Naakthaver

Naked oat

Nackthafer

Avoine nue

Avena sativa

Haver

Common oat

Hafer

Avoine

Avena strigosa

Evene

Bristle oat

Rauhhafer

Avoine rude

Camelina sativa

Dederzaad/ Huentut

False flax

Leindoer

Caméline

Cannabis sativa

Hennep

Hemp

Hanf

Chanvre

Cicer arietinum

Kikkererwt

Chickpea

Kichererbse

Pois chiche

Corylus avellana

Hazelnoot

Hazel

Haselnuss

Noisetier

Fagopyrum esculentum

Boekweit

Buckwheat

Buchweizen

Sarrasin

Hordeum vulgare ssp. distichon

2-rijïge gerst

2-row Barley

Zweizeilige Gerste Orge à deux rangs

Hordeum vulgare ssp. vulgare

6-rijïge gerst

6-row Barley

Mehrzeilige Gerste

Orge à six rangs

Isatis tinctoria

Wede

Woad

Färber-Waid

Pastel des teinturiers

Lathyrus sativus

Zaailathyrus

Grass pea

Saat-Plaerbse

Gesse blanche

Lens culinaris

Linze

Lentil

Linse

Lentille

Lupinus albus

Wie lupine

White lupin

Weiße Lupin

Tirmis

Linum usitatissimum

Lijnzaad/Vlas

Linseed/Flax

Saatlein/Flachs

Lin

Malus sylvestris

Wilde appel

Crab apple

Wilder Apfelbaum Pommier sauvage

Olea europaea

Olijf

Olive

Olive

Olive

Oryza sativa

Rijst

Rice

Reis

Riz

Panicum miliaceum

Pluimgierst

Broomcorn millet Rispenhirse

Millet

Papaver somniferum

Maanzaad

Opium poppy

Schlafmohn

Oeillee

Pisum sativum

Erwt

Pea

Erbse

Pois

Secale cereale

Rogge

Rye

Roggen

Seigle

Setaria italica

Trosgierst

Foxtail millet

Kolbenhirse

Millet des oiseaux

Triticum aestivum ssp. aestivum

Broodtarwe

Bread wheat

Saatweizen

Blé tendre

Triticum aestivum ssp. compactum

Dwergtarwe

Club wheat

Zwergweizen

Blé compact

Triticum aestivum ssp. spelta

Spelt

Spelt wheat

Dinkel/Spelz

Épeautre

Triticum monococcum

Eenkoren

Einkorn

Einkorn

Engrain

Triticum timopheevii

-

Zanduri wheat

Sanduriweizen

-

Triticum turgidum ssp. dicoccon

Emmer

Emmer

Emmer

Amidonnier

Triticum turgidum ssp. durum

Harde tarwe

Hard wheat

Hartweizen

Blé dur

Triticum turgidum ssp. turgidum

Engelse tarwe

Rivet wheat

Rauhweizen

Blé poulard

Vicia ervilia

Biere wikke

Bier vetch

Linsenwicke

Ervilier

Vicia faba var. equina

Paardenboon

Horse bean

Pferdebohne

Fève à cheval

Vicia faba var. faba (= var. major)

Tuinboon

Broad bean

Dicke Bohne

Grosse fève

Vicia faba var. minuta (= var. minor)

Duiveboon

Celtic bean

Kleine Ackerbohne Féverole

Vicia sativa

Voederwikke

Common vetch

Saat-Wicke

Vesce cultivée

Vitis vinifera

Druif

Grapevine

Weinrebe

Vigne

1.1 Plant taxonomy 21

1.1.7

Genetic research Until recently, the classification of plants was mainly based on morphological and anatomical features. Over the past 20 years, much research has been done on the genetics of plants, resulting in a better understanding of phylogenetic trees and in major shifts of plants between genera and even families. However, this work is still in progress and current floras only present a status quo of our increasing insight into phylogenetic relationships. For example, based on available genetic research, the traditional classification of the Brassicaceae family turns out to be rather artificial, and a revision may be expected. Genetic research is applied on two levels: the size of the genome and the nucleotide sequences of genes. The genome is the entire number of genes present in each cell of a particular plant taxon. A gene is a functional unit bearing the code for a specific feature of the organism. Genes are linked together in long molecules, called deoxyribonucleic acid (DNA). In organism with a large number of genes, the DNA is split up into several parts, each of which is called a chromosome. In comparison with mammals, the amount of DNA in the flowering plants varies tremendously. This is partly related to the presence of multiple copies of the chromosome set (see below). For example, the total genome of Arabidopsis thaliana contains 157,000,000 base pairs (157 megabases; Mb) representing 25,498 genes, which are subdivided over five chromosomes. Because of its relatively small genome size, Arabidopsis thaliana is one of the model plants for genetic research (fig. 2). Genes are concentrated in ‘gene islands’, and more than 80 percent of the genome consists of repetitive elements (e.g. retrotransposons).

Figure 2: Its small plant size, short life cycle, and small genome make Arabidopsis thaliana one of the model organisms for genetic research.

22 Handbook of plant palaeoecology | Cappers & Neef

Most flowering plants are classified as diploid, which means that the genome is present as a double set of chromosomes in the nucleus of each cell. Only the gametes (pollen and ovules) of the plants have a single copy in their nucleus, being the result of a special cell division called meiosis. The division of the genome into a unique set of the genome makes it possible that each fusion of a pollen grain and an ovule results in a diploid cell. In this way, the size of the genome remains the same for a particular plant species. The number of the unique (haploid) chromosome number is designated with the letter n (1 x n = n). Diploid organisms have two homologous chromosomes in each cell nucleus, which is designated ‘2n’. A homologous set of chromosomes contains the same genes in a fixed sequence, although the expression of both genes may differ. Arabidopsis thaliana has five different chromosomes. The genome of this diploid plant species can thus be designated as n = 5 or 2n = 10. The genome is designated with a capital letter. This is of special interest for illustrating the composition and origin of the genomes of plant species in which changes of chromosome numbers occur (viz. polyploids; see below). Aberrant numbers of chromosomes may be present in organisms. Humans have 23 different chromosomes (n = 23), but abnormalities occur due to the absence of, for example, one of the sex chromosomes (genome is 2n − 1) or the presence of an extra copy of chromosome number 21 (genome is 2n + 1). In both cases, the aberrant number of chromosomes results in an abnormal human phenotype. The former leads to Turner syndrome and the latter to Down syndrome. Because the aberrant chromosome number results in a combination of specific, clinically recognizable features, the term syndrome is used. In plant species aberrant numbers of chromosomes also occur. This phenomenon is of special interest with respect to the process of plant domestication. The term domestication syndrome has been introduced to describe the spectrum of traits that characterize domesticated plants. A disadvantage of this term is that it is based on a clinical perspective rather than on an ecological framework. See section 5.1.2 for a further discussion on the use of this term. Variations in chromosome numbers can be achieved by reducing or increasing the number of chromosomes. The reduction of the number of chromosomes is largely restricted to diploid plant species, resulting in double haploid plants (monoploids) in which every nucleus has just a single copy of each chromosome. In fact, the genome of a monoploid, which in essentially is a diploid organism, is similar to that of a haploid gamete (pollen and ovules). Monoploids can be artificially produced by special tissue cultures in which, for example, plants are grown from pollen or unfertilized ovules. The use of pollen is preferred because they are more easily obtained in large quantities. Because genes can still be represented by several copies in the haploid genome, the phenotypic expression is only predictable when dealing with major genes. This kind of crop manipulation has been successfully applied to, for example, Soybean (Glycine max) and Common tobacco (Nicotiana tabacum). Once such monoploids have been cultivated, the chromosome number is doubled again by treating the cells with chemicals. In this way, double haploids are created having similar (homozygotous) chromosomes.

1.1 Plant taxonomy 23

We are dealing with polyploids if more than two copies of the chromosome set are present. The genome of a triploid plant consists of three copies of the unique chromosome set (3n), that of a tetraploid has four copies (4n), that of a pentaploid has five copies (5n), that of a hexaploid has six copies (6n), and so forth. Each pair of chromosomes is designated with a capital letter. For example, the genome of the hexaploid Bread wheat (Triticum aestivum ssp. aestivum) is BBAADD, and can be written as 6n. This wheat species has 21 different chromosomes (7B, 7A and 7D), the total number of chromosomes being 6 x 7 = 42. Preferably, the genome is written as 2n = 6x = 42, in which n is the sum of a single set of chromosomes. This means that it behaves as a diploid plant (viz. the number of chromosomes of BAD = 3 x 7) and x is the number of one unique set of chromosomes (= 7). This description can be reduced to 2n = 42, in which the total number of chromosomes is mentioned but the polyploidy level remains obscure. The sequence of the genome designation in genomic formulas is determined by the taxon providing the cytoplasm (ovule); thus female parent x male parent. The BA genome of Triticum turgidum indicates that the B genome has been donated from a female Aegilops plant (present in the ovule) and the A genome stems from a male Triticum plant (present in the pollen grain). The hybrid offspring receives the organelles, residing in the cytoplasm, from the female parent. In plant cells, organelle genes are present in mitochondria (producing energy-rich ATP molecules) and chloroplasts (conducting photosynthesis). Recombination of mitochondrial DNA (mtDNA) and chloroplast DNA (cpDNA) does not occur during sexual reproduction, but changes in the DNA structure (viz. the sequence of the bases) result from rapid mutations. For this reason, studying the organelle genes is of much interest for understanding the evolutionary relationships of plants. The presence of more than two copies of the chromosome set is rather common in the plant kingdom. About one-third of all flowering plants are polyploids, and the proportion is even higher in economic plants. Because the number of chromosomes is large in polyploids, the nucleus and cell are often larger as well, resulting in plants that are more flexible and can adapt to different conditions and habitats. Polyploidy can be the result of chromosome duplication within a species, which is called autopolyploidy, but it can also be the result of a crossing between two different species (hybridization), which is called allopolyploidy. Autopolyploids have been cultivated because of the increase in flowers and fruits in particular. Fruits of Grape (Vitis vinifera) may serve as an example. The fruits of the diploid species (genome: 2n = 38) are normalsized, whereas those of the tetraploid cultivars (genome: 4n = 76) have fruits that are almost twice as large. Allopolyploidy has played an important role in the cultivation history of several crops, including Wheat (Triticum), ), Banana ((Musa), Sugarcane (Saccharum), Potato (Solanum), and Sweet potato (Ipomoea). Triticum urartu and T. monococcum are both diploid, having two sets of seven chromosomes, but their genome is distinct, and the related chromosomes show little affinity, as a result of which hybridization is rare (Table 7). A superscript letter designates the different character of both genomes: Au and Am. These letters are not fixed, and different combinations are used in publications. The two tetraploid wheat species, T. timopheevii and T. turgidum, both have the A genome, which is donated by Triticum urartu. The domestication of wheat also

24 Handbook of plant palaeoecology | Cappers & Neef

includes hybridization with Aegilops species. Recent research shows that the B and G genomes were probably donated by Ae. speltoides and the D genome by Ae. tauschii, although the origin of the B genome is still a matter of debate. The domestication history of wheat most probably included two independent hybridizations of Ae. speltoides with T. urartu. The first hybridization could have taken place some 300,000 years ago, by which the B genome was donated. During a second, more recent, hybridization between the species about 90,000 years ago, the G genome was donated. Due to the long time span since these hybridizations, the genome of the outcrossing Ae. speltoides has evolved, and identical alleles are only found in the G genome; hence it gets its own designation with the letter S. Most probably, specimens of Ae. speltoides having a genome that is almost identical with the B and G genomes cannot be found anymore (Kilian et al., 2009; Kilian et al., 2011; Wang et al., 2011). Table 7: Genome designation of Aegilops and Triticum species. The two Aegilops species have donated their genome by hybridization.

Plant species

Genome S

Aegilops speltoides

B

G

A

Aegilops tauschii

D

Triticum urartu

A

Triticum monococcum

Am

u

G

Triticum timopheevii Triticum turgidum

B

Triticum aestivum

B

Triticum zhukovskyi

D

S

Au Au Au

G

A Am u

2n = 2x = 14

2n = 4x = 28 2n = 6x = 42

Recently, artificial hybridization has resulted in new crops, such as Triticale (x Triticosecale) and several Cabbages (Brassica). Triticale is a hybrid of Wheat (Triticum) and Rye (Secale cereale), combining the high yield and protein content of wheat with rye’s the adaptation to more harsh environments and its high lysine content, one of the essential amino acids. Several hybrids exist in which genomes of different wheat species have been used. No valid names on a species level are available yet. Although the first crossings were made in the late 1890s, it only became a commercial crop in the 1970s. This commercially produced hexaploid Triticale (genome AABBRR) is a hybrid of the tetraploid Hard wheat (Triticum turgidum ssp. durum; genome BBAA) and the diploid Rye (genome RR). The octaploid Triticale (genome BBAADDRR), resulting from a cross between the hexaploid Bread wheat (Triticum aestivum; genome BBAADD) and Rye, has less favourable traits and is therefore not cultivated on a commercial basis. Within the genus Brassica, three diploid species are considered to be basic species: B. nigra (2n = 2x = 16), B. oleracea (2n = 2x =18), and B. rapa (2n = 2x = 20). Hybridization between B. nigra and B. oleracea resulted in the tetraploid B. carinata (2n = 4x = 34), hybridization between B. nigra and B. rapa in the tetraploid B. juncea (2n = 4x = 36), and hybridization of B. oleracea and B. rapa in the tetraploid B. napus (2n = 4x = 38). Each of these tetraploids is represented by several varieties, demonstrating the large phenotypic potential within this genus.

1.1 Plant taxonomy 25

The classification of plants has recently been improved by the analysis of DNA sequences. This kind of research focuses on the specific gene sequences, such as that of the gene rbcL (ribulose biphosphate carboxylase, large subunit). This gene is present in the chloroplast DNA and has the genetic code for the production of an enzyme that triggers an essential process in photosynthesis. Because of this essential trait, this gene is present in all angiosperms and is suitable for this kind of research. The phylogenetic relationship between plants is inferred by specific dissimilarities in the base sequences of the gene. The conserved regions of this gene are of special interest when comparing the same gene in different species, because they accumulate mutations over time. Assuming a rather constant rate of mutations, the dissimilarities in base sequences can be interpreted on a time scale and used for the reconstruction of phylogenetic trees. This kind of reconstruction is, however, hampered by specific events that can lead to a higher rate of mutations. The Angiosperm Phylogeny Group (AGP), which has published three overviews so far, governs this research (viz. APG 1 in 1998, APG 2 in 2003, and APG 3 in 2009).

1.2

Plant ecology

1.2.1

Seed production and seed predation Subfossil plant remains can be used to reconstruct the former vegetation and the past food economy. Ideally, archaeobotanical data records should consist of the scientific plant name; the plant part; the number of plant remains; the preservation condition; and the presence of specific features dealing with fragmentation, predation, and processing. The number of seeds or fruits, being part of most archaeobotanical records, can be indicative of the relative contribution of each plant to the former vegetation or food diet. This does not imply, however, that the observed differences in numbers coincide with the original representation in the vegetation or diet. Several variables have to be taken into account for a more reliable interpretation. For the reconstruction of the former vegetation in a particular landscape, these include seed production, seed dispersal, and the taphonomic processes that act on seeds after deposition over time. A complicating factor for the reconstruction of vegetation is that descriptions of vegetation are based on the surface area coverage of the individual species, whereas archaeobotanical records predominantly consist of seed counts. A possible pathway to link the archaeobotanical data record to the syntaxonomy will be discussed in section 1.3.3. Variables that determine the presence in the archaeobotanical data record of edible plants deal with crop processing and food preparation in particular. A meaningful conversion of seed counts to energy levels will be discussed in section 5.2.5. Seed production is the primary variable that determines the number of seeds recovered from an archaeological sample. Differences in seed production are, however, not mirrored in the archaeobotanical archive. Seed dispersal and all kinds of taphonomic processes are responsible for a shift in the proportions between plants. The potential seed production is primarily determined by the number of ovules present in the lower part of the pistil. Actual seed production is primarily determined by the degree of fertilization. This, in turn, depends

26 Handbook of plant palaeoecology | Cappers & Neef

on the type of pollination, the sexuality of the flowers (whether bisexual or unisexual), and the sex ratio of individual plants and populations. If plants are self-pollinating, such as cereals, seed production may be optimal, whereas in cross-pollinated plants the number of seeds is highly dependent on the success of pollen transport. The number of seeds produced in bisexual flowers, containing both female (carpel) and male (stamen) structures, may resemble the potential seed production better than the number produced in unisexual flowers, especially when the plant species has only carpellate or staminate flowers (dioecious species). An example of a dioecious plant is the date palm (Phoenix dactylifera). Humans prefer female plants because they produce the date fruits. An increase of female trees in a population is achieved by removing juvenile male trees. The difference in sturdiness of the leaf spines is used to determine the sex of a young tree. To increase the chance that a female tree will grow in a new spot, often several date seeds are planted together. This gender selection may result in palm tree populations containing as much as 99 per cent female trees. A disadvantage of such extreme gender selection is that the number of pollen produced by the male palm trees in a certain area is not sufficient for fertilizing all female flowers. As a result, the production of dates is reduced despite the preference for female trees. To optimize the fruit production in such unbalanced populations, artificial fertilization is practised. The earliest evidence of artificial fertilization of date palms is documented in a cuneiform text found in Ur (located in modern-day southeastern Iraq), dated to 2300 BC. In Egypt, artificial fertilization was probably introduced a couple of hundreds of years later, during the Middle Kingdom (Murray, 2000). Seed production is essential in the plant’s life cycle. Seeds are the new offspring, and biomass allocation to the seeds is of vital importance. Seed production may vary considerably among plant species and is partly the result of a trade-off between seed number and seed size. Members of the orchid family, for example, produce huge quantities of very small seeds. The necessity of this trade-off can be linked to the scarcity of safe sites for seed germination. Orchid seeds can only germinate if specific fungi are present in the soil. So although the chances of a seed falling into a location where such a fungus is present are small, the production of many seeds makes it work. Seed production can be reduced by seed predation. This is especially the case in plant species that have storage reserves in their seeds. They become an attractive food source particularly when such energy-rich seeds are produced in large concentrations. This is the case in plant families that have infructescences in which seeds are concentrated in dense clusters, such as the Daisy family (Asteraceae) and the Carrot family (Apiaceae). If seeds are not concentrated in large densities but have relatively large storage reserves due to their seed size, such as those of the Pea family (Fabaceae), they are also attractive to animals. Some animals, such as rodents, are generalists and feed on seeds from different plant species. Others have become specialists and are adapted to specific host plants, such as weevils that complete their life cycle in seeds of the Fabaceae. This kind of hazard is thus typical for the domesticated pulses as well. Once the seeds have been harvested and stored, a serious reduction of the yield may only become visible after some time, when the infected seeds have been reduced to an empty seed coat (fig. 3).

1.2 Plant ecology 27

Figure 3: Lentil seed infected with a beetle (Bruchus; left) and a full-grown specimen of the beetle (right).

Figure 4: The diaspore of Wood-sorrel (Oxalis acetosella) is a seed (left); that of Sea-kale (Crambe maritima) is a fruit (middle); and that of Rough cocklebur ((Xanthium strumarium) is the whole infructescence (right).

1.2.2

Seed dispersal Plants can propagate themselves vegetatively or generatively. Vegetative reproduction is nonsexual and includes the propagation by vegetative plant remains, such as shoots, underground stems (rhizomes), or bulbs. Generative reproduction includes the recombination of genes; the dispersal unit is a diaspore, which includes the seed and sometimes other plant parts, such as the fruit or other parts of the flower (fig. 4). Irrespective of the composition of the diaspore, the term seed is commonly used in relation to dispersal (viz. seed dispersal) and their presence in the top soil (viz. seed bank). The difference between seed and fruit is explained in section 3.2. In this chapter, the word seed is used in its broader sense. Although all ecofacts from an excavation are subject to some type of dispersal, mainly by the vector human activity, it is a typical feature of all seeds and fruits to have the potential of natural dispersal. A basic understanding of seed dispersal will therefore improve our interpretation of the subfossil seeds and fruits retrieved from excavations. It is a misconception to assume that plants, in contrast to animals and humans, are not mobile. Plants start their life as a seed, and it is in this particular shape that they are adapted to dispersal. Seed dispersal follows two different pathways: time and space. Seeds that are dispersed in time lack

28 Handbook of plant palaeoecology | Cappers & Neef

Figure 5: Examples of wind-dispersed seeds. Small seed of Marsh helleborine (Epipactis palustris) with an inflated seed coat (top); fruit of Rhodanthe humboldtiana with hairs (middle); and fruit of Silver birch (Betula pendula) with wings (bottom).

morphological features for transport and are only dispersed over small distances. Such seeds can survive for considerable periods in the soil due to, for example, germination delay by dormancy, before they finally germinate and fulfil their life cycle in the vicinity of the parent plant. Especially when such plants grew within settlements, archaeobotanical samples may contain huge numbers of their seeds. Those of Common orache/Spear-leaved orache ((Atriplex patula/prostrata; the one-seeded fruits from archaeological contexts are not identified to the species level) may serve as an example. Some 300,000 fruits were found in a single 3 l sample taken from a waste pit on the slope of the dwelling mound Heveskesklooster in the northern part of the Netherlands, dated to 800–1300 AD. The total volume of these fruits is 73 ml. Both species produce considerable numbers of fruits: the number of fruits from Common orache ranges from 100 to 6000 for a single plant (Cremer et al., 1991). Because the fruit wall is very decay-resistant, the fruits will easily become part of the subfossil record, even if they have germinated in the distant past. Such large numbers of seeds in archaeobotanical samples are sometimes interpreted as being indicative of economic use or even climatic events, as has been suggested, for example, for the fluctuation of seed numbers of Fat-hen (Chenopodium album) in samples from the site of Rojdi in north-western India. An increase in seed numbers during the Rojdi B period (2200–2000 BC) was interpreted as being the result of a short climatic change (Weber, 1992). It is more likely, however, that we are dealing with a normal fluctuation in seed numbers that have been retrieved from the local seed bank. Only if concentrations of seeds of this plant are found in, for example, storage facilities, can they be interpreted as food or seeds for sowing (Hansen et al., 2008; Kroll, 1998). Seeds that are dispersed in space are adapted to different dispersal agencies, so-called vectors—including water, wind, animals, and humans. It is these processes which are responsible for the transport of seeds and fruits from offsite localities to the settlement. This transport can be intentional, as is the case with, for example, the import of edible seeds and fruits. But seeds and fruits also arrive at a site unintentionally. This includes, for example, diaspores of arable field weeds that enter the settlement together with harvested crops. The dispersal of seeds in space may include several stages. The morphological and anatomical adaptations to seed dispersal are related to the primary dispersal vector. Seeds dispersed by the wind are light, and their transport is supported by wings and hairs (fig. 5). The fragile hairs and wings are mostly not preserved. Very small seeds, such as those of members of the Orchidaceae and the Orobanchaceae, are only seldom retrieved from archaeological contexts. The effectiveness of seed dispersal partly depends on the degree to which the pappus remains attached to the fruit. The pappus is the ring of hairs on the top of some fruits of the Asteraceae. Fruits of Thistle (Cirsium) species, for example, easily become detached from their pappus and are, therefore, not dispersed over large distances. Those of Sow-thistles (Sonchus), on the other hand, have a pappus that is tightly connected, and the dispersal potential of these fruits is much greater. Seeds that are adapted to water dispersal are relatively light because of the presence of air-filled tissues. Such seeds can float on the water surface for a considerable amount of time and can be moved by both wind and water currents (fig. 6). Because the soft outer tissue will easily be lost, subfossil plant remains may differ from the intact specimens (fig. 7).

1.2 Plant ecology 29

Figure 7: Whole fruit (top) and fruit without soft epicarp (bottom) of Branched bur-reed (Sparganium erectum).

Figure 6: Seeds of Yellow iris (Iris pseudacorus) are adapted to water dispersal because of their flat shape and light tissue.

Figure 8: Drift litter on the river side that is rich in plant remains. A fruit of Hazelnut (Corylus avellana) and a seed of Cucumber (Cucumis sativus) are visible, as well as stem fragments (Kop van Pannerden, the Netherlands; April 1992).

No special adaptations are necessary, however, when seeds are transported by running water. Seeds are transported submerged, and the dispersal capacity increases with the rate of flow. The analysis of drift litter material collected from beaches along the river Rhine and on the North Sea coast of the Netherlands shows that the dispersal potential of running water is considerable. A wide spectrum of diaspores as well as vegetative plant remains of both wild plants and economic plants was evidenced in the drift litter samples. The number of plant remains fluctuates, but it can be as much as 800 in a sample of 100 ml (figs. 8 and 9; Cappers, 1993). Seeds adapted to animal dispersal can be dispersed in two different ways. One pathway involves the consumption of the fruits, in which undigested seeds or fruit parts end up in the excrement. Seeds of the Twisted acacia ((Acacia tortilis), for example, have a firm seed coat that is resistant to 30 Handbook of plant palaeoecology | Cappers & Neef

Figure 9: Fruits of Sugar date (Balanites aegyptiacus), Zilla spinosa, and Neurada procumbens and droppings of sheep or goat in drift litter (Wadi Gimal, Egypt; 1995).

Figure 10: Dromedary browsing a Twisted acacia (Acacia tortilis) (left) and a dromedary dropping containing a digested fruit of which two seeds have germinated (right) (Eastern Desert, Egypt; February 1998 and January 1997, respectively).

Figure 11: Fruit of Sunflower (Helianthus annuus) transported by an ant (Yenisehir, Turkey; August 2010).

1.2 Plant ecology 31

Figure 12: Seeds collected by ants and stored in cavities. These cavities were hidden under a stone (north of Murtazaköy, Turkey; September 1998).

Figure 13: Spikelets of Wild barley (Hordeum vulgare ssp. spontaneum) collected by a Common vole around the entrance of a tunnel (Domuztepe, Turkey; August 2001).

digestion. When fruiting branches are eaten, fruits are swallowed as well, and the resistant seeds end up in the excrement, offering a safe site for germination (fig. 10). From the multiple fruits of Bramble (Rubus fruticosus) and Raspberry (R. idaeus) only the soft parts are digested; the one-seeded endocarps are dispersed with the excrement. The other pathway of seed dispersal by animals involves external transport. This kind of dispersal can be either unintentional or intentional. Unintentional transport by animals occurs when the diaspore remains attached to the skin. Such diaspores are covered with sticky glands, stiff hairs, or hooked spines. Intentional dispersal occurs when seeds are gathered for their nutritional value. Some plant species produce seeds that have a special nutritious tissue, for which reason they are intensively gathered by ants (fig. 11). Such an edible appendage can be an elaiosome or an aril, the latter being the outgrowth of the hilum or the funiculus. In this case, it is only the edible appendages that are eaten by the ants, and the seeds can germinate if the environmental conditions

32 Handbook of plant palaeoecology | Cappers & Neef

are adequate. The dispersal capacity of a whole colony of the Horse ant (Formica rufa), for example, has been estimated at 36,480 diaspores per year, assuming that ants are active 12 hours a day and can collect diaspores during 80 days a year (Sernander, cited in Bonn and Poschlod, 1998). Edible seeds can be buried in nest tunnels or fissures (figs. 12 and 13; Van der Pijl, 1969). Ants and field-dwelling rodents can thus be responsible for burying seeds in the soil. Depending on the species, seeds can be buried at considerable depths. Seed-granaries of ants have been observed at depths of more than 1 m in the profiles of trenches. The recovery of concentrations of viable seeds from archaeobotanical samples can thus be related to animal dispersal even when such concentrations are retrieved from deep layers. The second pathway of seed dispersal involves seeds that are dispersed in space. Such seeds are morphologically adapted to external dispersal vectors. The chances of such seeds ending up at a safe site, appropriate for germinating, are lower compared with the seeds that become buried near the parent plant. To compensate for this difference in survival prospects, seeds that are dispersed in space often outnumber those that are dispersed in time. Some plant species produce seeds that are partly dispersed in time and partly in space. In this way, two different pathways of seed dispersal are involved. Seeds without specific morphological features supporting dispersal by wind, water, or animals are deposited in the vicinity of the parent. Such seeds are capable of replacing the parent plant. Preconditions for a successful replacement are the removal of the parent plant and the breakdown of the dormancy of the buried seed. If the plant has a short life cycle, such a replacement can take place every year. It is also possible that a seed remains dormant for some years before it has the opportunity to germinate. The period of dormancy differs among plants, but for most plants it is relatively short. Some seeds, such as those of Rye grass (Lolium perenne), have a short lifespan in the soil; they are so-called transient seeds, and should germinate within a year. Other seeds can survive for more than a year and thus have the capacity of dispersal in time. Short-term persistent seeds have a longevity of fewer than 5 years in the soil, whereas long-term persistent seeds may survive for periods from 5 years up to 80 or 100 years. A good example are seeds of Fat hen (Chenopodium album), which can survive in the soil seed bank for several decades. A combination of both kinds of dispersal within the same plant species can be expressed by a difference in seed or fruit morphology. Lesser hawkbit (Leontodon saxatilis) demonstrates an example of such seed dimorphism (fig. 14). The combination of dispersal in time and space is a successful strategy of arable field weeds and is therefore of particular interest to studies dealing with the development of early agriculture. Seeds dispersed in time will be added to the seed bank of the field and will contribute to the established arable weed vegetation as long as the area is used as such. In addition, some seeds should end up in the seed stock that is used for sowing. As long as these seeds are sown in the same field, they still end up in the right place but cannot be considered to have been dispersed in space. However, when the seed stock is used to seed new fields, they are effectively dispersed in space by humans. Although such seeds still have to germinate and compete with crop plants and other wild plants, the wild plants involved have an ideal point of departure for becoming part of a specific crop vegetation. This adaptation will be further discussed in section 1.3.3.4.

1.2 Plant ecology 33

Figure 14: Flowering head of Lesser hawkbit (Leontodon saxatilis), with bracts enclosing the marginal fruits (left). The central fruits have already been dispersed. Centrally (middle) and marginally (right) positioned fruit, both depicted without pappus.

A reduced pappus and a rather plump shape characterize one row of marginally positioned fruits. These fruits are tightly enclosed by the bracts and will eventually be buried in the soil beneath the parent plant. The fruits positioned on the central part of the flowering disk are more slender and have a well-developed pappus. Being adapted to wind dispersal, they will be dispersed away from the parent plant.

1.2.3

Environmental conditions Successful germination, growth, and reproduction are only possible if a plant meets the proper environmental conditions. These environmental conditions comprise both biotic and abiotic requirements. Some plants have a symbiotic relationship with other organisms, facilitating, for example, the uptake of certain minerals. Other organisms can be involved in pollination and seed dispersal. If, for example, a pollinator of a particular plant is not present, such a plant can only propagate if vegetative reproduction is possible. Abiotic environmental conditions are related to climate and soil. The climate is characterized by temperature and precipitation, and both determine the availability of water. The soil structure is determined by the size of the soil granules, which affects not only the availability of water and air, but also the penetration of roots and the activity of soil organisms. The chemical composition deals with the mineral composition of the soil and groundwater and is thus related to the acidity, salinity, and water potential. The environmental requirements differ among plants. A plant with a small ecological range is adapted to specific environmental conditions. One with a broad ecological range is less demanding. The ecological range of a plant for a specific environmental factor is characterized by an optimum value and by values that mark the limits for this factor. If the optimum condition is met, a plant has the potential of reaching its maximum size. The ecological range can be determined for each environmental condition separately. Ellenberg et al. (1991) have published a synopsis of climatic and edaphic indicator values of many vascular plants. Although these values are based on measurements in central Europe, it has been shown that they are also valid for these plants growing in western Europe. Similar values have

34 Handbook of plant palaeoecology | Cappers & Neef

been published for southern Greece (Böhling et al., 2002), Poland (Zarzycki, 2002), and the Netherlands and Flanders (Runhaar et al., 1987 and 2004). The use of indicator values for the construction of ecological species groups will be dealt with in section 1.3.3.2. Environmental factors are considered to be stress factors if they are related to a shortage of light, water, or minerals; to extreme temperatures; or to restricted plant germination, growth, and propagation. However, plants may become adapted to such stress factors, and their ecological range may actually match these specific environmental conditions. The lower part of a salt marsh, for example, is regularly flooded with seawater, and only plants that can cope with this kind of inundation are capable of growing in such zones. Some types of vegetation have to deal with soil disturbance, during which part of or even the whole vegetation is destroyed. One could even argue that crop stands are subjected to such periodic disturbances, because some agricultural practices are aimed at the destruction of arable field weeds. Competition among plants occurs when individual plants have a simultaneous demand for specific environmental conditions. Such competition can be observed among specimens of the same plant species (intraspecific) and among specimens of different plant species (interspecific). In general, plants compete above ground for light and below ground for water and mineral nutrients. These environmental conditions are related to photosynthesis and the subsequent assimilation processes in which sugar is transformed into all kinds of organic molecules. Plants differ from animals and humans in that they are capable of making organic compounds from inorganic compounds. The inorganic molecules water (H2O) and carbon dioxide (CO2) are converted into a sugar molecule, glucose (C6H12O6). Oxygen (O2) is released into the atmosphere as a waste product. The energy that is necessary for this synthesis is obtained from sunlight. Because of the use of sunlight, this process is called photosynthesis. This differs from chemosynthesis, in which the energy is obtained from the oxidation of specific molecules. Chemosynthesis is used by certain microorganisms, mostly those living out of reach of sunlight, whereas photosynthesis is used by terrestrial as well as aquatic plants. Plants consist of a variety of compounds, such as vegetable oils, proteins, vitamins, and lignin. All these organic compounds are synthesized in subsequent biochemical processes, in which glucose is combined with specific elements. These elements are obtained from minerals that are dissolved in water and are absorbed through the roots. Carbon dioxide is present in the air and is absorbed through small openings (stomata) in the outermost tissue (epidermis) of the leaves and stem. Although its concentration in the air is a limiting factor in photosynthesis, carbon dioxide is accessible to all plants, and no real competition exists. Competition for light results in the arrangement of leaves (phyllotaxis) in such a way that all leaves can absorb sufficient amounts of sunlight. Plants that produce leaves that catch most of the sunlight may overshadow leaves of plants beneath. Especially when such plants grow fast and produce a dense foliage, such as Ground-elder ((Aegopodium podagraria), they become predominant in the vegetation. Competition for light is also strong in closed forests, where the tree canopy absorbs most of the light. Plants that grow on the forest floor

1.2 Plant ecology 35

either are shade-tolerant, that is, they have become adapted to low light densities, or complete their life cycle before the trees come into leaf. Competition for water is related to the availability of water. Three types of water sources can be distinguished: precipitation, soil moisture, and groundwater and atmospheric humidity. The amount of precipitation is determined by the climate, whereas the availability of water, being the ultimate condition for plant survival, is determined by different conditions. A shortage of water may thus be related to low precipitation, but environmental conditions may still be sufficient to enable plants to establish and survive. Competition will result in the establishment and reproduction of plants that are adapted to the environmental conditions. This kind of interaction can be observed in both natural and man-made vegetation types. Water availability is determined by differences in water potential and by extreme temperatures. The water potential is the potential energy of water per unit volume relative to pure water. Water has the tendency to move from an area with high water potential to an area with low potential. The water potential of an area lowers when, for example, solutes are added to the water. Because the concentration of solutes in the cells of a root is higher than that of the surrounding groundwater, groundwater will move into the cells of the root hairs. In salt marshes, the concentration of solutes in the groundwater is considerable. Only plants that are capable of lowering the water potential to a value that is more negative than that of the groundwater will still be able to take up water. This can be achieved, for example, by increasing the amount of solutes in the cells or by increasing the pressure in the cell (turgor) (fig. 15). Figure 15: Common sea-lavender (Limonium vulgare), with its purple-blue flowers, together with Sea purslane ((Atriplex portulacoides), with its whitish, mealy leaves, and fresh green specimens of Common glasswort ((Salicornia europaea) growing in the lower part of a salt marsh. Common sea-lavender is capable in lowering the water potential in its cells by regulating the concentration of solutes and the osmotic pressure (Schiermonnikoog, the Netherlands; August 2010).

36 Handbook of plant palaeoecology | Cappers & Neef

1.2.4

Water stress

1.2.4.1

Plant adaptations to water stress

Plants need water for their metabolic processes, including photosynthesis, and as a major compound of their tissues and a transport medium. Access to sufficient amounts of water may, however, be problematic. Plants have become adapted to this kind of water stress in different ways that include both morphological and physiological features. The adaptation of plants to water stress is diverse. Several principles can be distinguished, and specific combinations of such principles may find their expression in a particular plant species. Some climates, such as the maritime climate in north-western Europe, are characterized by precipitation throughout the year. However, the availability of water can become problematic when the temperature drops below zero. If the water columns in the transport vessels of plants become interrupted, they can no longer function. This happens when the groundwater is frozen while the evaporation of water through the stomata in the epidermis still continues. Adaptations to this kind of water shortage include the regulation of water evaporation, which can be realized by reducing the transpiration rate and by controlling the opening of the stomata. Plants that are capable of this kind of regulation are green yearround. Leaves of such evergreen plants can be reduced, such as the needleshaped leaves of members of the Pine family (Pinaceae). The abscission of leaves of evergreen plants is related to ageing. Because the leaves of these plants differ in age, the abscission of leaves is a continuous process and the plant remains green throughout the year. The presence of a thick waxy layer (cuticle) on top of the epidermis can also serve to reduce evaporation. This kind of adaptation can, for example, be seen in leaves of Ivy (Hedera helix) and Holly (Ilex aquifolium) (fig. 16).

Figure 16: Leaves of the evergreen plant Holly (Ilex aquifolium) are covered with a thick waxy cuticle (Hoogezand, the Netherlands; May 2009).

Another kind of adaptation to the unavailability of water due to freezing is the abscission of all leaves prior to the winter period. This process is triggered by the reduction of light intensity during autumn.

1.2 Plant ecology 37

The recovery of evergreen and deciduous plant species in palaeobotany can thus be used to make inferences about past climate. Although the pollen of Oak (Quercus) cannot be identified to the species level, it is possible to distinguish between those of evergreen and deciduous oaks. The pollen type of the evergreen oak in the Near East includes the Palestine oak (Q. calliprinos), the Evergreen oak (Q. ilex), and the Kermes oak (Q. coccifera) (fig. 17). The pollen type of the deciduous oak includes the Turkey oak (Q. cerris), the Tabor or Valonea oak (Q. ithaburensis), and the Pedunculate oak (Q. robur). Figure 17: The evergreen Kermes oak (Quercus coccifera) with fruits and waxy leaves (Geraki, Greece; July 2010).

Most plants have a C3-photosynthesis pathway, in which the first product of the CO2 fixation is an organic compound that exists of three carbon atoms (3-phosphoglycerate), hence its name. Simultaneously with the absorption of CO2, water will evaporate through the stomata. In fact, 97 per cent of the water that is absorbed by the roots will be lost in this way. In an environment where water is available in sufficient amounts, this kind of CO2 fixation is not a problem. But in areas with high temperatures, plants may close their stomata to reduce water loss. This in turn reduces the uptake of CO2 and subsequently also the synthesis of sugar (glucose). In areas with high temperatures, keeping open the stomata for the absorption of carbon dioxide may become problematic because too much water may be lost through evaporation. Plants have two alternative photosynthesis pathways as an adaptation to this possible loss: C4photosynthesis and CAM-photosynthesis. Plants with C4-photosynthesis and CAM-photosynthesis are characteristic of areas below 40°N latitude. CO2 fixation takes place under high light intensity and high temperatures and is more efficient with these alternative synthetic pathways. In both these alternative pathways, CO2 is fixed twice. The first fixation converts CO2 into an organic compound consisting of four carbon atoms (malate or aspartate). C4-plants take up CO2 during the day. It is converted to malate, which is then stored in large concentrations in inner cells around vascular bundles. Because the enzyme that is involved in the first CO2 fixation is very efficient, it is not necessary for the plant to keep the stomata open for a long period.

38 Handbook of plant palaeoecology | Cappers & Neef

Plants with a CAM-photosynthesis take up CO2 at night. In this way, the evaporation of water is minimized and the plant can even photosynthesize under drought conditions. A disadvantage of this metabolism is that the growth rate is extremely low. CAM is an abbreviation of Crassulacean acid metabolism; it takes its name from the Crassulaceae family, in which this type of photosynthesis was first demonstrated. Crop plants that were domesticated in the Near East, such as Barley (Hordeum vulgare) and the Wheat species (Triticum spp.), and in China, such as Rice (Oryza sativa), have a C3-photosynthesis pathway. Examples of crop plants with a C4-photosynthesis are maize (Zea mays), millet (Pennisetum spp.), sorghum (Sorghum bicolor), and sugar cane (Saccharum officinarum). An example of a crop plant with CAM-photosynthesis is the Pineapple ((Ananas comosus) (fig. 18).

Figure 18: Infructescences produced by a C3-plant (left: Barley; Hordeum vulgare); a C4-plant (middle: Maize; Zea mays); and a CAM-plant (right: Pineapple; Ananas comosus).

Some climates are characterized by precipitation that is limited to a particular part of the year. Areas with a Mediterranean climate, for example, receive most of their rainfall between October and April. Annual plants are adapted to this cycle by germinating when the autumn and winter rains start. Such winter annuals complete their life cycle before the summer drought starts. Cereals and pulses that have been domesticated in the Near East are therefore winter crops. The spread of agriculture towards regions with a continental and maritime climate has resulted in the development of crops that became adapted to the availability of water during the summer period. In addition to winter crops, summer crops became available. Although separate plant communities have been distinguished for the regions where both summer and winter crops are cultivated, this difference does not seem to be related to a difference in moisture regime. Moreover, a recent analysis of phytosociological descriptions of crop stands (relevés) has revealed that a distinction between arable field weeds associated with summer crops and winter crops does not exist (Schaminée et al., 1998).

1.2 Plant ecology 39

1.2.4.2

Plant growth and extreme water stress

The plant growth in the Eastern Desert of Egypt may serve as an example for the use of each of the three water sources (precipitation, groundwater and atmospheric humidity). This desert is characterized as hyperarid, with a mild winter and a hot summer, with average temperatures in the hottest month ranging from 33–40°C. The mean annual rainfall in the Red Sea coastal area, including the coastal plain and the Red Sea mountains, ranges from 25 mm in Suez (29°59' N), to 4 mm in Hurghada (27°14' N), to 3.4 mm in Quseir (26°07' N) (Zahran and Willis, 1992). Nevertheless, the Eastern Desert has a considerable and rich flora, which can be explained by the availability of sufficient water sources on the one hand and the adaptations of desert plants to water stress in particular on the other hand (fig. 19). The mean annual precipitation alone may, therefore, be misleading.

Figure 19: Shortage of water determines the distance between individual plants of Indigofera in Wadi Umm Athel (Eastern Desert, Egypt; February 1998).

Rainfall in the Eastern Desert is scanty, unpredictable, and concentrated in the winter period. The availability of rainwater is particularly determined by the almost-barren orographic relief. The barren mountains affects principally the redistribution of rainfall. As a result, the water supply of wadi branches is actually much more than what is calculated based on the mean annual precipitation. This concentration of water is also supported by the presence of impervious surface crusts, which are found in hyperarid regions in particular and are virtually water resistant. Short, heavy showers may produce huge quantities of water, but they are quite rare. Mostly, such cloudbursts affect only limited areas. After such heavy rains, silty, flat depressions are filled for weeks or even months with water and can be used for cultivating sorghum (Sorghum bicolor). Groundwater will be drained by an underflow towards the Red Sea or the river Nile. The persistence of this water flow is demonstrated by the permanent presence of slow-running water in a narrow gorge of Wadi Shenshef, even when preceding winter rains have not been recorded (fig. 20).

40 Handbook of plant palaeoecology | Cappers & Neef

Figure 20: Surface water in Wadi Shenshef near a narrow gorge west of the settlement (Egypt; January 1996).

Especially in the period after heavy cloudbursts, groundwater may be available on the surface of wadis. Annual plants with a short life cycle will use this shallow groundwater, which is available only for a relatively short period (fig. 21). Figure 21: Coastal plain along the Red Sea coast with green, fresh vegetation. After sporadic winter rains, water drains partly above ground and partly underground from the mountains to the Nile Valley and the Red Sea. The herbaceous vegetation marks the sandy wadis that have shallow groundwater flow (Wadi Kintsiurub, Egypt; January 1997).

Water that penetrates deep into the wadi escapes evaporation and will be available for plants with deep taproots (figs. 22 and 23). The amount of water that can sink into the ground will largely depend on the texture of the soil. Penetration and the coarseness of the soil texture are positively correlated. For this reason, sandy soils offer better water supplies in dry regions than do clay soils. Silty depressions are an exceptional case, because the concentration of the water may compensate for the high field capacity, which slows down water penetration.

1.2 Plant ecology 41

Figure 22: Scattered trees of Twisted acacia (Acacia tortilis) on the coastal plain of the Red Sea. These trees produce long taproots that penetrate up to 40 m into the soil to reach the deep groundwater. This water source makes them independent from local rain showers (Egypt; March 1998).

Figure 23: Surrounded by paving stones and asphalt, this Fig tree (Ficus) has become dependent on leaking water pipes (Cairo, Egypt; February 2003).

42 Handbook of plant palaeoecology | Cappers & Neef

The third source of water in a desert environment can be the atmospheric humidity. This water supply is an important source of water for plants growing in the Red Sea coastal zone. Evaporation of seawater during daytime replenishes the water content of the air. This water partly condenses again if the temperature of the air drops sufficiently at night during winter and early spring. Considering the fact that air temperature is lowest just before sunrise, morning dew (nada) may therefore be responsible for damp soils. Unlike precipitation, morning dew is a much more steady water source for plants. Referring to agricultural practices in Bactria, Africa, and Cyrene, Pliny (NH 18.50.186) even states that crops depend on the dewfall at night for their nourishment. Theophrastus (EIP: 4.3.7 and 8.6.6) also emphasizes the contribution of dew to plant growth in regions with no rainfall, including Egypt. According to Jabbur (1995:47), who studied Bedouin life in the Syrian desert, there is so much dew on plants in spring that some women collect the water in their waterskins. This atmospheric humidity is of value both on the flat coastal plain adjacent to the Red Sea proper and in the mountains, which intercept the clouds blown inland. High mountain massifs in particular may receive considerable amounts of rain in this way. This also explains the presence of several mosses and ferns on the higher levels of the Gebel Elba (fig. 24). Fertilization of these plants is only possibly under moist-to-wet conditions. Figure 24: A thalloid, fruiting liverwort, on the upper part of the north-eastern slope of the Gebel Elba (north-eastern Sudan; February 1999).

Some plants, such as the Nile tamarisk (Tamarix nilotica), are capable of initiating condensation of atmospheric moisture by means of special hygroscopic salt crystals under conditions of high humidity (fig. 25). For this reason, drops of water on the surface of the Nile tamarisk taste rather salty. The water condensation is triggered when the humidity of the air exceeds 76 per cent. Like dew deposition on the plant surface, which depends on a lower plant surface temperature in relation to the dew point temperature of the adjacent air, this salty water drips from the plant and can be absorbed by the shallow roots. The effectiveness of this self-induced sprinkling is evidenced by the many imprints of raindrops in the damp soil beneath such excretive plants.

1.2 Plant ecology 43

Figure 25: Nile tamarisk (Tamarix nilotica) in the salt marsh of the Red Sea covered by morning dew (Egypt; February 1996).

The availability of water is, however, no guarantee for the presence of desert plants. This is especially true of regions that do not profit from the dewfall. Here, the irregularity of rainfall is at odds with the presence and longevity of seeds in the soil. Seed banks in desert soils are characterized by a concentration of seeds in the upper 2 cm of soil, with a high degree of spatial heterogeneity and great seasonal and annual variability. Seeds tend to be concentrated in depressions where water collects and in wind shadows, such as below established plant species (Kemp 1989). Because dispersal distances are generally short, a specific type of ephemeral vegetation may exist for several years (Zahran and Willis 1992). Isolated spots that are completely dependent on unpredictable showers may lack a viable seed bank, and consequently no vegetation will develop after rainfall (compare figs. 26 and 27). Figure 26: Small lake in the Western Desert along the road between Luxor and the Kharga Oasis. Vegetation is not developed due to the lack of a seed bank (Egypt; November 1996).

44 Handbook of plant palaeoecology | Cappers & Neef

Figure 27: Wadi Shenshef in flower after a serious cloudbursts in November. The transition zone between the wadi and the terrace is dominated by Rumex vesicarius. On the terrace, a Roman building from the fourth to fifth centuries AD is visible (Egypt; December 1996).

Plant adaptations in areas with a desert climate are diverse. One possible adaptation that has arisen by natural selection is the storage of water in leaves, stems, or roots. Such plants are called succulents and are adapted to arid habitats as well as to saline environments. Succulent plants that grow in the Eastern Desert of Egypt, including the environment of Berenike, are Suaeda monoica, Halopeplis perfoliata, and Aizoon canariense. A good example of a halophytic succulent is Zygophyllum coccineum, which has swollen stems and leaves and is able to abort part of its leaves and even young shoots to prevent excessive water loss. Other plants have leaves with xeromorphic features. They are called xerophytes and are by definition adapted to arid regions. A common feature is the reduction in the size and/or number of leaves, as can be observed in Crotalaria aegyptiaca and Moringa peregrina. An extreme example of leaf reduction is demonstrated by Tamarix aphylla, a shrub that has only sheathlike leaves. Another method of drought survival is shortening the life cycle. Annual plants that take advantage of the availability of water independent of the season—either a rain shower or a damp period caused by dewfall—are called ephemerals. They include succulents, such as Zygophyllum simplex (fig. 28), and non-succulents, including grasses such as Aristida funiculata and nongrasses such as Lotononis platycarpa. Such plants can survive by quickly absorbing the rainwater and dew. Some of them have reasonably long roots, such as species of Aristida, whereas other plant species have only very small and shallow root systems, such as stunted but flowering specimens of Triraphis pumilio, Eragrostis ciliaris, and Coelachyrum brevifolium. These plants are frequently found in the environment of Berenike and Shenshef. Some plants may develop temporary so-called rain or extension roots, which grow in the upper part of the soil and absorb the water from a rain shower or dew (Girgis, 1971).

1.2 Plant ecology 45

Figure 28: Prostrategrowing specimen of Zygophyllum simplex (Eastern Desert, Egypt; January 1997).

Perennials may also behave as ephemerals. Conversely, some species, such as Zygophyllum simplex, that normally behave as ephemerals can prolong their lifespan under favourable conditions (Zahran and Willis 1992). They are as resistant to drought as a seed. Plants that obtain their water from the groundwater directly or through the capillary fringe are called freatophytes. Because in most cases groundwater is only available at considerable depths, freatophytes are predominantly perennial species that are able to develop deeply penetrating taproots. Regeneration of freatophytes will be especially successful during years with heavy showers, resulting in a moist soil from the surface down to the water table over a considerable time span. During this period, long taproots must grow to the water table. As soon as this is reached, the plants will have become independent of the unpredictable rainfall and can survive in an apparently unfavourable environment. A good example of a desert freatophyte is the Twisted acacia ((Acacia tortilis), which may have a taproot of more than 40 m. Owing to its deep, penetrating root, this tree is capable of growing on the fringe of wadi branches. Its presence is even indicative of the many wadi branches that dissect the flat coastal plains along the Red Sea. The growth of the root starts soon after germination. A seedling under a full-grown tree in Wadi Shenshef of only 2 cm high could be dug out to a length of 25 cm before its fragile root broke off. Examples of herbaceous species with long roots are Aerva javanica and Aizoon canariense.

46 Handbook of plant palaeoecology | Cappers & Neef

1.2.5

Agricultural practices The production of crops includes a variety of activities that are partly conducted in the fields and partly in the settlement. Some of these activities deal with the preparation of the soil and the manipulation of the field vegetation, while others deal with the processing of the crops. The agricultural practices are primarily aimed at optimizing the yield. This implies that the environmental conditions in the field are optimized in favour of the crop plants. Agricultural practices that are directly connected with the improvement of the crop yield are manuring, water management, and weed control. Agricultural practices that are applied may differ in both character and sequence, depending on the kind of crop, environmental conditions, and technical innovations. Manuring, for example, is not necessary if the soil is enriched with minerals from silt that is deposited during irrigation. In addition, ploughing, harrowing, and irrigating are not always practised. This section deals with agricultural practices related to crop cultivation. Because weed control is related to several of these practices, it is dealt with separately in the last section.

1.2.5.1

Manuring

Plants need some 16 elements in their metabolism. Although plants absorb a large variety of minerals, the amount that is needed of each of them differs considerably. Nitrogen (N), phosphorus (P), and potassium (K) are used in relatively large quantities and are classified as primary macronutrients. The other minerals concern three secondary macronutrients, namely, calcium (Ca), sulfur (S), and magnesium (Mg), and several micronutrients (trace minerals), of which only small amounts are necessary. Carbon and oxygen are absorbed as carbon dioxide (CO2) and oxygen (O2) from the air, whereas hydrogen is absorbed as water (H2O) by the roots from the soil. The most important minerals in the soil are thus nitrogen, phosphorus, and potassium. Nitrogen is found in many organic compounds, such as proteins (including enzymes) and nucleic acids (DNA and RNA). Phosphorus is also used for making essential compounds, such as nucleic acids and ATP (the latter used for the transfer of energy) and is a compound of cell membranes. Potassium is essential for the functioning of cells. The loss of mineral nutrients from the soil can take place through leaching, soil erosion, and the removal of part of the vegetation. The vegetation is removed by harvesting crops, with minerals being present in both the cultivated crop and the weed plants; by hay-making; and by grazing. When grazing is practised, minerals that are removed become incorporated into the tissues of animals. Removal of minerals by living animals occurs, for example, with milking, shearing, and the collection of dung as a source of fuel. Dung that is deposited on the field proper or is collected as a source of fuel will result in a depletion of the soil. The ploughing of the stubble and field weeds will reduce the loss of minerals to some extent. Burning the remaining vegetation may also accelerate the availability of nutrients but causes some losses by wind erosion as well.

1.2 Plant ecology 47

It depends on the nutrient availability and the uptake by the crop how many yields can be produced before the concentration drops below a critical point. The minerals that are most used in the plant’s metabolism will be the first ones to signal a nutrient deficiency. A loss of minerals will affect the concentration of macronutrients in particular. The replenishment of nitrogen is partly realized by certain bacteria that have the capacity of converting atmospheric nitrogen (N2) into several inorganic nitrogen compounds. These bacteria partly are free-living and partly form symbiotic associations with some plants, such as some members of the Fabaceae family, including pulses and clovers (Trifolium spp.). The fixed nitrogen is released into the soil when the symbiotic plants die. The increase of the soil nitrogen level will be thus be achieved by ploughing legumes into the soil. Owing to this capability of fixing nitrogen, legumes are considered natural fertilizers and can grow in poor soils. This kind of green manuring was already practised in antiquity, even though those practising it did not understand why it improved the soil. Theophrastus, for example, mentions in his Enquiry into plants (8.7.9) that beans, contrary to wheat, do not exhaust the ground, but reinvigorate it: ‘Beans, as was said, are in other ways not a burdensome crop to the ground, they even seem to manure it, because the plant is of loose growth and rots easily; wherefore the people of Macedonia and Thessaly turn over the ground when it is in flower’. Cato explicitly mentions lupines, beans, and vetch as crops that fertilize fields (On Agriculture: 37.2).This method of green manuring is still applied and improves both the mineral content and the structure of the soil. The replenishment of minerals can be achieved by letting farmland lie fallow. This is a strategy of both mobile farmers, those practicing so-called shifting cultivation, and sedentary farmers. Abandoned fields can recover from the depletion by the development of pioneer vegetation and by input via water and wind. It is also possible to fertilize the depleted fields by adding mineral-rich compounds. Fertilizing with dung is possible, although compensation for loss of minerals due to removing harvests from the field is only effective when this dung is collected from elsewhere. Dung produced by animals that are allowed to graze stubble fields does not compensate for the depletion of minerals by harvesting; it only accelerates the mineralization. Dung from cattle and pigeons has a long tradition as a fertilizer (fig. 29). Kitchen waste can also be used for manuring the fields (figs. 30, 31, and 32).

48 Handbook of plant palaeoecology | Cappers & Neef

Figure 29: Pigeon dung can be collected in considerable quantities from a dovecot (Fayum, Egypt; November 2009).

Figure 30: A donkey is used for the transport of stable dung to the field (Fayum, Egypt; November 2007).

1.2 Plant ecology 49

Figure 31: Mixture of dung and kitchen waste, including leftovers of Onions (Allium cepa), used as fertilizer (Fayum, Egypt; November 2007).

Figure 32: Heaps of stable dung ready for distribution over the field (Ezbert Dawud, Egypt; November 2007).

Flooding can contribute to the replenishment of minerals as well, especially when the water is not moving for a while and a mineral-rich layer of silt can thus be deposited on the fields. In Egypt, this kind of natural manuring was possible until the construction of the Aswan High Dam, which prevented the flooding of the Nile. In the Egyptian oases, irrigation water is obtained from groundwater that cannot compensate for the depletion of minerals. Here, mineral-rich soil is used instead. This can be sandy soil (marug) or a sediment of bentonite clay, silt, and sand (tafla), both of which are rich in minerals (fig. 33).

50 Handbook of plant palaeoecology | Cappers & Neef

Figure 33: Heaps of purple-coloured tafla ready to be spread over the nearby fields (Kharga oasis, Egypt; June 2011).

With an increase in population, fertilizers can become a limiting factor in the use of agricultural land. This has led to the exploitation of mineral-rich deposits from ancient settlements. In the Netherlands, for example, this kind of exploitation started in the eighteenth century. Originally intended for personal use and conducted on a relatively small scale, the digging of mineral-rich soils from terp (mound) settlements developed into a largescale and commercial enterprise from the middle of the nineteenth century onwards. This rich terp soil was not only used to fertilize depleted cultivated land but also to reclaim waste land with poor soils, such as heathlands. A deposit 3 m thick taken from 1 ha of a terp was sufficient for the reclamation of 375 ha of wasteland. A similar development took place in other countries. In Egypt, for example organic-rich soil from old sites (Sibakh kufri) became a substantial supplement to the traditionally used stable dung (Sibakh baladi). The removal of soil from ancient settlements started in the late nineteenth century and continued until the early twentieth century. In the Greco-Roman settlement of Karanis, a special railway was built, which could transport c. 200 m3 sediment each day. Because this digging was conducted under official licence, the company was able to continue the removal of soil even once the excavation by the University of Michigan had started. In fact, this exploitation of the centre of Karanis continued during the seven years that the excavations were conducted. To minimize further destruction of the archaeological record, the excavation was confined to the central part of the town. Sediment layers rich in organic content were collected for the company to prevent the company from demolishing buildings (figs. 34 and 35). It goes without saying that this kind of commercial digging has resulted in the large-scale destruction of the rich archaeobotanical archive, among other things. The destruction of ancient settlements gradually stopped when artificial fertilizers came onto the market. Although such fertilizers had already been developed by the middle of the nineteenth century, it was only after the middle of the twentieth century that they became available on a large scale. With this development, it was no longer the soil fertility but the availability of water that was the limiting factor—again.

1.2 Plant ecology 51

Figure 34: A gaping hole in the centre of Karanis due to the removal of mud bricks and organic deposits is visible in the background (Egypt; November 2007; looking south-east).

Figure 35: Clumps of organic sediment unearthed from trench 23 at Karanis. This kind of sediment was sought after by the Sibakhin as an agricultural fertilizer (Egypt; November 2010).

1.2.5.2

Sowing and planting

Seeds can be sown directly on the field or in seed trays. If seeds are scattered over the field, it is important that seeds are not secondarily dispersed by the wind or eaten by birds. Wind dispersal can be prevented by sowing wet seeds, so that they stick to the soil. If seeds have been soaked in water for a while, the germinating process is initiated and seeds will quickly produce roots to fix them into the soil. Seed predation can be reduced by harrowing or ploughing the field after sowing. If the seed is sown in seed trays, it is the seedlings that are planted in the field, and seed loss by predation is limited (figs. 36 and 37). Both sowing and planting can be done randomly or in rows. Because weeding is possible throughout the growing season if row sowing is practised, the weed vegetation that can develop in such fields differs from the vegetation that develops in crops that have been sown by broadcasting.

52 Handbook of plant palaeoecology | Cappers & Neef

Figure 36: Broadcasted seeds of Alexandrian clover (Trifolium alexandrinum). To improve the seed–soil contact, seeds have been soaked in water prior to sowing (Ezbet Dawud, Egypt; November 2009).

Figure 37: Planting Tomato (Solanum lycopersicum) seedlings in irrigated fields (Saqqara, Egypt; February 2010).

If broadcast sowing is applied, seeds will be scattered over the soil, and it is the experience of the farmer that minimizes the clustering of the seeds. If seeds fall too close to each other, competition will result in self-thinning, and a proportion of the seeds will thus not contribute to the next yield. This kind of competition also occurs with crops whose diaspores contain two or more seeds. Using spikelets of Emmer wheat (Triticum turgidum ssp. dicoccon) for sowing results in competition between two seedlings. And fruit clusters of Beet (Beta vulgaris) contain on average three to five germs, resulting in a close competition for both water and light. The number of seeds that germinates is partly regulated by a number of chemical compounds in the seed, resulting in a delay of germination of some seeds for several years (fig. 38).

1.2 Plant ecology 53

Figure 38: A farmer shows the mixture of Beet seeds (Beta vulgaris) and an artificial fertilizer he is distributing over his field (Ar Rub’, Egypt; November 2008).

Sowing Emmer wheat can be done with dehusked grain kernels to avoid competition and to optimize the sowing ratio. Recently, alternative approaches have been developed to overcome the disadvantage of using multigerm clusters for sowing. One method is to break up the clusters into fragments that contain one fruit each. Another possibility is to manipulate the number of fruits in a cluster by genetic engineering. Sowing spiny diaspores, such as those of Carrot ((Daucus carota), is also problematic because such diaspores stick together. One possible solution is to remove the spines of the seeds used for sowing. Today, such diaspores are pelletized, so that the removal of spines is no longer necessary. The coating of such diaspores can also contain chemicals that regulate the germination and even pesticides that protect the seedling against diseases and fertilizers. Crops that have small seeds can be broadcasted more evenly by mixing the seeds with sand.

1.2.5.3

Hoeing, ploughing, and harrowing

Ploughing the field loosens the soil. This allows the roots to grow easily into the ground. It also allows water to penetrate more easily into deeper soil levels and aerates the soil. In this way, more oxygen becomes accessible, which is necessary for the active water uptake by the roots and for the soil organisms that are partly responsible for the breakdown of organic residues. If the soil is turned, remains of the previous crop and weed plants are buried. Burying the dung and the sown seeds can also be achieved by ploughing. It is also possible with ploughing to bring a mineral-rich layer to the surface. This kind of improvement of the nutrient availability is only possible if such a buried layer is present and a plough is used that can reach this layer. In the initial period of agriculture, the soil was treated with simple tools, such as a piece of antler or a bent piece of wood used as a hoe (fig. 39). These were replaced initially by the ard, which later developed into the plough. The ard has a share that can be drawn forward into the ground (fig. 40). With a plough it is possible to turn the soil that is cut by the share (fig. 41). In many ploughs, a vertically placed coulter in front of the share cuts the soil. An oblique mouldboard is placed behind the coulter and turns the soil. If furrows are ploughed close to each other, the tilled soil can be used to fill the adjacent furrow (fig. 42).

54 Handbook of plant palaeoecology | Cappers & Neef

Figure 39: Wooden hoe (length: 56 cm). The metal head is no longer present (unknown provenance).

Figure 40: An ard with a horizontal sole, vertical stilt, and straight beam used for yoking cattle. The sole-head is covered with a metal blade. The angle between sole and beam can be adjusted, thus allowing both shallow and deep ploughing (Ezbet Dawud, Egypt; November 2006).

Figure 41: The mouldboard and sole of this plough from Indonesia are made from a single piece of wood. The metal ploughshare, strengthening the sole head, has been removed.

1.2 Plant ecology 55

Figure 42: Two cows pull a plough similar to the specimen depicted in figure 40. The straight yoke fits in front of the withers, which have a slight hump. The share is driven into the soil by pushing on the stilt (Fayum, Egypt; November 2008).

Both ard and plough can be pulled by animals. Especially when the soil is covered with vegetation or contains much clay, much power is needed. Oxen and horses are stronger than cows. Nevertheless, cows are used because ploughing is limited to a short period and cows produce milk. Animals used for ploughing need a high quality fodder such as barley grains. The economic value of a domestic animal capable of pulling the plough is expressed poignantly by the Egyptian architect Hassan Fathy (2000: 92), who promoted the use of mud brick in Egyptian architecture: “The peasant family’s whole life depends upon one or two cows and an acre or so of soil. If the cow dies or the crop fails, the family must starve, for there is no insurance scheme to save him, no dole, no benevolent governmental soup kitchen.… He will never taste a green vegetable because all his land is growing cash crops. He is in the grip of creeping famine; although the Nile will never fail and the crops are always assured, this in Egypt, where twenty-seven people live off every six acres of farmland, merely assures the peasant of the same inadequate nourishment as last year. To hold on to even his present miserable living standard he must treasure every last leaf and grain of saleable crop and treat his cows as jealously and tenderly as his children—more so, in fact, for he will say that if a baby dies he can make plenty more free, but if a cow dies he must pay to replace it.” Excavations may reveal ploughing patterns. A precondition for the preservation of such patterns is that holes and furrows are filled with different sediment. The use of a hoe was recognized, for example, at Swifterbant, in the central part of the Netherlands (fig. 43). This field, dated to 4300–4000 BC, is the earliest evidence of this kind of tool in north-western Europe. An ard may produce a crisscross pattern, whereas a plough produces a pattern characterized by parallel strips.

56 Handbook of plant palaeoecology | Cappers & Neef

that is produced by an ard may show a crisscross pattern whereas the pattern made with a plough is characterized by parallel strips.

Figure 43: Hoe marks seen from above (top) and in profile (bottom) (Swifterbant, the Netherlands).

Turning the soil with an ard or plough may result in a rather uneven surface of the topsoil. Harrowing can be applied to crumble the soil and to level the surface (fig. 44). This improves the soil structure and increases the water capacity. Harrowing is also used for burying the seeds. Levelling the surface is also important if fields are flooded and water has to be distributed equally over the parcels. Both dry levelling and wet levelling can be applied. The implement used for levelling the soil before irrigation should be able to smooth the soil surface to a high degree (fig. 45).

Figure 44: Levelling with a zahhafa. A trunk of a date palm (Phoenix dactylifera) is used for levelling the soil prior to sowing. The farmer stands on the beam to make the implement heavier (Sinnuris, Egypt; November 2008).

1.2 Plant ecology 57

Figure 45: Levelling with a kasabiya. A wooden box that is open in front is used to scrape off the mounds and to fill the hollows. This implement is used prior to irrigation (Fayum, Egypt; November 2009).

1.2.5.4

Irrigation

As a rule of thumb, it can be stated that rain-fed agriculture is possible if annual rainfall is 300 mm or more. This amount of rainfall is also sufficient for the development of woodlands. If rainfall is 250–300 mm per annum, it is still sufficient for the production of crops, but an additional water supply may be necessary depending on the local hydrological conditions and the kinds of crops that are cultivated. In areas where the mean annual rainfall drops below 250 mm per annum, irrigation is always necessary. For lowland Iraq, Thalen (1979) calculated an absolute minimum requirement of 250 mm precipitation per annum for stable rain-fed agriculture, a requirement which should not be taken too literally; local factors, such as the capacity of the soil to retain water, and dew, which can affect the microclimate, can also contribute. Another aspect that can have an impact is the distribution of rainfall throughout the year. For Syria, Wirth (1971) mentions that on a welldeveloped soil with a favourable distribution of precipitation during the winter months, barley still grows well in areas with 200 mm precipitation and wheat in those with 250 mm. Heavy showers can release more than 50 mm of rainfall within an hour, whereby much of the water is just runoff, sometimes causing severe soil erosion on the slopes. The water available for irrigation is often limited, and its distribution has to be arranged by establishing special rules. The amount of water available to each farmer can be determined in several ways. In the Moroccan oasis of Ouijjane, for example, located south-east of Tiznit, the movement of the sun is used to determine the distribution of the available spring water, which is collected in a large basin. Perpendicular to a wall a number of stones are buried in a straight line (fig. 46: top). During the time that it takes for the shadow of the wall to move between two stones, a farmer is allowed to use water for irrigation. The distance between the stones varies and corresponds to the size of the arable fields of the farmers who are authorized to use the water. On an overcast day, sticks with special markers are used instead (fig. 46: bottom). Like the stones, the marks have fixed distances. In ancient Egypt, the quantity of available water from the Nile could be determined by nilometres. These nilometres were constructed at various locations along the river. By measuring the level of the Nile during the flood season, an estimate could be made of the expected harvest, and thus the tax could be determined as well.

58 Handbook of plant palaeoecology | Cappers & Neef

Figure 46: Wall and row of stones (top) and stick with marks (bottom) used for determining the length of time allotted for irrigation (Oase de Ouijjane, Morocco; April 2007).

The application of irrigation is related not only to the available water, but also to technical applications, such as the construction of a system of canal irrigation and the development of special equipment for lifting the water. Such applications are partly adapted to local conditions and may vary between regions. Agriculture in Egypt was initially dependent on the natural flooding of the Nile. This took place every year over a period of three months, from August to October. Due to extensive rainfall in the source area, the Nile flowed over the natural levees and inundated the fields. The flood water was captured in depressions embedded by natural levees or artificial dams (fig. 47). Such basins had an inlet opening upstream and an outlet opening downstream. When the water level of the Nile was high enough, the inlet was opened and the basin filled up with water. When the basin was filled with a sufficient amount of water, the opening was blocked and the water allowed to get absorbed into the soil. Next, the drain hole was opened and the remaining water could flow back into the Nile, taking with it also the salt 1.2 Plant ecology 59

that had accumulated on the soil surface. This form of basin irrigation was applied until the construction of the Aswan High Dam, which prevented the flooding of the Nile. Figure 47: Remnant of the mud brick wall between Itsa and Shidmuh that once enclosed an artificial lake in the south-east of the Fayum used for basin irrigation. It is assumed that this artifical basin was made in the beginning of the Ptolemaic period (Itsa, Egypt; May 2004).

In order to ensure that the water flows evenly over a field, the field can be divided into small plots by earthen walls. Each compartment is individually irrigated so that the water retains enough strength to reach the entire field, even if the ground is not completely levelled. The water can flow into the first plot after an opening has been made in this part of the dam. If sufficient water has entered into the first plot, the dam adjacent to the second compartment is opened. Only then is a dam created across the main channel, blocking the flooding of the first plot. In this way, the whole field is irrigated (fig. 48). Figure 48: Irrigation of compartmented field. The fifth parcel is being inundated (Ezbet Dawud, Egypt; November 2010).

60 Handbook of plant palaeoecology | Cappers & Neef

To irrigate fields that cannot be inundated by manipulating the flooding, different kinds of devices can be used to lift the water artificially. For example, in Egypt, probably during the second half of the eighteenth dynasty (Amarna period), the shaduf became in use. With this device, a bucket of water can be lifted without much effort by using a lever. In Egypt, wood from the Acacia and Sycomore fig (Ficus sycomorus) was used for its construction. With a shaduf shaduf, water can be lifted to a height of 2.5 m, and a farmer could raise 3 m3 per hour as an average rate during a full working day. In the Ptolemaic period, the water wheel was developed. These water wheels have compartments that lift up the water. If these compartments are constructed at the rim, the water is tipped out at the top of the wheel’s revolution. Terracotta pots could be fixed to the rim for catching the water, hence its Arabic name, sāqiya. Water wheels can be driven by draught animals or by water. They can be water-driven only if the water wheel is not too heavy and water flows at a fairly high velocity. The lower part of such wheels is submerged and turned by the running water. Water is collected in the many compartments of the outer rim and released at the top (fig. 49). Figure 49: Water-driven wheel with compartmented rim, made entirely from wood (village east of El Mandara, Egypt; November 2008).

The lifting capacity of these wheels depends on the diameter of the wheel, the capacity of the buckets and the kind of draught animal or the velocity of the stream. A total volume of 20–22 m3 of water can be lifted up to 1.5 m each hour; the volume is reduced to 8–10 m3 if lifted up to 9 m. The lifting capacity is reduced when the water is suspended near the hub, as is the case with the spiral-scoop water wheel (fig. 50). This double-sided metal water wheel has volute-shaped compartments in which the water is lifted. This kind of lifting device is popular when only a low lift is required.

1.2 Plant ecology 61

Figure 50: Compartmented metal water wheel driven by a blindfolded donkey. The energy comes from Alexandrian clover (Trifolium alexandrinum), which is cultivated in the fields visible in the background (Sinnuris, Egypt; November 2008).

This improvement in irrigation by using lifting devices increased the yield because land located on a higher level could be reclaimed and because two crops a year could be produced (Butzer, 1976). It is thus with these lifting devices that perennial irrigation was established in Egypt and that, in addition to winter crops, summer crops could be cultivated. Although water wheels are still used in Egypt, they are increasingly replaced by mobile diesel engines that are transported to the fields by donkeys (fig. 51). Figure 51: Irrigation powered by a diesel engine. The metal frame enables it to be transported by a donkey. The metal water wheel is still used by farmers who cannot afford an engine (c. 5 km west of El Fayum, Egypt; November 2010).

Irrigation can be inferred from both the established vegetation and the seed bank. Indicator species in the established vegetation are tolerant of inundations and are good competitors in wet soils. An example of such a field weed is Purslane (Portulaca oleracea). This herbaceous plant can grow up to 50 cm high, but it is also able to grow horizontally over the soil surface (fig. 52). The plant has somewhat succulent leaves and is eaten as a vegetable. Purslane grows on moist to wet soils, such as spots where spilled water

62 Handbook of plant palaeoecology | Cappers & Neef

collects, in gullies where rainwater accumulates along roadsides, and in fields that are irrigated. In irrigated fields, Purslane can predominate in the lower vegetation zone. Seeds of this plant may end up in the dung of animals that graze on fields after harvesting. When this dung is used as fuel, seeds of these indicator species may eventually be found in ash deposits. Figure 52: Prostrategrowing specimen of Purslane (Portulaca oleracea) in a road verge fringe (Geraki, Greece; July 2010).

The seed bank, in addition, may present both seeds of weeds that are indicative of irrigation and seeds of aquatic plants that grow in water used for irrigation. Seeds of aquatic plants will be spread over the surface of the field during irrigation and will subsequently become mixed with the seed bank by ploughing. These seeds will not germinate, but transport of clay from irrigated fields to the settlement could result in them being retrieved from archaeobotanical samples. The soil may become saline due to prolonged irrigation. This is because even fresh water contains small amounts of minerals, including salty compounds. During exposure to the air, the water evaporates and the concentration of salts may gradually increase in the top soil if it is not washed out. The concentration of salt may increase to a level at which the intake of water becomes problematic and crop failure results. Since barley has a shorter life cycle than wheat, this cereal is slightly better adapted to these kinds of soils. Summer grain has a shorter life cycle in comparison with winter grain, but its cultivation depends on the availability of summer varieties and the availability of sufficient water in summer.

1.2.5.5

Harvesting

Harvesting methods and harvesting implements differ depending on the crop that has to be taken from the field. Harvesting can be done by hand stripping, by using a basket and seed beaters, by uprooting the whole plant, or by reaping. Hand stripping and collecting with a swinging basket is used mainly in the harvesting of wild plants. Contrary to domesticated crops, wild plants are able to release their diaspores. This means that the ripe diaspores are easily separated from the parental plant, making it possible to collect

1.2 Plant ecology 63

them by hand stripping or by using a swinging basket. A disadvantage of this kind of harvesting is that only a small portion of the total seed production can be collected. Uprooting and reaping is used for the harvesting of both wild and domesticated crops. Uprooting implies that the whole plant is pulled out. This kind of harvesting is widely used for different reasons. Pulses are mostly uprooted because they are low-growing plants and the fruits are not concentrated in an infructescence (fig. 53). The fruits are removed by hand and threshed. Cereals are uprooted if they grow in low densities (fig. 54). Fibre plants such as Flax (Linum usitatissimum) are uprooted because the whole culm can be collected. Moreover, the strong fibres do not allow reaping. Uprooting is also applied to crops that easily lose their seed if harvesting is done by reaping. This is the case, for example, with Sesame (Sesamum indicum), in which the fruits open at maturity. Despite their long history of cultivation, most cultivars are still able to disperse their seeds. The ripe seeds are easily shaken out of the fruits when they move in the wind or are cut with a sickle. To reduce seed loss during harvesting, sesame plants are uprooted before the fruits open. Because the whole plant is uprooted, after-ripening is possible by reallocating the moisture from the seed and fruit to the vegetative plant parts. A disadvantage of uprooting is that soil particles remain attached to the roots and soil is brought to the threshing floor. Figure 53: Uprooted domesticated Chickpea (Cicer arietinum) (Güzelyurt, Turkey; August 2010).

64 Handbook of plant palaeoecology | Cappers & Neef

Figure 54: Uprooted hulled 6-row Barley (Hordeum vulgare ssp. vulgare) (north of Tamalout, southern Morocco; April 2007).

Reaping is done with a sickle or a scythe. The scythe developed later than the sickle, because it cannot be used in fields that have stony soils. Stones can be present in reasonable quantities in new fields, but are removed by farmers over the years. A sickle consists of a handle and a curved blade that are located in the same plane. It is suitable for cutting both short and long crops at different heights. The handle of a scythe is longer and is angled in relation to the long blade. This implement is especially used for mowing hay, but was used for reaping crops as well. Sickles are rather diverse in shape, which is partly related to the kind of crop and partly to its independent development in different parts of the world, in which cultural and social factors play a role as well (figs. 55, 56, 57, and 58). The sickle can have a smooth-edged blade or a serrated blade. The latter is used for crops with tough culms. 1.2 Plant ecology 65

Figure 55: Finger knife used for harvesting rice. A sharp piece of serrated metal is fixed in a handle made of polished buffalo horn (southern China).

Figure 56: Finger knife used for harvesting rice. The handle consists of a triangular-shaped piece of wood containing a metal blade, and a piece of bamboo (Kali, southern China).

Figure 57: Sickle and carved wooden holder (Nepal).

66 Handbook of plant palaeoecology | Cappers & Neef

Figure 58: Sickle (Afghanistan).

The use of a sickle is most probably related to the density of the crop and uneven ripening of the plants. Reaping with a sickle implies that culms are gathered with one hand, and this only makes sense if a sufficient number of culms can be collected. To increase the number of culms that can be cut in one go, the shape of the sickle can be adapted. One possibility is to increase the size of the metal blade. Sickles also have been developed that can be used both for gathering and for subsequently cutting the culms, such as rice sickles used in Cambodia. To increase the number of culms that are cut simultaneously, it is also possible to extend the fingers of the hand with harvesting claws (fig. 59).

Figure 59: Curved wooden implement with leather sleeve, which is placed on a finger. Four of these implements are used to extend the reach of the hand. Its Turkish name, ellik, means for the hand. Total length: 22 cm (Kahramanmaras, Turkey; May 2009).

A wild population benefits from an uneven ripening because the seed dispersal is spread over a longer period, thus reducing the risk of unsuccessful propagation. A farmer, however, benefits from a synchronized ripening of the crop, which is considered a domesticated trait. Nevertheless, uneven ripening may still occur within a particular field. In fact, one single unripe culm in a handful of culms that is gathered for harvesting will hinder the breaking off by hand. This may result in seed loss, especially when freethreshing cereals are harvested. By using a sickle, a handful of culms is easily cut, even if some culms are not yet ripe and brittle. Hulled cereals can be harvested when the entire spikes are ripe. Because the rachis has become non-brittle in domesticated crops, all spikelets remain attached to the parental plant. This implies that basically the whole yield can be 1.2 Plant ecology 67

harvested with a sickle or by uprooting. In naked cereals, on the other hand, the dispersal unit is no longer a spikelet but the grain kernel. Because seed dispersal has become possible again, ripe grain kernels will be easily scattered around during harvesting. To reduce this kind of yield loss, naked cereals are preferably harvested before the ripening has finished. Completing the ripening process is possible if reallocation of water from the grain kernels to the vegetative plant parts remains possible. This is achieved by harvesting the plants close to the soil, preferably in the morning, when dew prevents shattering of grain kernels. After-ripening takes place on the field proper (fig. 60). Figure 60: Harvesting Bread wheat (Triticum aestivum) with a sickle. This free-threshing wheat is harvested close to the soil to facilitate after-ripening (Ezbet Basili, Egypt; April 2002).

Cereals can be harvested on three height levels. The culms can be cut close the ground, they can be cut about mid-way, or they can be cut just below the ears. In each case, the harvesting may include two stages. If the culms are cut close to the ground, the plants can be collected in the field and the ears can be cut off. If the crop is cut at the middle of the culm or if only the ears are cut off, the remaining part of the culm can be harvested in a second stage. All three possibilities are mentioned, for example, by Varro (On Agriculture: 1.50) and Columella (On Agriculture: 2.19–20). Cutting the culms at the middle was, most probably, often practised in ancient Egypt in harvesting Emmer, judging from the wall paintings in burial chambers (fig. 61). Figure 61: Wall painting from the mastaba of Mereruka at Saqqara (sixth dynasty), showing the reaping of Emmer. The culms are cut mid-way. This is a reproduction on Papyrus, painted by Nadia Kotb.

68 Handbook of plant palaeoecology | Cappers & Neef

Figure 62: Field with maize (Zea mays) that has been harvested. From a distance, it seems as if the complete plants are present (left). A closer look shows that the maize cobs have been removed (right) (Ezbet Dawud, Egypt; November 2008).

The level at which the crop was cut during harvesting can be inferred from the kinds of field weeds mixed with the grain kernels. It is, however, not sufficient to use the average height of the field weeds for determining this level. Any interpretation should also take into account the distribution of the fruits along the stems, as well as the possible creeping habit of some wild plants. From some crops, such as Sorghum (Sorghum bicolor), Sunflower (Helianthus annuus), and Maize (Zea mays), infructescences can be collected without diaspores of associated field weeds (fig. 62). Some contamination with diaspores may occur, however, during further crop processing. Such diaspores are not related to the crop and cannot be used to infer agricultural practices or environmental conditions.

1.2 Plant ecology 69

Figure 63: Empty ears of Bread wheat (Triticum aestivum), the result of bird predation (Kozan, Turkey; May 2009).

70 Handbook of plant palaeoecology | Cappers & Neef

Figure 64: Flower-heads of Sunflower (Helianthus annuus) with ripening fruits are protected from bird predation with plastic bags (Çiflik, central Anatolia, Turkey; August 2010).

During the entire growing season, and close to harvesting time in particular, crops are vulnerable to predation. Agricultural experiments in salt marshes have, for example, shown that crops can be damaged by insects, cattle, and birds (figs. 63 and 64). The fencing of fields proved to be necessary to prevent such a loss of yield. Birds need trees in the vicinity of the fields to use as shelter; in the absence of trees, depredation by birds is no longer a serious threat.

1.2.5.6

Threshing, winnowing, and sieving

Threshing is the first process after harvesting. It is aimed at separating the edible parts from the non-edible parts. Threshing can be followed by winnowing and sieving, during which the edible parts are cleaned to a higher degree. While threshing, if practised, takes place only once, winnowing and sieving can be applied several times before the grains are used for food preparation. Crop processing following harvesting can be rather complicated and is determined by the kind of crop, its use, the harvesting method, and the cultural traditions, which may be partly related to environmental conditions. Hulled and naked (or free-threshing and hull-less) cereals are threshed using different methods. In hulled cereals, the grain kernels are tightly enclosed by the chaff, whereas naked cereals have loose chaff. The threshing of hulled cereals is aimed at separating the spikelets from the rachis, whereas the threshing of naked cereals is aimed at separating the grain kernels from the ear. The threshing of naked cereals can be completed in one go. The grains are easily separated from the chaff on the threshing floor—hence its alternative adjective, free-threshing. The threshing remains of naked cereals consist of culm fragments; chaff (namely, glumes, lemmas, and paleae); and diaspores of field weeds (fig. 65). The lighter fragments can be removed by winnowing. The difference in husk tightness determines both the use that the threshing remains can be put to and the archaeological contexts that should be sampled.

1.2 Plant ecology 71

Hulled cereals need two stages of threshing. The first threshing results in the fragmentation of the culms and spikes. Subsequent winnowing and sieving separate the spikelets from the threshing remains. Threshing remains from hulled cereals consist of culm fragments (the number depends on the reaping height) and diaspores of field weeds (fig. 66). A second threshing is necessary for dehusking the grain kernels. This can be done by pounding and grinding. To facilitate the removal of the chaff, spikelets can be slightly roasted. Figure 65 : Threshing remains from Bread wheat (Triticum aestivum) produced by a modern threshing machine. These remains are fragmented to a high degree and are relatively poor in diaspores from associated field weeds (Tunis, Egypt; September 2002).

Figure 66: Threshing remains from hulled 6-row Barley (Hordeum vulgare ssp. vulgare) produced by a modern threshing machine. These remains contain a large number of diaspores from associated field weeds. Barley is only represented by some smaller sterile spikelets (Calabria, Italy; July 2007).

72 Handbook of plant palaeoecology | Cappers & Neef

The degree of fragmentation of the threshing remains depends on the implements used. The use of a flail or roller will bruise the crop. The same is true if animals, such as horses or oxen, are used for trampling the grain out of the crop (fig. 67). More advanced implements, such as the threshing sledge, having stones or metal on its underside, combine separating with fragmenting. Flailing and eventually rolling can be done inside, whereas trampling and sledging need more space and are done on a threshing floor that is located outside. Threshing may be delayed because of the pressure of work during the summer period. In that case, the unthreshed harvest is stored and threshing is done during the winter period at regular intervals. Figure 67: Wall painting from the tomb of Mena at Thebes (eighteenth dynasty), showing threshing of ears of Emmer using longhorn cattle. This is a reproduction on Papyrus, painted by Nadia Kotb.

When hulled cereals are ear-harvested, threshing is not necessary if the grain is to be used as fodder or for making beer, especially in the case of 2-row Barley. If used as fodder, the whole yield can be fed to the animals. There are several advantages to this. The spikes can be fragmented into their individual spikelets, and they can also be separated from the culm fragments. Fragmentation of the harvested crop also reduces the amount of space that is needed for storage. The spikelets have a higher nutritional value. And finally, spikelets are larger and are more easily chewed than the grain kernels themselves. For beer making, spikelets are isolated from the yield, but separating the grain kernels from the chaff (dehusking) is not necessary. Grain kernels are allowed to germinate, and this occurs naturally within the chaff. Only when grain is to be used for human consumption should an effort be made eventually to separate the grain kernels from the chaff. Threshing can be done off-site or on-site. In the latter case, the yield can be stored and threshing can be done in periods when less labour is required on the fields. Threshing is preferably done on a threshing floor that is clean, flat, and solid (fig. 68). The size and quality requirements for the floor also depend on the implements used for threshing. Cato (On Agriculture, 91) and Varro (On Agriculture, 11), for example, describe the requirements for a threshing floor. To facilitate the drainage of rainwater, the central part should be somewhat elevated, and the solid and well-packed floor may be coated with amurca, being the last liquid that is produced during olive pressing. This black water has antiseptic properties and protects against weedy plants, insects, and moles.

1.2 Plant ecology 73

Figure 68: Overview and detail of a village threshing floor. The floor consists of large, flat stones. The metal pole in the centre was used for fastening the rope used to restrain the animals that were used for threshing. Having fallen out of use, the threshing floor has become a storage place for fuel wood (Geraki, Greece; July 2010).

Some crops, such as Sesame, can be threshed without special tools. When the uprooted plants have become sufficiently dry, they can be turned upside down, and the seeds can then be shaken from the fruits (fig. 69). The threshing of many crops is, however, done by animals or with special tools. By walking over the harvested crop, animals can tread the spikelets from the ears and the seeds from the fruits with their feet. Traditional tools used for threshing are flails, threshing sledges, and rollers (fig. 70). A flail is made from two rods of uneven length, connected to each other with a short chain or other strong, flexible material, such as the dried skin of an eel, which is extremely durable. The threshing sledge was used until recently around the Mediterranean and in the Near East. The most common type consists of a

74 Handbook of plant palaeoecology | Cappers & Neef

wooden board with many small holes on the bottom, in which sharp stones or metal blades have been inserted (fig. 71). The use of sledge flints illustrates that even in modern times implements could be still partly made of stone. In Egypt, a different type of threshing sledge evolved: the nurag. This sledge consists of a frame that is furnished with metal or wooden rollers. Each roller carries several large metal disks (fig. 72). Figure 69: Seeds of Sesame (Sesamum indicum) are shaken from the fruits (Tunis, Egypt; October 2003).

Figure 70: Threshing fruits of Chickpea (Cicer arietinum) using a roller. Picking the fruits from the plants and crushing the fruits is done on the roof (Murtazaköy, Turkey; August 2010).

1.2 Plant ecology 75

Figure 71: Bottom view of threshing sledge with flint inserts. This specimen originates from the Çanakkale Province in western Turkey. Length: 185 cm; width: 40 cm. (purchased in Istanbul, November 1994).

Figure 72: Threshing sledge furnished with rollers carrying disks (Salehia, Egypt; May 2005).

Winnowing is a method of cleaning the crop that makes use of differences in weight. The threshed crop is thrown into the air or scattered from a basket, fan, or sieve, allowing the wind to blow away the lighter straw waste and seeds of weedy plants. The heavy grain and heavy diaspores of wild plants fall back to the same spot (figs. 73 and 74). 76 Handbook of plant palaeoecology | Cappers & Neef

Figure 73: A wooden winnowing fork is used for throwing threshed Chickpea (Cicer arietinum) into the air (Yozgat, Turkey; August 1997) (photo by R. Neef).

Figure 74: Decorated and strengthened winnowing fan from the Swat valley (northern Pakistan). Total length: 163 cm; only the blade and part of the stalk are shown.

1.2 Plant ecology 77

Sieving is a method to separate the crops seeds from their chaff and associated field weeds by differences in size. This can be done with a nested combination of a coarse sieve and a fine sieve, the latter with perforations slightly smaller in diameter than the crop seeds (fig. 75). The residue of the top sieve, having a coarse screen, will consist of large waste fragments, whereas the residue of the bottom screen will consist of the prime crop seeds. The mesh sizes of both sieves have to be adapted to the crops that are to be cleaned. A high level of accuracy is required for the production of a sieve in order to combine a successful concentration of crop seeds and the elimination of field weeds. This is achieved by an even and appropriate mesh size. Determining the size distribution of a former yield may be problematic due to the cleaning of the crop, since only the prime specimens will have been retrieved on the lower sieve, whereas the smaller tail specimens may have passed through the lower sieve. Such discarded tail seeds may be recovered in trash layers or as a kind of settlement noise in a random sampling, but cannot be used to estimate the size distribution. Figure 75: Cleaning Bread wheat (Triticum aestivum) with two sieves. In this way, three fractions are produced: a sieve residue on the top, coarse sieve; a sieve residue on the bottom, fine sieve; and a waste fraction that has passed through the bottom sieve and is scattered on the ground (Shawashna, Egypt; May 2004).

Today’s sieves have metal frames and a standardised mesh size. In the past, however, sieves were handmade and the mesh screens were made of small strings of botanical or zoological origin (fig. 76). Figure 76: Detail of Roman sieve from Karanis (left; Agricultural Museum, Cairo) and handmade modern sieve (right; Fayum, Egypt; May 2004).

1 cm 78 Handbook of plant palaeoecology | Cappers & Neef

Winnowing and sieving reduce the number of diaspores of the associated weed plants, but the diaspores that resemble the crop seed in size and weight will be difficult to eliminate. In this way, both these methods of crop cleaning unintentionally resulted in the selection of field weeds (fig. 77). This implies that for the reconstruction of former agricultural practices, which mainly depends on the analysis of the associated field weeds, archaeological contexts could be samples representing the stock supply, the threshing remains, or the waste fraction. Until recently, sieves were still handmade and mesh sizes were not uniform. This uneven mesh size and the possible presence of unrepaired holes explain how a minor proportion of the crop seeds could also end up in the waste fraction (fig. 78). Figure 77: A Flax (Linum usitatissimum) sample from Argentina measuring 240 ml contains 14 ml (almost 6 per cent) of diaspores of weed plants. The diaspores of these wild plants have been picked out and are shown separately. It is striking that the majority of the diaspores of the weedy plants are smaller than those of the associated crop.

1.2 Plant ecology 79

Figure 78: Waste from Faba bean (Vicia faba var. equina) that passed through a sieve used for cleaning the stock prior to selling. The waste fraction typically consists of a mixture of the crop; a few grains from other crop species, in this case Sorghum ((Sorghum Sorghum bicolor bicolor)) and Cow pea (Vigna unguiculata ssp. unguiculata); many diaspores from weed plants, including those from Beet (Beta vulgaris) and Melilotus messanensis; and soil particles. The majority of the Faba beans consist of seed fragments from full-grown specimens and whole but immature (aborted) seeds. The presence of a few mature seeds is indicative of a hole in the screen (El Fayum, Egypt; November 2006).

Figure 79: Grain kernels of Wheat (Triticum) are dried on the field after threshing (Central Anatolia; c. 2000).

1.2.2.7

Drying and storing

Because most crops are harvested only once a year, proper storage of the yield is of vital importance to the farmer. Special care has to be taken to prevent spoilage by microorganisms and animals. Seeds must be thoroughly dry before they can be stored for a long period. The moisture content of grain kernels, for example, should be less than 14 per cent, a condition that is called dead-ripe. In this condition, they are less vulnerable to mould. Traditionally, seeds are air-dried (fig. 79). During the drying process, crop seeds may become contaminated with diaspores of wild plants that are different from the associated field weeds. This kind of contamination is likely to occur if seeds are dried away from the field. Such contamination is easily recognized if the crop can be harvested without diaspores of field weeds, such as is the case with Sorghum and Sunflower (figs. 80 and 81). But with other crops, such as barley and wheat, it may be difficult to judge upon such contamination.

80 Handbook of plant palaeoecology | Cappers & Neef

Figure 80: After-ripening of heads from Sunflower (Helianthus annuus) (Fayum, Egypt; November 2008).

Figure 81: Fruits of Sunflower (Helianthus annuus) spread on a street to dry (Yenisehir, Turkey; August 2010).

1.2 Plant ecology 81

After drying, the seeds can be packed and transported to the storage facility. This is also the moment to quantify the yield (fig. 82). The volume is determined by using grain measures. Such measures have a fixed volume and are mostly available in different sizes, depending on the kinds of crops that have to be measured. Shape and volume also differ among regions (fig. 83). Figure 82: Wall painting in the tomb of Meryneith showing officials recording the size of the yield after threshing (top) (Saqqara New Kingdom necropolis; 1350–1325 BC). Measuring the amount of Bread wheat (Triticum aestivum) collected on reed mats (bottom) (Tunis, Egypt; October 2003).

82 Handbook of plant palaeoecology | Cappers & Neef

Figure 83: Nicely decorated cone-shaped grain measure from Egypt made of vertical wooden slats and covered with metal. Volume: 2150 ml (purchased in Cairo, c. 2005).

The yield can be stored in the settlement proper or outside the settlement. On a domestic scale, only relatively small amounts have to be stored. The yield can be stored in small storage facilities, such as bags, basketry, pottery, and even furniture (fig. 84). It is also possible to store the yield in separate storerooms or bins (figs. 85 and 86). Such storage facilities were filled from above and could be emptied through a hole at the base. During this filling, the small-sized diaspores of field weeds tend to roll down. If these diaspores are to be used for the reconstruction of former agricultural practices, it is necessary to take subsamples throughout the volume, including the from the bottom layer, of such storage facilities. A subsampling of several layers may also be considered for large granaries, to take into account the possibility that crops from different fields were stored on top of each other.

1.2 Plant ecology 83

Figure 84: Grain chest from the Swat Valley (northern Pakistan). Volume: 80 litres.

Figure 85: Storage of grain and threshing remains on a roof. The threshing remains are from Sorghum (Sorghum Sorghum bicolor bicolor) and are to be used as fuel. The large storage bins have a capacity of 300 kg and are used for the storage of grains and pulses. Today, such bins are becoming obsolete and are used for mounting antennas (village south-east of Tamiya, Egypt; May 2005).

84 Handbook of plant palaeoecology | Cappers & Neef

Figure 86: Storage bins made from clay. Left: Wall painting in the tomb of Meryneith. Most probably the repeated pattern represents the spikelets of the grain that is stored (Saqqara New Kingdom necropolis; 1350–1325 BC). Right: Recent bin that is filled from above and then sealed. Below is a hole through which small amounts can be unloaded. This hole has been sealed with clay (Qarun, Egypt; November 2010).

The storage of the yield within the settlement offers the advantage of close protection and control. A disadvantage may be the large volume of the total yield and the high number of all kinds of animals that are around and have to be kept out of the storage facilities. Trash deposits in a settlement will not only be visited by domesticated animals, such as sheep, goats, chickens, and dogs, but will also attract mice and rats. An alternative location may be present outside the settlement. An example is the food storage of the dwellers of the Neolithic settlement Kom K, in the Fayum. The food was kept in baskets that were buried in a sandy outcrop about 2 km from the settlement. The baskets were sealed with firm lids made from wetted calcareous sand. Once the sand was dry, the food was well protected against vermin. The importance of safe storage is also illustrated by the communal storage of food in well-protected locations. In Morocco, for example, such communal storage facilities are transformed into real fortresses found in almost inaccessible locations (fig. 87).

1.2 Plant ecology 85

Figure 87: Large fortress for collective food storage (called an agadir agadir) built on a mountain top (Amtoudi, Morocco; April 2007).

Part of the yield risks being destroyed during storage by insects and rodents. Insects such as weevils, beetles, and moths may get access to the storage facility and spoil considerable parts of the yield. Some of these animals may develop in the seed, and the infestation may occur before the seeds are even harvested. This is a serious hazard in the storage of pulse seeds. When stored, the host may develop and finish the seed and, depending on its life cycle, may consume other seeds as well. Other insects bore into the full-grown seed and consume its interior. Even small amounts of waste grain can be a source of insect infestation. It is desirable, therefore, to clean the storage facility before filling it with the new yield. The plaster on the inside of a granary can be coated with amurca, which is a repellent against insects and small mammals. Naked cereals are stored as dehusked grain kernels. But it seems that hulled cereals were stored both as spikelets and as grain kernels. During ethnoarchaeobotanical research in Eastern Anatolia, Hillman (1981) documented the storage of Emmer wheat (Triticum turgidum ssp. dicoccon) after two threshing stages, thus as dehusked grain kernels. According to Hillman, the storage of hulled wheat, such as Emmer, after dehusking is typical for areas with dry summers, at least for Turkey. In parts of Turkey that have wet summers, the hulled wheat is put into storage as spikelets, which are only dehusked on a small scale prior to food preparation. However, it seems that in ancient Egypt, Emmer wheat was always stored in the husk (Murray, 2000). Although the storage of dehusked grain has the advantage that it reduces the storage space by half, it is more vulnerable to decay in this condition. It is therefore more likely that hulled cereals were only fragmented into individual spikelets by a single threshing and subsequently stored. Dehusking could be practised on a small scale prior to food preparation.

1.2.5.8

Weed control

To reduce the competition for light, water, and minerals, the farmer can try to reduce the number of field weeds. Field weeds germinate from seeds that are present in the topsoil of the field. The living seeds in this part of the biosphere are collectively called seed bank. The number of viable seeds in the topsoil (the top 15–40 cm) of arable fields can be high and ranges from several hundred up to more than 130,000 seeds per m2 (Cavens & Benoit, cited in Leck et al., 1989).

86 Handbook of plant palaeoecology | Cappers & Neef

Most of the seeds present in a seed bank originate from the field weeds that grow in the field proper and succeed in completing their life cycle. The seed production per m2 may differ among arable weeds and is related to the individual reproductive effort, defined as the individual’s net investment of resources in reproduction (Bazzaz & Ackerly, cited in Fenner 2000), and the patchy distribution of field weeds. Seed production of White clover (Trifolium repens), for example, ranges from 910 to 109,000 seeds per m2 (Turkington & Burdon, cited in Leck et al., 1989). In addition, seeds may enter the seed bank by seed dispersal. This could be dispersal either by the agency of wind, water, or animals or by the agency of humans in using seed for sowing that is contaminated with seeds of weedy species. Weed control is possible in several ways. The standing weed vegetation can be destroyed by using a hoe or a plough. However, hoeing and ploughing with an ard will not effectively reduce the number of field weeds. Loosening the soil may even initiate the germinating of field weed seeds present in the seed bank. The exposure to very short flashes of sunlight may result in the breakdown of inhibitors that control seed dormancy. A plough, on the other hand, turns the soil, and this will destroy part of the established weed vegetation (fig. 88). The introduction of the plough will have reduced the proportion of perennial weed plants in the fields, such as Heath grass (Danthonia decumbens). Figure 88 : Although the soil has been turned by a plough, parts of the arable field weeds are not completely covered by the soil. Such plants may still complete their life cycle (Tufanbeyli, Turkey; May 2009).

1.2 Plant ecology 87

Whether it is possible to remove arable field weeds during the growing season depends on the kind of crop and the size and shape of the field. When broadcast sowing is practised, plants grow in an irregular pattern and their density mostly does not allow walking through the field. This kind of sowing is, for example, used for cereals. If the plots are large, weed control during the growing season is limited to the fringes of the field. Especially noxious weeds will be removed in this way, such as thistles and the poisonous Corn cockle ((Agrostemma githago) (fig. 89). Figure 89: Uprooted Corn cockle (Agrostemma githago) from the fringe of a cereal field (Republic of Macedonia; May 1998).

88 Handbook of plant palaeoecology | Cappers & Neef

Fields in which plants grow in rows facilitate the removal of field weeds with a hoe throughout the growing season. Rows can be achieved by ridge sowing; by planting plants out from breeding beds; or by growing plants from vegetative propagules, such as bulbs, corm, or tubers. This kind of weed control is also possible in cereal fields where the individual plants are spaced out due to the environmental conditions. In the sorghum field shown in fig.  90, the following weed plants were recorded: Forsskålea tanacissima, Farsetia ramosissima, Aristida funiculata, A. adscensionis, Tribulus terrestris, Dichanthium foveolatum, Astragalus eremophilus, A. vogelii, Trichodesma ehrenbergii, Zilla spinosa, Acacia tortilis, and Morettia. All these species are desert plants that are also found outside the cultivated areas. Figure 90: Ababda nomad weeding a field of Sorghum bicolor in the coastal area of the Red Sea (Areb Saleh, Egypt; January 1997).

1.2 Plant ecology 89

Weed control can also be performed by burning the field weeds so as to reduce the number of seeds that will be added to the seed bank. This can be done before sowing or after harvesting. Burning of weedy plants is only practised when after-grazing is not practised or when inedible plants, such as Camel thorn ((Alhagi graecorum), predominate (fig. 91). To improve the effectiveness of burning, plants can be uprooted with a hoe and collected on heaps. But even this process does not result in the destruction of all fruits (fig. 92). Burning of such concentrations of weed plants results in circular ash layers spread over the field. The surface of such fields is characterized by large, vague black spots after ploughing. Figure 91: Burning heaps of weeds on a field prior to irrigation (Ezbet Dawud, Egypt; November 2008).

Figure 92: Detail of burned heap of Camel thorn ((Alhagi graecorum), showing unburned stems and ripe fruits (Ezbet Dawud, Egypt; November 2008).

90 Handbook of plant palaeoecology | Cappers & Neef

The burning of stubble vegetation after the harvest is typical for environments that offer sufficient grazing land, making it unnecessary to use the stubble fields for after-grazing (fig. 93). The effectiveness of burning depends on the dryness of the wild plants and the density of organic matter. Field weeds that are still fresh have a good chance of surviving the flames. The same is true for field weeds that are present in less dense parts of the vegetation. The ash layer that is produced will improve the mineral composition of the field in the short term. But a disadvantage may be that part of the ash is winddispersed out of the field, thus resulting in a reduction of the mineral content comparable with the removal of the harvest. Figure 93: Burning of cereal stubble field. Some spots remain unburned (vicinity of Aksaray, Turkey; October 1998).

The burning of weeds may result in contamination of the archaeobotanical archive (fig. 94). Contamination with recent charred material especially occurs when shallow archaeological sites are covered by agricultural fields. This kind of contamination with recent charred material has been demonstrated by Smith (2007) in the vicinity of Çadir Höyük. Incorporation into the settlement layers is supported by ploughing and animal disturbances, such as the digging of burrows by rodents. Charred remains of crop plants can be considered as subfossil if they concern crops that are no longer cultivated, such as Emmer and Einkorn. But prudence is called for when, for example, charred remains of Barley, Bread wheat, or Hard wheat are found. Contamination of recent charred wheat could be evidenced, for example, in samples from the Neolithic tell Kurdu (Amuq area, Turkey). Some samples yielded grain kernels and rachis fragments that were completely charred, whereas other plant remains were partly charred or even uncharred. All plant remains of this crop as well as the accompanying charred rachis nodes of Aegilops were considered as modern and rejected from the record (Cappers, in press).

1.2 Plant ecology 91

Figure 94: Culm fragments and spikes of hulled 2-row Barley (Hordeum vulgare ssp. distichon) after burning of the stubble. Some spikes are almost untouched by the flames, whereas others have become charred (Demirci, Turkey; August 2010).

The vegetation of stubble fields used for grazing after the harvest consists of small field weeds and large but prostrate-growing plants. Stubble grazing reduces the coverage of field weeds after the harvest. This kind of weed control is especially practised if pasture land is limited or even absent, as is the case in a desert environment such as Egypt. The degree to which the number of diaspores is reduced depends on the number of viable seeds that pass the digestive tract, as well as the removal of dung from the fields. Collecting dung can be practised for the production of dung cakes as a source of fuel. A reduction of seed input to the seed bank can also be realized by cleaning the seed stock. A first reduction of the number of diaspores of field weeds is accomplished with threshing. After fragmentation of the crop, threshing remains can be separated from the seeds by winnowing and sieving. As we noted above, winnowing makes use of differences in weight, whereas sieving is based on differences is size. Unfortunately, neither type of cleaning will produce a pure stock of seeds. Some seeds of field weeds that match the crop seeds in size and/or weight will show up in the seed stock. Seeds of wild plants can be doubly undesirable if they are poisonous, such as the seeds of Corn cockle ((Agrostemma githago), or spiny, such as the spikelets of wild oats ((Avena spp.). The removal of all field weed seeds is time consuming and does not make sense. But a complete cleaning of seed stocks is an option because the quantity is much smaller and a full cleaning also prevents a return to the seed bank.

92 Handbook of plant palaeoecology | Cappers & Neef

1.3

Flora and vegetation

1.3.1

Landscape, flora, and vegetation Both flora and vegetation can be related to landscape. The main difference between these concepts is that a flora is always an abstraction, whereas a vegetation is related to the tangible landscape. In other words, one cannot walk through a flora, but one can walk through a vegetation. A flora is mainly based on morphological descriptions, and an enumeration of its plant species gives only a general impression of the landscape it relates to. A vegetation is the established plant cover in a certain part of the landscape. It takes into account both species diversity and species abundance. The original vegetation, which can be considered to be the natural vegetation, is determined by both abiotic and biotic factors (fig. 95). Abiotic factors include those related to the soil and climate. Whether a given plant will grow in a particular soil type is partly determined by the physical and mineralogical composition of the soil. The water availability is primarily determined by climate and is affected by precipitation and temperature.

Figure 95: Relationship between landscape, flora, and vegetation.

natural landscape

cultural landscape

landscape

abiotic factors

geology

biotic factors

climate

plant

animal

edaphic

human anthropogenic

flora

vegetation

morphology

taxonomy

syntaxonomy

ecology

autecology

synecology

Climatic change and human impact can be responsible for changes in the vegetation. Much of the current landscape that is inhabited by humans has undergone serious alterations. The natural vegetation is often changed into vegetation that is adapted to human interference. These vegetation types have become part of the cultural landscape and thus differ from the vegetation that would potentially have been present without the impact of humans (figs. 96 and 97).

1.3 Flora and vegetation 93

Figure 96: Vegetation bordering a cereal field, consisting of Papaver rhoeas and Sinapis arvensis (Angeren, the Netherlands; July 1990).

Figure 97: Landscape characterized by overgrazing and over-cutting of trees. The potential herbaceous vegetation is only present within the walls of the cemetery. The Pistachio trees (Pistacia spp.) at the back are protected because they are holy trees near the shrine of Elif Ana. Their presence on top of the hill indicates that the availability of water is not a limiting factor (vicinity of Mamaz Tepe, Turkey; August 2000).

1.3.2

Flora In its simplest form, a flora is a list of plants present in a particular area. The plant species are listed according to a hierarchical classification, and each level (taxon), with the exception of genus and species, can be recognized by a standard suffix. The area that is covered by a flora can differ and may include, for example, a continent, country, province, or nature reserve. A flora usually contains an identification key and a brief description of each species. Ideally, this description should be complete for each plant species, and the morphological features that are described should be presented in a fixed sequence. Additional information on distribution and ecology can be included as well. More often, however, the description of the plants is limited and confined to the features that have diagnostic value and are used for identification to the species level.

94 Handbook of plant palaeoecology | Cappers & Neef

A flora is also characterized by time. It describes the plant species based on a survey in a given period. But the species composition in each area is subject to changes that are the result of natural plant dispersal, the transport of diaspores by human activity, and the changeability of the landscape. As a result, some species disappear, whereas others invade from other regions. Therefore, it is desirable to update a flora on a regular basis. A flora distinguishes between the indigenous (native) plants and nonindigenous (alien) plants. The term indigenous has a relative meaning and depends on the period under consideration. Invaders may become established and will be considered as naturalized after a while. In the Netherlands, for example, a plant species is considered to have become naturalized when it has become part of the natural vegetation and is capable of reproduction without the interference of humans (fig. 98). Figure 98: The Wild tulip (Tulipa sylvestris) is indigenous in southern Europe and was introduced in central Europe in the sixteenth century. The earliest record of this tulip from the Netherlands dates back to 1811 (Mennema et al., 1985).

The neolithisation of a region, for example, was responsible for the introduction of new field weeds together with the seeds for sowing. Weed plants that are considered to have been introduced into Europe with sowing seeds during the early Neolithic include Rye brome (Bromus secalinus), Barren brome/Drooping brome ((Anisantha sterilis//A. tectorum), False cleavers (Galium spurium), and Hairy tare (Vicia hirsuta) (Bakels, 2009). This kind of seed dispersal has been strongly reduced today because seeds for sowing have to be offered for sale without any contamination. Invaders that are currently successful are dispersed by soil rather than seeds for sowing, such as the Earth almond (Cyperus esculentus). The term agriophyte (from the Greek agrios, which means wild, and phyton, which means plant) is applied to plants that have been introduced into a new area by the agency of humans and have become part of the natural vegetation, where they are capable of propagating without interference of humans (Lohmeyer & Sukopp, 1992). Agriophytes can be subdivided into archaeophytes and neophytes. Archaeophytes are plants that were introduced before Columbus reached the New World (1492). Plants that have become established in a new region after 1500 are called neophytes.

1.3 Flora and vegetation 95

Archaeobotanical research may contribute to our knowledge of the indigeneity of plant species, as could be demonstrated for Rough cocklebur ((Xanthium strumarium) by Brinkkemper and Kuijper (1993). Originally, it was thought that this species, characteristic of disturbed sandy soils, was indigenous to western and central Asia. The recent recovery of subfossil diaspores in Europe indicates, however, that we are, in fact, dealing with an indigenous species. The connection of the diaspores with the coastal area and the fluviatile districts in particular, indicates that this plant most probably was distributed by main river streams as early as the Iron Age. It can also happen that ornamental plants escape from gardens and become invaders in existing (semi-)natural vegetation. Once such plants have become wild and succeed in successful propagation, they may become part of the indigenous flora as well. An example of a successful escape is Canadian fleabane (Conyza canadensis). This garden plant originates from North America and belongs to the earliest plants introduced from this area into Europe. Canadian fleabane was introduced in the Netherlands by the eighteenth century. It produces many small fruits that are wind-dispersed. Today, its distribution in the Netherlands covers most of the country. The necessity of updating a flora is also related to new insights in plant taxonomy. Up until about 15 years ago, the classification of plant genera into families was based on morphological and anatomical characteristics, developmental biology, and the presence of certain chemical compounds. The relationships between plants, expressed in the organization of plant families in classes and orders, was therefore partly speculation. This explains why there are several phylogenetic classifications that express the evolutionarily-defined relationships between plant taxa. Ecological research has also had an impact on plant taxonomy, although its influence is vanishing. Subspecies and varieties based on a combination of plant morphology and ecological differentiation have been temporarily accepted in the Dutch flora, for example, for Perennial sow-thistle (Sonchus arvensis), Yellow rattle (Rhinanthus minor), and Greater yellow rattle (Rhinanthus angustifolius). A fundamental change in the classification of flora was triggered by molecular research. Some 15 years ago, taxonomists around the world, united in the Angiosperm Phylogeny Group (APG), began to analyse specific parts of the DNA of plants. DNA research provides an additional basis for the phylogenetic classification of plants. Current floras have started to accept these new insights, and the impact is considerable. Whereas changes in previous editions of floras were largely at the level of the rearranging of species within genera, molecular analysis shows that a considerable number of genera have to be classified under different families and that even some families have to be rearranged. The Figwort family (Scrophulariaceae) may serve as an example. The (half-)parasites that are traditionally attributed to this family (Euphrasia, Melampyrum, Pedicularis, and Rhinanthus) are now classified in the Broomrape family (Orobanchaceae), and other genera in this family, such as Antirrhinum, Digitalis, Gratiola, Limosella, Linaria, and Veronica, are now housed in the plantain family (Plantaginaceae). The Digital Seed Atlas of the Netherlands (Cappers, Bekker & Jans, 2006) follows a taxonomic classification that is based on recent DNA research, and photos in the atlas provide visual confirmation that the new taxonomic organization is often supported by the morphology of seeds and fruits. Thus, Mare’s-tail ((Hippuris 96 Handbook of plant palaeoecology | Cappers & Neef

vulgaris), originally placed in a separate plant family (viz. Hippuridaceae), is now regarded as a member of the plantain family (Plantaginaceae). In the atlas, the fruit and seed of this species are depicted next to those of Riparian weed ((Littorella uniflora), and the resemblance in their morphology is striking. Also notable is the similarity in the seed morphology of Veronica and Plantago. Both genera have species whose seed is slightly concave (e.g. Veronica hederifolia and Plantago lanceolata) and species with flattened seeds (e.g. V. officinalis and P. coronopus).

1.3.3

Vegetation The vegetation of a landscape is determined by environmental conditions that involve both abiotic and biotic factors. The ultimate species composition in a given area is determined by specific environmental requirements of these plants and by seed dispersal. The landscape can be characterized by a description of its vegetation, which in turn is indicative of the specific combination of environmental factors. Research on plant communities can be related to structure and texture (synmorphology), changes over time (syndynamics), geographical distribution (synchorology), history (synchronology), classification (syntaxonomy), and ecology (synecology). For vegetation reconstruction of the past, it is important to understand the method by which a vegetation is currently being described. One approach is to define plant communities by describing the floristic assemblage in a representative part of the vegetation. Each plant community, in turn, is indicative of the environmental conditions of the landscape. Because plants are adapted to specific environmental conditions, plant communities can also be used to define these conditions. The ecological characterization of a plant community is considered to be more precise than that of the individual plant species, because the combined ecological range of the taxa involved is narrower (Schaminée et al., 1995). Another approach in characterizing the vegetation is to define ecological species groups of species of known habitat. Each group indicates a specific combination of environmental conditions, and the included species are representatives of sites in the landscape that meet these conditions. Such sites, which are defined by environmental conditions and vegetation structure, are called ecotopes (Runhaar et al., 1987). Both approaches will be dealt with in more detail in the next sections.

1.3.3.1

Plant communities and syntaxonomy

A vegetation type is characterized by a combination of plant species. Syntaxonomy is the discipline that deals with the description and classification of vegetation types. Vegetation types are hierarchically defined on several levels, similar to the classification of plant taxa. A specific level of a vegetation type is called syntaxon (plural syntaxa). The most important levels in descending order are class, order, alliance, and association. An association is the basic floristic unit of this classification system, because it is a description of the vegetation as it can be seen in the landscape. The nomenclature of vegetation types is based on its characteristic plant species. The syntaxa are named after the dominating taxon, and each syntaxonomic rank is characterized by its own suffix. The suffix of a vegetation class is -etea and that of an association is -etum. In the Netherlands, the number of classes that are currently distinguished comes to 43, and these have been subdivided into 228 different associations (Schaminée et al., 2010; Table 8). 1.3 Flora and vegetation 97

Table 8: Syntaxa at the level of class that are currently recognized in the Netherlands. The number of associations is mentioned for each class. The first number refers to the number of syntaxa that are well developed, and the second number refers to the number that are not well developed (derivative communities) as a result of declining groundwater level, eutrophication, and fragmentation (aer Schaminée et al., 2010). Open water, marsh vegetation, and moist heathland

Brackish and salt water Ruppieta

2+0

Zosteretea

2+0

Fresh water Lemnetea minoris

3+2

Charetea fragilis

8+1

Potametea

18+8

Liorelletea

8+4

Marshes Montio-Cardaminetea

3+0

Phragmitetea

19+9

Parvocaricetea

8+4

Raised bogs and moist heathlands Scheuchzerietea

4+6

Oxycocco-Sphagnetea

5+3

Grassland, fringe vegetation, and dry heathland

Open to almost closed grasslands Sedo-Scleranthetea

2+0

Koelerio-Corynephoretea

12+13

Nutrient-poor, closed grasslands Nardetea

4+2

Festuco-Brometea

1+0

Moderately nutrient-rich grasslands Molinio-Arrhenatheretea

12+12

Plantaginetea majoris

7+5

Fringe vegetation Trifolio-Geranietea sanguinei

2+0

Melampyro-Holcetea mollis

2+0

Dry heathlands Calluno-Ulicetea

98 Handbook of plant palaeoecology | Cappers & Neef

6+1

Table 8: (continued): Syntaxa at the level of class that are currently recognized in the Netherlands. The number of associations is mentioned for each class. The first number refers to the number of syntaxa that are well developed, and the second number refers to the number that are not well developed (derivative communities) as a result of declining groundwater level, eutrophication, and fragmentation (aer Schaminée et al., 2010). Coastal and inland pioneer vegetation

Tidal flats Thero-Salicornietea

3+1

Spartinetea

2+1

Tidal mark Cakiletea maritimae

2+2

Salt marsh Asteretea tripolii

14+4

Saginetea maritimae

2+0

Beach dunes Ammophiletea

2+1

Inland pioneer vegetation and ruderal tall herb vegetation Asplenietea trichomanis

4+4

Stellarietea mediae

9+6

Bidentetea tripartitae

4+2

Isoeto-Nanojuncetea

4+1

Artemisietea vulgaris

9+9

Tall herb vegetation, shrub, and woodland

Tall herb vegetation Convolvulo-Filipenduletea

4+6

Galio-Urticetea

6+5

Epilobietea angustifolii

1+0

Shrubs Lonicero-Rubetea plicati

3+1

Franguletea

2+2

Rhamno-Prunetea

9+4

Moist woodlands Salicetea albae

3+2

Alnetea glutinosae

2+4

Vaccinio-Betuletea pubescentis

2+4

Dry woodlands Vaccinio-Piceetea

3+6

Quercetea robori-petraeae

4+4

Querco-Fagetea

6+3

1.3 Flora and vegetation 99

Two different approaches can be applied to describe a vegetation. One possibility is to look at the established vegetation as a whole, the other is to take the ecological requirements of the individual plant species as a starting point. Although the latter approach does not have a large tradition, it is of particular interest for the reconstruction of former plant communities (Cappers, 1994). The reconstruction of former vegetation as part of landscape reconstruction is often the aim in archaeological research. The primary archaeobotanical data, however, have to be considered as reduced lists of the former flora, quantified by the number of pollen and diaspores. The question is how to translate these data into a reliable reconstruction of the former vegetation, taking into account that they will have differed to an unknown extent from today’s vegetation. In deciding on a methodology that can be applied to the archaeobotanical record, it is important, first, to understand the methods that are used in describing a vegetation. The history of describing and classifying plant communities goes back to the early nineteenth century, when Alexander von Humboldt (1769–1859) suggested that growth-forms of plants could characterize plant communities. Several traditions in describing and classifying plant communities have developed in Europe since then, each with a different approach to vegetation description, including the Scandinavian school, the Danish school, and the French–Swiss school. One of the earliest more or less detailed descriptions of a plant community came from the Swiss geologist and naturalist Oswald Heer (1809–1883), who also produced the first archaeobotanical publication, dealing with the recovered charred and waterlogged plant remains from Neolithic lakeside dwelling sites (Heer, 1865). One of the schools whose method is widely adapted in Europe is the French– Swiss school, initiated by Josias Braun-Blanquet (1884–1980). The vegetation description of this school is based on the selection of a representative surface of a vegetation (relevé), which is used for a detailed description of the plant’s abundance in all vegetation layers. The size of the relevé depends on the species richness and species distribution. Its size can be determined by duplicating the smallest area that seems to be appropriate for a description and to fix the size by which the similarity between the different-sized relevés does not increase any more. This way of determining the appropriate size is similar to the way in which archaeologists fix the volume of an archaeobotanical sample that should be checked for its botanical composition (Table 9). The description of a relevé takes into account the setting in the landscape and the soil type, and it includes an enumeration of all the established plant species with their relevant ecological parameters. These parameters concern the layering of the vegetation, the abundance and degree of coverage of the species, the sociability, and the fertility and vitality (Schaminée et al., 1995). Most plant communities show a vertical stratification, and the number of levels is considered to be indicative of its complexity. The principal layers of a vegetation may include the moss layer, the herb layer, the shrub layer, and the tree layer. Abundance is a measure for the density (or number of individual sprouts) of the population in the relevé. Coverage of a species refers to the percentage of the surface that is covered by the above-ground plants parts. Sociability is an expression of the horizontal pattern of the species and is estimated by categorizing the degree of clustering. Fertility deals with the generative development, whereas vitality is related to the vegetative development.

100 Handbook of plant palaeoecology | Cappers & Neef

Table 9: Minimum relevé area values for various plant communities (m2) (aer Westhof & Van der Maarel, 1973). Plant community

Tropical swamp forest Temperate deciduous forest Steppe Agricultural field Calcareous grassland Heathland Hay meadow Mobile coastal dune Pasture Salt marsh Dune grassland Hygrophilous pioneer vegetation Terrestrial moss vegetation Epiphytic vegetation

Minimal plot size (m2)

2000–4000 100–500 50–100 25–100 10–50 10–50 10–25 10–20 5–10 2–10 1–10 1–4 1–4 0.1–0.4

The inventory of the vegetation in the vicinity of three Roman sites in the Eastern Desert of Egypt may serve as an example. Inventories have been made of the salt marsh vegetation and the vegetation in Wadi Mandit and Wadi Umm el-Mandit, all in the vicinity of Berenike; of the surroundings of the watering station (hydreuma) Kalalat, located at the transition between the coastal area and the mountainous area; and of the vegetation in Shenshef, located in the Red Sea mountains. The plant composition in the relevés is summarized in Table 10. Although the size of the relevés is relatively small for a desert environment, the number of species is surprisingly high. It varies from 5 to 19, with a mean value of 13. The plant cover can also be reasonable high, as it varies from 10 to 85 per cent, with a mean value of 37 per cent (fig. 99). It has to be realized, however, that these inventories were made after a heavy rainfall, resulting in exceptionally rich vegetation, which found its expression in both the plant cover and the proportion of ephemeral species. Figure 99: Relevé no. 11 (area: 4 m2) in the central part of a sandy wadi (vicinity of Berenike, Egypt; January 1977).

1.3 Flora and vegetation 101

Table 10: Floristic composition of 16 relevés (2 x 2 m2) in the vicinity of three Roman sites in the Eastern Desert of Egypt. Cover and number of individuals is indicated with a sign or number (r =