9 Investigating the Faunal Record from Bronze
Age Cyprus: Diversification and Intensification

Matthew Spigelman

Abstract

This paper uses data from published faunal reports to reconstruct strategies of faunal

exploitation practiced in Cyprus during the Early Cypriot – EC, Middle Cypriot – MC and Late
Cypriot – LC periods of the Bronze Age. Paul Croft published the data for Sotira and Marki (1996,
2003, 2006), David Reese for Alambra and Athienou (1996, 2005), Brian Hesse, Anne Ogilvy
and Paula Warnish for Phlamoudhi-Melissa and Vounari (1977, 1978), Rita Larje for Nitovikla
(1992) and Nils-Gustaf Gejvall for Kalopsidha (1966). The basic comparison of represented species
presented here between the EC and MC assemblages has also recently appeared in Paul Croft’s
discussion of the faunal assemblage from Marki (2006). My contribution in this paper has been to
bring together data from the EC, MC and LC periods, in terms of both species diversity and kill
patterns, and to contextualize these findings within models of animal husbandry practices and social
systems.

This paper consists of two sections. The first compares data on species diversity found at sites

of the EC and MC periods with that from the LC period. It is shown from these data that EC and
MC inhabitants practiced a diversified faunal strategy, utilizing a suite of animal resources, both wild
and domesticated. This diversified local economy contained inherent buffering mechanisms against
resource failure. It is argued that these internal buffering mechanisms obviated the need for large-scale
storage or extended social networks.

This diversified faunal economy is sharply contrasted by the LC data. LC data reveals a shift

to a faunal strategy focused upon a limited number of domesticated species, utilized primarily for their
secondary products. This intensified economic strategy allowed for higher productivity but removed the
internal buffering mechanisms provided by a diversified economy. It is argued that LC inhabitants
developed strategies of social and/or physical storage to buffer against the increased risks of catastrophic
resource failure from this intensified strategy.

The second section of the paper investigates the kill patterns for sheep and goats in the EC

through MC and the LC periods. Strategies of herd management can be reconstructed through an
analysis of the ages at which animals are killed (Silver 1963; Crabtree 1989). It is found that in the
EC through MC periods sheep and goats were utilized for both meat and secondary products. In the
LC period, however, herd management strategies shifted to the intensive production of secondary
products.

Species Diversity at EC and MC
Period Sites

Sizeable faunal assemblages have been
recovered and published from the EC
through MC period sites of Sotira,
Marki and Alambra. These assemblages

correlate well with each other for
species diversity (the number of species
represented) and richness (the relative
proportions of these species to one

120

another). Species variability found at
each site is measured using number of
identifiable specimens, or NISP. The
NISP measurement counts every bone
that can be assigned to a specific
element and species. As such it is well
suited to smaller assemblages, such as
some of the ones incorporated here.
The alternative measure of minimum
number of individuals, or MNI, was not
used as it can give distorted results for
small assemblages. The size of the
assemblages is listed on the chart below
the site names. Assemblage size is
determined by bone preservation,
recovery techniques, volume of
excavated material and habits of the site
inhabitants. No attempt has been made
here to disentangle these processes.

Moving chronologically, we will first
look at the EC assemblages from Sotira
and Marki. Both assemblages show the
utilization of sheep and goat, cattle, wild
deer and pig, in that order. These
similar patterns of species use
demonstrate that a common strategy of
faunal exploitation was practiced in the
EC period. The location of Sotira and
Marki in different regions of the island
suggests that this strategy was present
over a wide geographical area. The site
of Marki spans the EC and MC periods.
The faunal assemblages from these two
periods show continuity in both species
diversity and richness.

The MC assemblage from Marki can be
compared with the contemporary one
from Alambra. Here too we find
comparable species diversity. The
assemblage from Alambra, however,
contrasts slightly with those from Sotira
and Marki in richness, as deer are better
represented than cattle. At all three EC
and MC site, Sotira, Marki and Alambra,
we see a consistent pattern of species
diversity suggesting a common strategy

of exploitation that spanned great
temporal and geographical distances.

The EC to MC faunal assemblages
briefly described here are characterized
by species diversity. Species diversity is
a risk averse strategy that seeks to
buffer the effects of resource failure by
relying on a suite of resources. The
diverse EC-MC faunal assemblages
show evidence for two quite different
mechanisms of risk aversion. The first
of these is the keeping of a range of
domestic animals: sheep, goats, cattle
and pigs. Resource failure in any one
species would have been buffered
against by increased exploitation of the
others.

The second risk averse behavior is the
hunting of wild deer. Hunting would
have taken place outside of the agro-
pastoral economy. The regular
exploitation of wild resources by agro-
pastoral economies often serves as a
means of providing additional resources
during the period(s) of the year when
stored resource are running low and the
next season’s crop is not yet ready for
harvest (Stein 1989). The hunting of
wild deer may have served this purpose
in the EC-MC economy. Agro-pastoral
groups also exploit wild resources
during times of extreme resource
failure. Hunting most likely served
both of these roles in the EC-MC
economy. Hunting is a high skill
activity and as such must be maintained
and passed on to subsequent
generations on a regular basis. If
hunting is expected to form a buffering
mechanism in times of dire need then
the necessary skills must be maintained
on a regular, most likely yearly, basis.

Species Diversity at LC Period Sites

Moving chronologically forward into
the LC period we can investigate faunal

121

assemblages from the sites of Nitovikla,
Phalmoudhi-Melissa and Vounari,
Athienou-Pampoularis and Kalopsidha.

The plotting of species percentages for
LC sites reveals a shift to a faunal
economy focused almost exclusively on
sheep and goats. Sheep and goats make
up over 90% of the assemblages at all
sites save for Nitovikla. Cattle are
found in all faunal assemblages but at
exceedingly low levels. The hunting of
wild deer is only attested to at
Athienou, and only in very small
amounts. Pig remains are similarly
sparse, only seen by a small number of
bones at Melissa.

In contrast to the EC and MC periods,
the LC faunal economy is neither
diverse nor rich. It is instead
characterized by the intensive
exploitation of a limited number of
species, namely sheep and goats. This
dramatic shift reveals the abandonment
of the buffering mechanisms inherent in
a diversified economy. Sheep and goat
now dominate the pastoral sector, and
the practice of hunting wild deer has
been all but abandoned. An intensified
economy offers the prospect of more
efficient use of resources, both human
effort and grazing land, but sacrifices
security.

The shift to an intensified LC faunal
economy necessitated the development
of new mechanisms for buffering
resource failure. Increased physical
storage is evidenced at this time
through the appearance of storage
pithoi (Pilides 1996). Physical storage
can take place within social structures
that range from egalitarian to highly
ranked. The development of social
storage, however, requires the
formation of long distance
relationships. These relationships can
be called upon during times when

resource failure affects one local but not
another. Increased interaction between
communities leads to, and is aided by,
the development of distinct local
identities. This process of identity
formation represents the development
of horizontal social structure between
communities.

Herd Management Strategies at EC
and MC Period Sites: Sheep and
Goat Kill Patterns

The shift from the EC and MC to the
LC faunal economy also witnesses
changes to the strategy of sheep and
goat exploitation. In this second
section of the paper these shifting
strategies are modeled using evidence of
kill patterns as derived from data on
epiphysial fusion. Variable states of
epiphysial fusion provide a means of
assessing the age at which an animal
was killed. The ends of bones, the
epiphyses, fuse to the shaft of the bone
at a given point in the maturation
process. Some epiphyses are already
fused at birth while others do not fuse
until the animal reaches maturity. Using
known ages of epiphysial fusion for
specific bones the presence of unfused
bones provide a terminus ante quem for
the age of the animal at death (Silver
1963). Bones are grouped into early,
middle and late fusing elements
corresponding to 0-1 year, 1-2.5 years
and 2.5-3 years (Crabtree 1989). A kill
pattern for an assemblage can be
constructed by graphing the ratio of
fused and unfused bones for each of
these groups. At the left of the graph
the herd starts at 100%. The first data
point shows the percentage of the herd
killed before reaching 1 year of age.
The second data point shows the
percentage of animals killed before
reaching 2.5 years and the third data
point shows those killed before 3 years.
If the graph were extrapolated to the

122

right it would eventually reach zero
upon the death of all animals. The
graph represents an average of the kill
pattern for the entirety of the period in
which the assemblage was formed.

Kill patterns can be interpreted using
known lifecycle events and animal
husbandry practices, as observed
through ethnoarchaeological research.
Lifecycle events, such as age of
reproductive maturity, maximum meat
yield and decline in female fertility,
occur at predictable ages. The killing of
an animal either before or after any
such lifecycle event provides insight
into the goals of the exploitation
strategy.

Sheep and goats are notable in that they
can be utilized for both their meat as
well as for their secondary products,
such as milk and wool or hair. Sheep
and goats approach their maximum
meat yield at some time between 2.5
and 3 years of age (Redding 1981). An
animal killed before this point will
provide less meat while an animal
allowed to live past it will continue to
consume resources with little to no
additional meat yield. Animals that are
kept beyond 3 years of age indicate their
use in producing secondary products
and/or in maintaining the demographic
security of the heard through the
production of offspring.

The kill patterns of sheep and goats at
the EC and MC period sites of Sotira,
Marki and Alambra all show 10 to 20%
of animals dying before reaching one
year of life. These early deaths can be
attributed to natural causes and are
most likely an underestimate as fragile
infant bones are less likely to have
survived. Ethnographic accounts
indicate that on average 30% of sheep
and 45% of goats die of natural causes

within their first 6 months of life
(Redding 1981).

These same kill patterns show only a
small percentage, less than 5%, of the
herd killed between the ages of 1 and
2.5 years. This low amount is to be
expected in a meat producing economy,
as these juvenile animals have not yet
reached their maximum weights. The
kill patterns, however, change drastically
after 2.5 years, with one quarter to one
third of the herd killed between the ages
of 2.5 to 3 years. It is during this brief
period that killing an animal for its meat
is most efficient (Redding 1981).

The kill patterns from Marki and
Alambra indicate that 45 to 60 percent
of sheep and goats lived past the age of
three, reaching full maturity. Because
of the small size of the Sotira
assemblage it is not possible to know
what percentage of animals lived to
maturity there. Mature animals
provided both the reproductive capacity
to continue the herd and a source of
secondary products. Female sheep and
goats give birth for the first time at two
years of age and every year thereafter
until the age of 6 or 7, at which point
female fertility begins to decline. Milk
production is triggered in pregnant
females. Young animals require
mother’s milk until the age of 2 to 3
months of age at which point they can
be weaned and the milk taken for
human consumption. Additionally,
sheep would have provided wool and
goats hair, both usable for textile
production.

Similar exploitation strategies at all three
sites, Sotira, Marki and Alambra,
reinforce the view of self-reliant
settlements utilizing sheep and goat for
their meat and their milk, wool and hair.
There is no evidence for the over or
under production of any one of these

123

products, and as such there is no
evidence for the import or export of
goods. This finding does not
demonstrate that EC through MC
settlements were egalitarian
communities per se but it does indicate
an absence of mechanisms for the
production of significant amounts of
surplus through faunal exploitation.

Herd Management Strategies at LC
Period Sites: Sheep and Goat Kill
Patterns

Of the LC faunal assemblages only
those from Athienou and Melissa
contain sufficient evidence to
reconstruct the kill patterns of sheep
and goats.

The settlement site of Melissa shows a
kill pattern that indicates the intensive
exploitation of sheep and goats for their
secondary products. The kill pattern
reveals 20% of the herd dieing within
their first year of life, a similar
percentage to those found in the EC
and MC. A new pattern of exploitation
is revealed by the killing of some 30%
of the herd between the ages of 1 and
2.5 years. Equally dramatic is that all
animals that reach 2.5 years of age are
allowed to reach full maturity.

The killing of juvenile animals, under
2.5 years of age, sacrifices meat yields,
as these animals have not yet begun to
approach their maximum weight. The
killing of young animals, however,
serves to free up resources for the older
mature animals. These older animals
are the primary producers of secondary
products. If milk production by goats is
the goal then adult females are needed
and young males are culled. For wool
production by sheep, castrated adult
males are most productive so young
females are culled. These culling
practices allowed for the more efficient

converting of resources into secondary
products, however, they produce
skewed herd profiles (both for sex and
age) increasing the risk of demographic
collapse due to sickness or accidental
deaths.

The kill pattern for sheep and goats at
Athienou has a similar shape to that of
Melissa, however, it is located at the
base of the graph, as it is almost
exclusively comprised of young
individuals. The Athienou kill pattern
reveals that a full 85% of the sheep and
goats were killed before reaching 1 year
of age. Additionally, over 95% of the
animals were killed before reaching 2.5
years of age. Viewed on its own the
Athienou kill pattern represents an
unsustainable herd management
strategy. Interpreted in conjunction
with the Melissa data, however, it
reinforces the model of a faunal
economy focused on the intensive
production of secondary products at the
expense of meat production. The
young animals killed at Athienou would
have further helped to create a herd
composed of wool producing mature
male sheep and milk producing mature
female goats.

The assemblage from Athienou is dated
to the LC II period, however, a similar
assemblage of predominately juvenile
sheep and goats was excavated at
Kalopsidha-Trench 9 and is dated to
the LC I period. This assemblage was,
unfortunately, published without data
on epiphysial fusion, however, it was
stated that 77% of the animals were
“infantile or subadult” (Gejvall 1966).
While this method of classification
makes the reconstruction of a kill
pattern impossible, the high percentage
of young animals suggests that the
pattern shown at Athienou in the LC II
period can be extrapolated back into the
LC I at Kalopsidha.

124

As previously stated, the kill patterns of
sheep and goats at Kalopsidha and
Athienou represent an unsustainable
herd management strategy if viewed in
isolation. They are sensible, however, if
viewed as the transport of young sheep
and goats from other sites to specialized
– non-settlement – locations for
slaughter. This pattern raises the
possibility that Athienou and
Kalopsidha were aggregation localities
where members of different
communities came together to build
relationships between their
communities. We see at Athienou and
Kalopsidha an intersection of the duel
needs of the intensified LC economy:
the killing off of young sheep and goats
to conserve resources and the building
of long distance relationships to
facilitate the buffering mechanism of
social storage.

To summarize, the faunal economy of
the EC through MC periods is
characterized by the exploitation of a
diverse range of species in a risk averse
manner. This strategy included both

domesticated and wild animals. Each
site was able to provide its own internal
mechanisms of buffering resource
failure. The LC period sites show the
intensive exploitation of sheep and
goats for their secondary products.
This strategy was more efficient but
removed the buffering mechanisms
inherent in the earlier system. The
development of new buffering
mechanisms required increased contact
between settlements and strengthened
local identities. The new non-
settlement sites first seen in this period,
such as Athienou and Kalopsidha, could
have served as the location of inter site
interaction, facilitating the development
of social storage networks.

Acknowledgements

I would like to thank my advisors Rita
Wright and Pam Crabtree for their
helpful comments on this paper and
New York University for the Student
Travel Grant that has allowed me to
attend these meetings.

Figure 1 NISP for EC and MC period sites.
Data from Croft (1996, 2003, 2006) and Reese (1996).

0

25

50

75

100

Sotira-Kaminoudhia (EC)

(n=426)

Marki-Alonia (EC I-III)

(n=6107)

Marki-Alonia (MC I)

(n=4410)

Alambra-Mouttes (MC)

(n=991)

%

Ovis/Capra (Sheep/Goat)

Bos (Cattle)

Dama (Deer)

Sus (Pig)

Other

126

Figure 2 NISP for EC, MC and LC period sites.
Data from Gejvall (1966), Hesse et al. (1977, 1978), Larje (1992), Reese (2005).

0

25

50

75

100

Phlamoudhi-

Vounari (MC III -

LC I) (n=76)

Phlamoudhi-

Melissa (LC I)

(n=394)

Nitovikla (LC I)

(n=82)

Kalopsidha

Trench 9 (LC I-II)

(n=742)

Phlamoudhi-

Melissa (LC II)

(n=68)

Athienou-

Pampoulari (LC

II) (n=897)

%

Ovis/Capra (Sheep/Goat)

Bos (Cattle)

Dama (Deer)

Sus (Pig)

Other

127

0

10

20

30

40

50

60

70

80

90

100

Early Fusing

(before 1 year)

Middle Fusing

(before 2.5 years)

Late Fusing

(before 3 years)

Anatomical Elements

%

F

u

sed

Sotira - EC (n=43)

Marki - EC I - EC III (n=697)

Marki - MC I (n=356)

Alambra - MC (n=231)

Figure 3 Kill patterns of sheep and goats at EC and MC period sites.
Data from Croft (1996, 2003, 2006) and Reese (1996).

References

Crabtree, P. 1989. West Stow, Suffolk: Early Anglo-Saxon Animal Husbandry. East Anglian
Archaeology Report No. 47. Suffolk: Suffolk County Planning Dept.

Croft, P. 1996. Animal Bones, in Frankel, D. and Webb, J. (eds.) Marki Alonia. An Early
and Middle Bronze Age Town in Cyprus. Excavations 1990-1994. Studies in
Mediterranean Archaeology, Volume 123. Goteborg: Paul Astroms Forlag.

Croft, P. 2003. The Animal Remains, in Swiny, S., Rapp, G. and Herscher, E. (eds.)
Sotira Kaminoudhia: An Early Bronze Age Site in Cyprus. Cyprus American Archaeology
Research Institute Monograph Series, Volume 4. Boston: The American Schools of
Oriental Research, 439-448.

Croft, P. 2006. Animal Bones, in Frankel, D. and Webb, J. (eds.) Marki Alonia. An Early
& Middle Bronze Age Settlement in Cyprus. Excavations 1995-2000. Studies in
Mediterranean Archaeology, Volume 123, number 2. Goteborg: Paul Astroms Forlag.

Gejvall, N-G. 1966. Osteological Investigations of Human and Animal Bone Fragments
from Kalopsidha, in Astrom, P. (ed.) Excavations at Kalopsidha and Ayios Iakovos in
Cyprus. Studies in Mediterranean Archaeology, Volume 2. Lund: Paul Astroms Forlag,
128-132.

Heese, B., Ogilvy, A. and Wapnish, P. 1977. The Fauna of Phlamoudhi – Melissa: An
Interim Report. Report of the Department of Antiquities, Cyprus, 5-29.

Heese, B., Ogilvy, A. and Wapnish, P. 1983. Report on the Fauna from Phlamoudhi
Vounari, in Al-Radi, S. (ed.) Phlamoudhi Vounari: A Sanctuary Site in Cyprus. Studies in
Mediterranean Archaeology, Volume 65. Goteborg: Paul Astroms Forlag, 116-118.

Larje, R. 1992. The Bones from the Bronze Age Fortress of Nitovikla, Cyprus, in Hult,
G. (ed.) Nitovikla Reconsidered. Medelhavsmuseet Memoir 8. Stockholm:
Medelhavsmuseet, 166-175.

Pilides, D. 1996. Storage jars as evidence of the economy of Cyprus in the Late Bronze
Age, in V. Karageorghis and D. Michaelides (eds.) The Development of the Cypriot Economy
from the Prehistoric Period to the Present. Nicosia.

Redding, R. 1981. Decision Making in Subsistence Herding of Sheep and Goats in the Middle
East. Unpublished PhD Thesis, University of Michigan.

Reese, D. 1996. Subsistence Economy, in Coleman, J., Barlow, J., Mogelonsky, M. and
Schaar, K. (eds.) Alambra: A Middle Bronze Age Settlement in Cyprus; Archaeological
Investigations by Cornell University. Studies in Mediterranean Archaeology, Volume 118.
Goteborg: Paul Astroms Forlag, 217-226.

Reese, D. 2005. Faunal Remains from Israeli Excavations at Athienou-Pampoulari tis
Koukkouninas. Report of the Department of Antiquities, Cyprus: 87-108.

129

Silver, I. 1963. The Ageing of Domestic Animals, in Brothwell, D. and Higgs, E. (eds.)
Science in Archaeology. New York: Praeger, 283-302.

Stein, G. 1989. Strategies of Risk Reduction in Herding and Hunting Systems of
Neolithic Southeast Anatolia, in Crabtree, P. and Campana, D. (eds.) Early Animal
Domestication and Its Cultural Context. MASCA Research Papers in Science and
Archaeology. Special Supplement to Volume 6, 1989. Philadelphia: The University
Museum of Archaeology and Anthropology, University of Pennsylvania, 87-97.