nevertheless more dispersed in this rural part of lowland Macedonia,
with a distribution of 1.4 sites per square kilometre. Crossing the ridge of
mountains that separates the Langadas area and the Echeidoros valley
from the River Strymon, surveys conducted since the 1970s have pro-
duced a slightly greater concentration of population—2.14 sites per
square kilometre around Sandanski, just north of the Bulgarian–Greek
border, 2.7 sites per square kilometre in the Blagoevgrad area farther
north. This can be compared with a maximum concentration of 2–3 sites
per square kilometre in sampled areas of central Bulgaria.^55 The statistics
of ceramic scatters may translate into a range of historically documented
locations, not all of which were necessarily habitation areas.
Any attempt to relate survey data with known historical place names,
or with known types of site, is not particularly meaningful without some
consideration of land use and local ecologies. The phenomenon of
settlement mounds (referred to astoumbasortells, depending on local-
ity) is probably the most graphic illustration of fundamentally different
ecologies in northern Greece and the east Balkans, compared with
central and southern Greece. Such man-made mounds are still promin-
ent topographic features, notwithstanding erosion over many centuries.
These mound settlements, whilst being the most easily recognizable sites
in the Bronze and Early Iron Ages, were by no means the only kinds of
site identified from these periods, as rescue archaeology and intensive
surveys have shown. Since the summits of these mounds are often
eroded, leaving few clear traces of continuity into historical times, and
there are usually few traces of related activity in their immediate vicinity,
the challenge for researchers has been to try and understand how such
compact, conservative patterns of community organization evolved into
new organizational structures during thefirst half of thefirst millennium
bc, and how these new forms of collective organization relate to the
urban patterns discernible from the sixth centurybconwards. This
challenge applies as much to inland and Aegean Thrace, as it does to
lowland (and upland) Macedonia.^56
(^55) Śliwa and Domaradzki 1983 (joint Polish–Bulgarian survey of the middle Strymon
valley); Domaradzki 2001; B. Athanassov et al.,AOR2009 [2010] 676–8; 2010 [2011] 555–8;
AIA/APA Annual Meeting, Philadelphia 2012 (preliminary report); Kazanluk area survey:
G. Nehrizov and J. TsvetkovaAOR2008 [2009] 746–76; 2009 [2010], 144–6, 649–51;
G. Nehrizov, J. Tsvetkova, A. Sobotkova, E. Bozhinova,AOR2010 [2011] 571–4; Nehrizov
and Tsvetkova,AOR2011 [2012] 550–2.
(^56) Baralis 2010, 248–50, 256 (on settlement dispersal along the north Aegean coastline).
Herdsmen with golden leaves—narratives and spaces 67