The extent to which CaS of the extensive pastures and Macchie reach along waysides far into the Nebiler settlement (cf. Fig. 5 and Table 2) confirms the wide ecological amplitude of natural vegetation in the area under investigation.

Since the vegetational surveys of Table 2 recorded hardly any Macchie species (but cf. chapter 5.6 and 5.8), it should be emphasised that even in the very centre of the settlement Macchie remnants, mainly of Quercus coccifera are to be found in the corners between houses and sheds and near walls on the field side.

The lianas Ephedra fragilis and Asparagus acutifolius also occur sporadically throughout the village at or on dry stone walls (field side).

Even species which are only found in the Macchie in a very devastated state, such as Quercus coccifera and Pistacia terebinthus were found in the village in some cases as tall trees (cf. the observations of Schwarz 1936). Thus in the northern section of the 2nd transect there is a larger stand of Pistacia terebinthus which is coppiced, and a little way away a 7m high Quercus coccifera tree with dense, hairy-white foliage.

There are clear differences in the site distribution of the Macchie remnants from the periphery to the centre of the settlement. At the outer margin, shrub complexes are spread around the cornfields in a mosaic pattern (cf. Figs. 4, 10 and 11) generally characterising piles of stones collected from the field, but the nearer they are to the centre of the settle-ment the more they are confined to the loose walls around the fields (cf. Fig. 11). The species which are most resilient and best able to expand remain pressed up against the walls. The Macchie species originally encountered on both sides of the enclosing walls disappear from the waysides near the village centre, and can only be found sporadically on the field-side of the walls.

The bush complexes of the Macchie remnants also display zoning (cf. Fig. 11). Thus the wayside polyhemerobic, trodden and compacted soils are included in a narrow zone of CaS of the trampling communities. This is followed by the heliophyte extensive pasture species in a well developed species spectrum in a more euhemerobic zone, which can reach to the foot of walls or can mediate to bush complexes with the CaS of the border communities.

Scabiosa reuteriana, Geranium purpureum var. and Tordylium aegaeum are as frequently represented here as outside the settlement. Beneath the bushes themselves are almost all the shade-requiring plants which are recorded together with the shrub complexes of the devastated Macchie. Species spectrum and soil situation here show the transition from euhemerobic to mesohemerobic conditions, as exists in mixed zone B (cf. Fig. 10) between extensive pastures and heavily devastated Macchie.

In the shrub and bush complexes of the fields and walls we find repeated on a small scale the hemerobic zoning which occurs on a larger scale for the settlement situation (cf. Fig. 5 and the following subsection). The rare untrodden sites, isolated and thus largely undisturbed in the settlement centre are almost as rich in species as complexes with similar habitats in the Macchie. Missing on the whole are the species of the fourth group (see Table 1) which have their centre of distribution in the Macchie or partially also in the border area of the vegetation patches (cf. third group of Table 1). Furthermore, only 50% of the species of the Cistus-Micromeria communities determined in the Macchie are also represented in the settlement area.

Relative to the area and the proportion of CaS of Macchie, Phrygana (Cistus-Micromeria communities) and the extensive pastures the overall species spectrum in the core of the set tlement declines rapidly and is dominated by the CaS of the ruderal and segetal communities.

In the research area there are essentially two forms of land-use and intensities, which are ordered in zones corresponding to the land-use gradients. The heart of the settlement with some buildings, piles of rubble, waste heaps and small garden allotments is the centre of highest eutrophication with the most intensive use, and characterised by the dominance of the CaS of the Chenopodium murale communities (cf. Fig. 12, zones A+B). Although durable weed communities are mostly encountered here (cf. Blume & Sukopp 1976), the inner settlement area must be classified polyhemerobic, if only because built-up areas, the pathway system with its intermittent trampling communities and the high proportion of hemerochores.

Even if the degree of naturalisation of the ruderal and segetal hemerochores cannot be further evaluated because of the lack of comparative investigations, it could nevertheless be proved that many of the segetal species only occur facultatively, and occur naturally in the original vegetation of the East-Mediterranean.

At its margins, with the adjacent farm land and grazing areas (of extensive pastures) the polyhemerobic settlement centre goes over to an euhemerobic zone (cf. Fig. 12, zone C), which is separated by the two major forms of use, dry farming (intensive) and grazing (extensive). The proportion of extensive pasture and Macchie species in the vegetation increase considerably, and the ruderal and segetal species in the mixed zone B (cf. Fig. 5) disappear completely in the transition from the eu- to meso-hemerobic zones. Using the distinction between alpha- and beta-euhemerobic stages made by Blume & Sukopp (1976), then the cornfields belong to the alpha-euhemerobic zone and the well-developed extensive pastures to the ß-euhemerobic zone. Ruderal communities are completely absent from the latter zone.

The ß-euhemerobic zone is essentially characterised by the high grazing pressure and trampling, which have led to a depletion of species and a negative selection of species in terms of grazing potential. The typical indicator plant for this stage of hemeroby, at least under the pertaining habitat conditions is Evax eriosphaera, which occurs massively (cf. also Tables 1 and 8). Anthropo-zoogenous influences almost completely govern the ecosystem of the extensive pastures, so that it is culturally determined as an alpha-euhemerobic zone.

The structure and floristic composition of the vegetation in the meso-hemerobic zone are largely determined by human activities. Mosaics of bush complexes with Daphne gnidium and a high proportion of thero- and hemi-kryptophytes settling the clearings (cf. Tables 1 and 8) indicate the intensity of use. Though the floristic composition of the vegetation in terms of phanerophytes and chamaephytes is not very different from the present-day potential natural vegetation or from the "climax plant community", the high level of species indicating degradation and the structural state of the devastated Macchie offer a good measurement criteria for the degree of hemeroby of this vegetation zone, characterised by Daphne gnidium.

The meso-hemerobic zone, largely determined by cultivation, is succeeded by relatively natural vegetation, which is very close to the presumed "natural potential vegetation". The floral composition differs only slightly from the Macchie to be expected under Pinus brutia forest. However, there are clear signs throughout this zone of anthropo-zoogenous influence from grazing and wood-cutting. Although there can only rarely have been any homogeneous vegetation without human influence around Nebiler, the high proportion of indicator species for an extreme stage of degradation and character species of the community are a good indicator of 'cultivation influence'.

From observation of Oberdorfer (1954) and Simonis (1954) about changes to the species spectrum in the course of a vegetation period, together with the investigations by Major & Pyott (1966) and Thompson & Grime (1979) of density and frequency of seed banks of grass and herb communities it seemed advisable to take soil samples for seed bank investigations.

Oberdorfer (1954b) had determined seasonal plant communities on the Balkan peninsula, but the germination tests carried out over a period of 1½ years failed to establish such communities for the area under investigation.

Although an average of 7.178 seedlings/m² were found, additional information regarding the dominance of well known species was not found and it was not possible to recognise any really new species combination (cf. Table 7). Of the species occurring in the seed bank (cf. Table 6) more than 90% belong to the therophytes, and nearly half of these to the CaS of the extensive pastures and Macchie clearings. Among the CaS of the latter group those the largest numbers of seedlings are mainly species which flower between February and May.

Even though it has not been clarified whether the seeds with the greater tendency to germinate came from the preceding vegetation period (cf. Thompson & Grime 1979), it is questionable whether such germination rates would be as high under natural conditions.

Continuous water supply and the absence of any of the disturbance experienced by the extensive pasture species throughout the winter, represent fundamental changes to the conditions under which the seed-bank annuals are competing. The investigations of Simonis (1954) established a close dependence of the large populations in plant communities with many therophytes (Brachypodietum ramosi and Barbula gracilis-Onobrychis caput galli assoc.) on the water supply and the resultant ability to compete.

Since, according to Simonis (ibid.) a marked decline of annuals tending to mass development at the onset of the dry period can be seen as a measure of vitality and competitive ability, then firstly the growth period of these species (e.g., Sagina apetala, Galium murale, Arenaria leptoclados, Psilurus incurvus, Veronica arvensis, Capsella rubella and Stellaria media) must be at the start of the year when precipitation is plentiful, and they must also be capable of establishing a permanent seed bank (Schneider & Kehl 1987). Once more it has to be pointed out that before the samples were taken the surface debris was removed (see chapter 4.6) to exclude temporary seed banks, so that all seeds collected must have already been in the soil, or at least just under the surface.

Returning to the mass development of some species, observations from Spring 1980 must also be taken into consideration. The much cooler and wetter April and May meant that many plants started flowering much later. Even eroded areas, completely free of vegetation in 1979, were densely covered in 1980, with Plantago cretica or Pl. lagopus frequently dominating. Above all Psilurus incurvus, found in the seed bank and easily overlooked on the ground, and also Poa bulbosa showed a regular distribution in the Macchie clearings, extensive pastures and wayside areas of the settlement. Sagina apetala, Galium murale, Arenaria leptoclada and Gagea bohemica, frequent in the seed bank investigations, were not found, but in the extensive pastures the circum-Mediterranean geophyte Ornithogalum umbellatum was frequent, though not determined in the vegetation survey in early 1979.

Although in the present investigation the species spectra for the 1st and 3rd transects were not changed significantly by the germination tests, these have highlighted many questions of vegetation dynamics. Of the some 10% new species only 2% occurred in more than 5 of the 40 soil samples between relevees nos. 20 and 88. Many species with high population densities in the vegetation surveys could not be determined by the seed-bank investi-gations. In particular Cynodon dactylon, Hordeum leporinum and the frequent Plantago species, Pl. cretica and Pl. lagopus or Evax eriosphaera would have been expected as seedlings. Whether seeds of these species only belong to the temporary seed bank, or were removed with the surface debris, or merely failed to germinate under the test conditions is unclear. Interestingly, Poa bulbosa only grew from bulbs, and no germination from seeds could be determined (cf. Sukopp & Scholz 1968).

The floral-geographic assignment of the most common species in the seed bank shows a remarkable similarity of potential distribution. Gagea bohemica and Erophila verna are elements of pontic steppes of the dry continental climate (Horvat et al. 1974); Sagina apetala is a middle European geoelement according to Walter & Straka (1970, from Horvath et al. ibid.); Psilurus incurvus occurs in the sub-Mediterranean areas of the Macedonian steppe (Horvath et al., ibid.), and Capsella rubella with Polycarpon tetraphyllum are widely distributed in north Mediterranean and middle European plant communities. Veronica arvensis also belongs to the middle and south European geoelements. The only purely east Mediterranean species found in the seed bank was Verbascum leptocladum, an endemic of the Pamphylian plain (Huber-Morath, correspondence).

In summary, the ability to germinate can be established for sub-Mediterranean, south European and Pontic steppes annual floral elements. However, this statement does not take account of the temporary seed bank, which was not registered, so that important information about the potential species spectrum is missing.

When establishing the seed bank as part of the vegetation, the aspects to be taken into account include the dormancy, the necessary resting periods, germination conditions and quantitative seed production of various populations as survival strategies.

Synecologically the vegetation is in a state of equilibrium, or in an artificial steady state between anthropogenous influence and regeneration potential which would otherwise lead to the formation of a forest (cf. Schwarz 1936: 417, Regel 1943: 83, Oberdorfer 1970: 277, Walter 1973: 129, Zohary 1973: 502, Specht 1981: 259).

The high level of species diversity in the degraded margins of the Macchie is a result of the numerous variety of habitats, often covering small areas which have often arisen with the development of mosaic of shrub complexes. This did not lead necessarily to an increase in the biomass suitable for grazing use, since the migration of ecologically more resistant species to extreme habitats such as Evax erisophaera to the extensive pastures, was accompanied by a process of negative selection (Zohary 1973: 652, Radke 1976: 411). Extreme xerophytes, avoided by livestock, could spread much more, and had a corresponding advantage over competition (cf. Greuter 1975: 162/190, Walter 1956: 271). This is basically true for all those indicator species of an extreme stage of degradation which together with Daphne gnidium characterise a species group combination. This also includes Verbascum leptocladum, which only infrequently shows bite marks. In order to assess the species spectrum and the high species diversity in relationship to biomass production relevant for grazing, it is necessary to conduct thorough investigations of grazing behaviour, also considering seasonal development of vegetation. Relevant work on the nutritional value of Mediterranean therophyte and chamaephytes and their selection under grazing pressure was not found in the literature.

The concept of hemeroby, introduced by Jallas (1956) and augmented and extended by Sukopp (1969, 1972) and Blume & Sukopp (1976) as a term for the totality of direct and indirect human intervention in and influence on the ecosystem, i.e., as a site factor, can well be used to classify the retrogressive site and vegetation development.

On the other hand the criteria developed for the classification of hemeroby in central and northern Europe cannot be transferred directly to the area under investigation, since the relationship of the extent and the significance of vegetation changes to the intensity and duration of anthropogenic influence depends upon the habitat in question and its vegetation (cf. Ellenberg 1954 in Sukopp 1972: 113). As frequently emphasised, a decisive factor for assessing the degree of hemeroby in the area under investigation is the robust nature and regenerative capacity of many Mediterranean species and the more original floral composition of the vegetation. Thus modified assessment criteria are required for determining the changes to the species stand due to neophytes (cf. Sukopp ibid.) i.e., to plant migration under human influence, and for deducing the significance of the proportion of therophytes in the spectrum of life forms, as well as the significance of soil forming and changing processes at anthropogenously influenced sites.