 Good morning, my name is Brent Whitford. I'm a PhD student at the University at Buffalo Effective last week, so brand new to the program. Thank you all for coming out, and I hope that you enjoy the presentation My talk today will focus a little bit on the Balkan Peninsula, but more specifically on the Strumer River Valley The Strumer River Valley is located in southwest Bulgaria and northern Greece It extends 415 kilometers from north to south and effectively transects both the Mediterranean and temperate climatic and vegetational zones As one of the earliest regions settled by farming communities During the Neolithic period of Southeast Europe at around 6200 BC It represents a transitional area in which the first agricultural communities may have adapted to more temperate environmental conditions Such adaptations, it's been argued, preceded the expansion of early agricultural practices further north into continental-type climates So in this paper, I will elaborate on the role of the Strumer River Valley as such a probable transition zone during the early Neolithic period of Southeast Europe and Specifically, I'll assess its potential to stimulate adaptation and transmission As early as the late 1960s It was recognized by Nandrists that the Strumer River Valley played a particularly important role in the advance of Neolithic lifeways From the Aegean to temperate Europe and more recently It's been increasingly hypothesized that the transitional ecological characteristics of the Strumer River Valley Served as a stimulant to adaptation and transmission However, before such a hypothesis can be sustained The ecological characteristics of the Strumer River Valley must first be compared with that of the neighboring regions In order to assess whether adaptations derived from settlement in the Strumer River Valley could have in fact been transmitted to other regions With little to no adaptive difficulty So all regions having similar ecological characteristics with the Neolithic niche of the Strumer River Valley May then be interpreted as suitable locations from which to exploit related adaptations And thereafter evaluated against the cultural evidence on the Neolithic process of Southeast Europe more generally So in order to do this So in order to compare the ecological characteristics of early Neolithic settlement in the Strumer Valley with that of Southeast Europe as a whole I use an approach based on species distribution modeling Maximum entropy species distribution modeling or Maxent for short lends itself quite well to such tasks Essentially the Maxent algorithm Quantitatively identifies the impact and relative influence of different ecological factors over the determination of settlement locations The output of the Maxent procedure consists of a logistic habitat suitability grid ranked in percentage that is graphically projected onto the study region Those areas having a habitat suitability score of greater than zero percent Can then be interpreted as inclusive within the ecological niche of settlement distribution further excuse me Furthermore Maxent also incorporates functions that serve to model the potential distribution of a population Beyond its current known boundaries So by first training the model according to the specific boundaries of a known settlement distribution Maxent can subsequently project the results across a wider region and effectively identify all areas having similar ecological characteristics It then follows that those areas having similar ecological Characteristics should be suitable locations to which the original population may be transmitted while facing little to no adaptive difficulty I therefore train the model using only the characteristics of early Neolithic settlement in the Strumer River Valley and Subsequently project these results to the whole of Southeast Europe To verify whether there are any similarities All right, so in order to produce a Maxent model You need two things the first is a data set consisting of the locations of archaeological settlements And the second is a data set consisting of ecological variables that represent the study region Regarding the locations of early Neolithic settlements the early Neolithic period is confidently attested at 33 Archaeological sites in the Strumer River Valley. I retrieved the precise coordinates of each of these sites based on a rigorous review of published maps and Archaeological site registry is located within the relevant regional museums Regarding the ecological variables. I started I selected 10 climate and terrain variables in order to represent the ecological setting of Southeast Europe In short these consist of elevation above sea level terrain slope topographic ruggedness index Topographic wetness index taxonomic soil type mean annual temperature temperature annual range temperature seasonality precipitation seasonality and annual precipitation Present day values were selected as they closely approximate the conditions of the region post 6200 BC Furthermore, I selected these particular variables because of their straightforward affiliation with regards to their non-mediating effects on The composition of domestic plant and animal assemblages as well as for their relevance with regard to the suitable placement of sedentary agricultural settlements a Second step in the data selection procedure involves verifying for instances of statistical Colinearity between the ecological variables So as not the bias the model in favor of collinearly related sets You can think of this as elevation above sea level Typically is correlated with temperature. So like the higher you go the colder it gets Usually in this case the two variables vary in a linear way and if you include them both in your model You're going to bias you're going to favor those collinear sets So first you should check for statistical collinearity and Remove such variables So all continuous ecological variables were subjected to Pearson's test for Collinearity one half of each pair presenting an r-square value greater or equal to 0.8 That's a fancy way of saying that they were collinear was removed I prioritized the removal of variables that were found collinear in more than one instance and In this case soil type was the only variable excluded from the test because it's represented as categorical rather than continuous and Therefore can't be collinear in the Luke Euclidean space So so following the results of Pearson's tests of the original 10 that I selected six remained These are slope Topographic wetness index annual mean temperature temperature annual range Precipitation seasonality and soil type in order to assess the influence of the ecological setting on determining the locations of archaeological Settlements the software max and randomly samples background points from across the study region and compares their values against that of the Occurrence points or the locations of archaeological settlements Considering that my goal is to generate a habitat suitability grid based only on the distribution of settlements in the struma valley I Produced what's called a bias grid in order to isolate the struma valley and a small portion of the surrounding regions Which accounted for a total of 18,000 square kilometers That's the small bounding box you can see on the screen is the bias grid So this ladder bias grid serves to restrict the location or pardon me the selection of background points to within that bounding box Although the software recommends sampling 10,000 Pseudo absences from background considering that my added a regular relatively small sample size of archaeological settlements And that the bounding boxes also doesn't cover a very big area I chose to reduce the sampling to 1000 background points to avoid excessive model overfit The Maxent modeling software also allows for the adjustment of additional parameters that can impact model performance In order to aid with the selection of appropriate modeling parameters muskarela at all developed an E and Pardon me the E and M eval package for our studio By running the E and M evaluation function that's included in the E and M eval package. It's very difficult to say The Maxent model is run across Pardon me each possible combination of modeling parameters and provides evaluation metrics pertaining to overall model performance test and training performance as well as on test point omission rates, so The combination of parameters yielding the lowest omission rates and Mean difference in model performance paired with the highest overall model performance should then be selected for producing the final model In this case, I selected the combination of linear quadratic hinge and product feature classes at a regularization multiplier of 2.5 in accordance with the results of the E and M evaluation procedure So the resulting habitat suitability grid provides some valuable insights into the ecological characteristics of early Neolithic settlement in the Strummer River Valley That can thereafter be evaluated more widely in terms of its potential for adaptation and transmission during the Neolithic process of Southeast Europe High habitat suitability values are driven by a moderate to high topographic wetness index a moderate contrast and precipitation seasonality a high temperature annual range for settlement on soils of the Louvysol, Fluvisol and Vertisol types for settlement on flat terrains and in regions having a moderate mean annual temperature When taken in unison the characteristics of early Neolithic settlement in the Strummer Valley Therefore indicate a settlement distribution that is primary primarily defined by a temperate continental type climate and mediated by topographic and soil type considerations Therefore the ecological characteristics of early Neolithic settlement in the Strummer River Valley Strongly suggests that adaptation to more temperate climates and to harsher seasonal conditions had either already begun And or was fully developed by the time that early agricultural practices spread throughout the valley Here I need your help a bit I rely on your knowledge of the geography of Southeast Europe to more or less identify the different countries in this In this figure, of course, I'm willing to discuss during the discussion period if you have any questions anyhow So furthermore the ecological characteristics when projected to the whole of Southeast Europe are found to share a great deal of similarities with many of the neighboring regions Considering the the entire extent of the habitat suitability grid so those areas with a greater than zero percent habitat suitability regions with a comparable ecological setting include Central and West Greek Macedonia Bulgaria the Republic of Macedonia Kosovo Serbia the vast majority of Hungary the continental regions of Slovenia Croatia Bosnia and Herzegovina Montenegro and Albania as well as the extreme West and Southeast of Romania Basically everywhere that isn't blue on the figure in each of the latter cases the only exceptions are the high reaches of the Alpine regions It should be noted that the habitat suitability grid also encompasses some minor portions of Austria and Italy and Appears to continue to the Northeast into Moldova and the Ukraine although the latter regions were not fully included within the boundaries of the study area and Thus cannot be properly assessed In total the habitat suitability grid is distributed over 37 percent so over one-third of the ecological variability present in the whole of Southeast Europe Indicating that early agricultural practices could have generally been transmitted to a wide range of areas from the Strummer River Valley with little to no adaptive difficulty Regarding now the areas in blue or those having zero percent probability for settlement the whole of Greece Excluding only central and West Creek Macedonia is found to share no strong similarities with the distribution of settlement in the Strummer Valley The latter is significant considering that central Greece was the first region of continental Europe to have been settled by early agriculture Communities and therefore indicates that a period of significant adaptation to more temperate conditions must have indeed taken place elsewhere Furthermore the exclusion of East Greek Macedonia and Greek race I suppose this part here There's some interesting implications when one considers that the early Neolithic period has yet to be confidently identified in either of these regions Also, the exclusion of the Eastern Adriatic lend some support to the hypothesis that ecological factors Distinguished in the other neolithicization process from that of temperate Europe which there is characterized by the impressive complex I'm going to speed up a bit because I was just given a two-minute warning, but in general that whole distribution Coincides with the distribution of what's called the volcano Anatolian type white on red painted ceramics So essentially the hallmark of the Neolithic of Southeast Europe I'd like to take a moment however and talk about the areas that only have a 50 percent suitability or more and to summarize basically what that means is When we look at a more refined distribution or the stronger niche characteristics of the Strummer Valley It coincides mostly with the areas that we know as the starchable complex to the west of the Strummer Valley and Of the Kremikovci-Karanovo one complex to the east The only exception here is the region of Dobrogya and Romania which to date has not yielded any evidence of early Neolithic settlement to my knowledge So what that means to me is that rather than a north-south conduit the Strummer Valley rather Was more of an east-west sort of transmitter during Southeast The Neolithicization process of Southeast Europe And that changes really the discussion of what people normally say that it's a that it was a north-south conduit and so on so forth most interesting is the the complete Absence or exclusion of the regions of the Kouros and Krish culture and Hungary and Romania Which typically are taken as coinciding within a larger archaeological complex called the starchable Kouros Krish In this case you can see that ecological factors more or less Strongly distinguished between the three and that maybe we should start thinking of them in in different terms in terms of the temperate So in conclusion the ecological Characteristics of early Neolithic settlement in the Strummer River Valley share a great deal of affinity with that of temperate Southeast Europe as a whole thereby supporting the hypothesis that the Strummer Valley could have served as a region of adaptation and transmission during the early Neolithic period Though in this case no more or no less than many of the other regions It is significant that the ecological characteristics of the Strummer Valley and Encompass almost all of the regions in which the earliest pottery of the Balkano Anatolian type is distributed However, the more refined hypothesis that the Strummer Valley was well positioned to stimulate the spread of early agricultural practices northward Is here tentatively undermined When one considers only those regions having a habitat suitability score of greater than 50 percent the Strummer Valley rather shares a greater affinity with the Distribution of starchable and Kamikochi kind of a one-type sites thereby suggesting that the Strummer Valley served primarily as an east-west conduit of adaptation and transmission rather than northward Thank you