 Hi, everyone. Welcome to the Virtual Archaeological Research Facility. In case I don't know any of you personally, my name is Lucy Gill, a PhD candidate in anthropology here at Berkeley, and I'm one of the brown bag organizers this semester. Before I introduce our speaker today, I will begin with a land acknowledgement modeled on the statement developed by the Native American Student Development Office in partnership with the Moacma Ohlone tribe. We consider this a working formulation to be replaced with language reflecting the particular position of the ARF community and developed in collaboration with appropriate stakeholders. The Archaeological Research Facility sits on the territory of Huchin, the ancestral and unceded land of the Chichenyo Ohlone, the successors of the historic and sovereign Verona band of Alameda County. We acknowledge that this land remains of great importance to the Ohlone people, that every member of the ARF community benefits from the continued occupation of this land, and that it is our responsibility to support Indigenous sovereignty and hold the University of California accountable to the needs of American Indian and Indigenous peoples. So before we get into this week's brown bag, I'm going to turn it briefly over to Sarah Kanza to make announcements. Thanks Lucy. I just wanted to flag really quick that next week, March 11th is UC Berkeley's Big Give. It's an annual online fundraising extravaganza and we are participating this year with a triple match that's been offered by two anonymous donors in support of archaeology at Cal. So come back to our website and you'll also be getting emails from me before Big Give happens. But on March 11th, if you go to give.berkeley.edu slash ARF, you can make a donation and your donation will be tripled in support of archaeology. Thanks for that Sarah. So before I introduce our speaker today, I just wanted to say that next week we will have the pleasure of being joined by Katie Kinkoff, who will be talking about her experience as a new professor in the CSU system and her ongoing research on materializing disability justice. So please join us for that back here at the virtual ARF. Today we are joined by Jordan Brown, who is also my brown bag co organizer this semester, and he is a geo archaeologist working on his PhD in the Department of Anthropology at UC Berkeley. He works primarily in pre and proto historic Southwest Asia, in what is now Eastern Jordan and Northern Iraq. However, he also has an abiding interest in statistical methods in archaeological data analysis, and the integration of archaeological ecological geomorphological and climatic data. So while I look forward to hearing about his work in Southwest Asia some other time. Today we'll be hearing about a project a little closer to home about a novel statistical approach for determining the harvesting seasonality of shellfish on the California coast. This is a collaborative project drawing on many different data sets. So towards the end of today's talk will be joined, I believe by a couple of guest stars, familiar faces to many of you so definitely stay tuned for that. Also this work will be presented at a session at the Society for California Archaeology meetings this Friday afternoon, along with other work by folks from the Cal lab and the almond and tribal band so check that out. And without further ado, take it away Jordan. It's a pleasure to be here and something other than my usual capacity. And of course, do stay tuned for those, those special guests. You saw one of them on the screen just now. But yeah so so what I'll be talking about today is a case study and archaeological quantification, looking at shell fishing seasonality on the California Central Coast during the Holocene. And there are a lot of different aspects to this project and to the research that's related to it. I'm going to be focusing fairly narrowly on the isotopes and how to quantify them, since that reflects how I got involved in this project. But before I launch into that I do want to just give a quick thanks to the Cal lab for getting me involved in the first place it's about a year ago now that we started chatting about this kind of stuff. And the SCA presentation session that Lucy mentioned has been about that long deferred. So this has turned into a longer project which has been fortunate. And I will agree. So yeah very very many thanks to Kent and Rob Cuthrell and, and, and microphone and Alec Apodaca for their work on this and you'll get to hear from some of them toward the end. So, why are we interested in this study here. We're interested in in shellfish on the coast in this case we're talking specifically about California muscle which are really sort of a dominant rocky intertidal species here and and very very important in in the gathering that that we see throughout the Holocene important research resource for folks. The sort of best single argument perhaps you can make is there are enormous shell mounds all along the coast. And shellfish feature very prominently and that and that's sort of a, you know, a shorthand for a lot of other things that are going on with these with these organisms and people's relationships to them and the environments and what they in which they grow. And that that we're going to be interested in here, some sort of broader questions to be thinking about again I'm going to be focusing fairly narrowly on sort of this as a math problem. But we'll get hints at some of the other other stuff. I was thinking about how does shell fishing seasonality harvest seasonality change through time how is it related to variation in space across a region. Or, you know, how are these activities concentrated during the year that's the basic meaning of seasonality that I'm using here, not strictly and you know one of four seasons. And how does that relate to all manner of social factors, trade networks, economic structures, food ways, and anything you like like there's this is pretty deeply integrated in life on the coast. And then, finally, and, and, and perhaps, most importantly is, you know, looking at these long term archaeological data sets. Particularly that are, you know, these are the sites and the ancestors of the folks in the onwards and tribal band and and looking for you know lessons from those ancestors for how we steward the coast today, and particularly how the members of the Donald Mitzen can be involved in that and in fact leading those efforts by looking to these long term data sets. So, I will give a brief overview of what's going on in in seasonality determination from shellfish. The basic thing is that we're drilling out samples of the shell carbonate. And relying on the fact that there is a very well studied relationship between the oxygen isotopic content of carbonate that's precipitated by by mollusks in this case California muscle. And the temperature of the water in which that that carbonate is is precipitated. And, and so this allows us to look at temperature changes over the life of a muscle. Very, very far back in time. So this is a little picture of our basic scam sampling scheme courtesy of Alec Apadaca and and a little picture of a thin section as well where you can see sort of the different growth bands. These creatures grow by accretion one band on top of the other as they as they age. So we're kind of creating a biotic calendar in a way of sea surface temperature in the form of carbonate oxygen isotope composition. Now, I want to give a quick history of seasonality analysis using isotopic data sets. And it is a it has a long history at least relative to the relatively young field of of stable isotope geochemistry. One of the very first papers that discussed this carbonate oxygen isotope paleo thermometer as it's referred to is from Yuri at all this is the dude who invented radio carbon dating so thanks Harold. He was very, very busy. And, and in this paper they sampled from the growth rings of a fossil Belknight. And I'm not sure if it's exactly a specimen but some this species became the sort of standard for isotopic analysis of carbonate in general anyway. They plot up their values and they get this nice curve, and they say, we think this is seasons that's why we see WS here for winter summer. And so that was an exciting thing to realize and couple of decades later, Nicholas Shackleton realized that not only was that a fancy thing for paleo climate studies, it was also great for archaeological studies. And this is a little graph from the very, very first archaeological isotope seasonality study where Shackleton sampled modern shells of a particular genus or species. I'm forgetting the name at the moment, not my delis California honest not not California muscle and establish sort of what the distribution of isotope values was over the whole course of the year, and then sampled some archaeological specimens and said look, here's the values cluster on one side of the year. This is a, a positive isotope value so this is, this is cold. I won't go too deeply into into that but winter harvesting is what you conclude. And then, somewhat after that, killing Lee and burger and then burger perhaps, and then killing Lee. Again, both both geochemists with some archaeological interests. Take a look at actually specifically California muscle in Southern California and say, we think we can do one better than shackleton shackleton just said this harvesting was happening during cold part of the year I can't say exactly when. These analysis are too uncertain. killing Lee says, no, look, if we take a bunch of samples, not just, you know, a couple from the very end of the archaeological shell. If we take many, we can trace this whole curve through a long span of time. We can make some assumptions about how fast the shell is growing but, but we feel we can do this and then we go to our archaeological shells and say, huh, where in the squiggle that they end up. This is where they're harvested and made estimates of a monthly resolution on the basis of that. Shackleton didn't love this, and thought that they were spending way too much money on sampling for conclusions that were far too uncertain, and in an attempt to demonstrate this deep and colleagues colleagues included Shackleton plotted, samples on the same day from, you know, shells, modern shells, and said look, all three of these isotope profiles were harvested on the same day. Do you really think you can say to a monthly degree of resolution, you know what, when when a single archaeological shell was harvested. So that's their their skepticism here. I would say that there's a little bit of shape similarity in this and killing me as argument was about shape. So, I think it doesn't fully end the end the discussion as you'll see my opinion is today. So, let's bring us up to the present day very quickly. The two major techniques in use are so called terminal growth band plus x sampling tgb plus x which means you take the final growth band value one isotope sample here, and then you move back at regular intervals say one millimeter two millimeters two millimeters and, and you plot those values in a sort of time series. And this spacing is relatively course, you can see here. And some other authors have pointed out that you might be missing important parts of the trend important parts of the shape. And, and so there are other two advocate micro sampling and micro milling. In order to really resolve as clearly as possible, the isotopic variation in shell over time. But there's a trade off here and the authors that I'm mentioning, recognize this, because this approach is a lot more costly and a lot more time consuming. And it provides some benefits but then you have a real trade off to think about in regards to archaeological site and and region sampling because if we're trying to say something about how people interact with the resource. It's all along the coast, then just getting one shell right is not enough. So that's why I am interested in the terminal growth band tgb plus x approach. If we can do a better job of understanding the uncertainties that are involved here and of trying to get useful estimates of seasonality out without sort of pre dividing up the year into sort of bins that we can see easily in the samples. This is a problem that Shackleton confronted, and that everybody who has ever done a tgb plus x study has confronted. You're going shell by shell saying like, Okay, I see a high value at the start high temperature value once I've converted it from the isotopes. And then it goes up, but then it goes down, and I'm guessing at how much time that includes and so I'm trying to match that up to seasonal curve but there's a lot of sort of expert judgment required and that's also hard to scale. If we want to do a large sample study. So, what are our goals in in sort of revising these methods or perhaps trying to build upon them as a better way to put it. The the sort of King King goal perhaps that that all these things could fall under is reproducibility. So I think that's the concept of other researchers being able to reanalyze your data using your same methods but perhaps varying some assumptions. And part of this is automation, making sure that it's not just you saying here I plotted this thing, but trying to make that an automatic process if you can, which also helps with scalability. And then there's a little on the uncertainties involved because we know there's a great deal of uncertainty. Even just the baseline we're using for the annual sea surface temperature curve in our region of interest is an important decision see that later. And then also transparency and accessibility having, especially because this is a collaborative process in many ways, not only with other researchers but also with the woods and tribal band and directed towards you know their stewardship the folks who are, you know, sort of having who are our audience here are both, you know, all these groups really need to have the ability to evaluate what we know and how well we think we know it in in an interactive and and sort of clear cut way so we don't end up with like isotopes just seeming like you know this is the dictate from on high and either take it or leave it but yeah so this is this is an important thing we'll see coming up throughout. So the very basic thing that you've got to do is when you're trying to get seasonality out of shell isotopes is get from the isotopes to see surface temperature which we have an equation for and then from distance along a shell into calendar time, which is a bit of a more authoritarian problem for this pilot study that I'm going to be talking about today, we make a simple assumption of a constant on average growth rate of millimeter and a half per month, which is plausible for California muscle on the central but there are lots of things that I would like to look at me more detail about this later. But for now that's what we do and this can get you from a comparison of your archaeological oxygen isotope data shell to a historical reference data set of sea surface temperature, which we have thanks to the Hopkins marine station at Pacific Grove at the southern end of our day. So to give a sense of the specifics of our approach that's sort of the general things that you've got to do on one way or another but what sort of particular about what we do is to take our historical data and transform it into simulated isotope space, which is to say that since we have this historical data set of what sea surface temperature was doing over a long period of time. We could, in principle predict what a shell should look like that was art but you know harvested on any given day, given the assumptions of our sea surface temperature and isotope relationship and of our growth rate assumptions for for this tax. So if we do that, then we can say okay well here's the temperature values that should be relevant to these different growth bands that we're sampling and our TGB plus four in this case approach so terminal growth band plus four others back in time. And here's what a shell that's harvested on that day I think it was. Oh, I think it was like may first but I picked what that isotope profile should look like. And we can do that for a whole database. And so you've got, oh, may 10. Here's your May 10 squiggle and isotope values that you hypothesize. Here's a September 11 here's January 13 these are all totally arbitrary just pulled out of the broader data set I do this for the whole, the whole like 100 year span that we have a Pacific Grove, this temperature data. And as you can see these are the different growth bands time proceeds from left to right. And, and then the next thing that you can do, which is also somewhat particular to our approach is to standardize each of these growth profiles simulated shell growth profiles. And say, we know that there's, you know, perhaps like some overall, you know, centennial trends in sea surface temperature change, we don't want those to affect our seasonality analysis. So we're going to just say we want the shape. We don't care about the specific values so we're just getting the shape out here, plus or minus from the average it's very simple approach. And then we take our archaeological shell over here. I believe it's issue Shell number nine. And we compare it to our historical database and say, does it match. Where can we find matches. And then what we do is we plot up those matches because we know the day on which those simulated shells were simulated to have been harvested. And then we make that plot of like here's here's where we find matches for shell number nine. And so we see, you know, this histogram presenting certain peaks throughout the year. And that gives us a guess at when this shall be harvested. So then we do that for all 40 of our specimens or potentially quite many more. This is really just the click of a button. That's a nice automation thing. And we get a sense for between these different sites. The three sites are SCR seven SCR 14 SMA 216 SCR seven is a middle Holocene site. You know, on the on the order of 6000 years ago, 64 or so. And SCR 14 is as well as SMA 216 are both late Holocene sites. So primarily over the last last millennium. And all these sites are just north of Santa Cruz. So this is worth noting on the opposite side of Monterey Bay from the temperature data set and in a more open coastline environment. So anyway, this is the histograms that we see. And I submit that this is real uncertainty about when these shells were harvested. What we can then do is combine those histograms into general pictures of harvesting activity as we think, you know, based on this very small sample size. In these sites. Yeah, like it's going on. Well, I like I'll, I'll, I'll bring you and you and Mike on in a little bit sorry to give give away our special guests. But if you want to hop off for the moment you can now you got zoom pro over here. But anyway, so we see these sort of seasonal curves from the histograms here is if we plot a kernel that's kernel density estimate over those data points just really interpolating densities to them. And we see some trends here. And now that's exciting the the basic basic takeaway here is kind of kind of works. These might be real things but I'll come back I'll come back to that in a moment. So just another fun thing that you can do if you assume that if you go back to these plots and you say well let's take the modal value for each of these shells, the highest sort of count of matches on these in these histograms. For our archaeological shells, then you can take that as like our best guess at the, the specific time of year. That each of these shells was harvested and if you do that, then you can sort of count back along this growth band here right, and you can say okay well we assume that date, because from our the mode of that histogram, then we go back and we can say based on our growth rate how far we think that that was, we can take the guess at what this date was and so on. And then you can treat each shell as a time series and plot up the temperature values, as though you had a real little thermometer hanging out in, you know, in this region of the Santa Cruz coast, hundreds and thousands of years ago. And what is fun to see about this is at least for the, for the subplots B through E. There's something that resembles actually a seasonal sea surface temperature curve, it's sort of got the right shape. There's something that we're not just, you know that there isn't something going terribly wrong in our analysis somewhat as an independent check on this. It's not entirely independent but both part of what makes it somewhat independent is the fact that we actually plotted this in terms of the absolute temperature values, rather than those relative ones, and the curve still seem to hang together somewhat. We're hovering around some reasonable temperature values for the most part. One thing that I will note is these five dots up here represent one shell, which has some very balmy temperatures for the Santa Cruz coast. That's in Celsius over here. So, pretty pretty toasty waters there getting up close to 20 degrees Celsius. Those are are off by some factor, but we actually kind of know what factor they're off by because there's this gap here. And despite the fact that the absolute temperature values the absolute isotope values, don't match up with what we expect. We picked up on the shape and our analysis assigned it to a plausible time of the year, similar to these strands down here which are, you know, plausible absolute temperature values. One thing that killing Lee and Berger did in their 1979 piece was note that carbon isotopes can also be used as a potential proxy for upwelling intensity. This is a complex relationship that I don't have time to go into, but I made the plot and it's kind of fun to take a look at and speculate about that this is something we could potentially try to take trace also with these the same data set. And there's a, you know, it's just sort of an example of how much not only information on, you know, human interaction with the environment but also on other environmental variables that may impact, you know, these seascapes over the long term. How much of that sort of information is sort of ensconced in these little shells. So, fun stuff. All right, so now we get to some, some caveats, which is, this is a mini caveat. You notice there are missing shells here. They didn't find good matches in our historical temperature data set. And that is unfortunately the nature of a limited, a limited data set even though it's 100 years long. And the temperature conditions that you go through, it's not necessarily a, you know, a balanced representation of different annual and multi-decadal cycles. El Nino variation is not necessarily the full range of it is not necessarily included in that century and we know there's also changes over the Holocene and sea surface temperature so some of the shells didn't find good matches. At least within the goalposts I set up, which there is a better way of doing this full disclosure. I was plotting this stuff up this week. So I didn't have time to work out all the kinks yet but so that's also why all the, all the graphs have labels in like lower case and saying things like day. Day 300 of the year is about November 1. So, you know, just to give you some bearing there. Anyway, so what do you do about this? Well, you try to come up with a true functional relationship between the variables that you're interested in. Again, not something I'll go into in great depth here. But I took a hack at it. Oh, this is the real way to do it. Yeah, this is so foreshadowing of collaborations. My, I'll show you my version first. This is this is my sort of hacky way of trying to get estimates for these shells that we're missing. Two, four, five, seven, 10 and 23. You can get an estimate. That's great. It does affect somewhat the overall distribution, but not usually, but the other thing that's going on here is that I'm essentially still treating each of the growth band samples as like an independent observation, which they're not because those, the shape is really what we're doing and those shapes hang together. So this is just a sort of tag at like how you might go about this. The real way to do it is, can I go back? There we go. Is to do some fun Bayesian statistics, which is a favorite theme, but I'll leave for another time in any detail. Okay, we've got one growth band estimate here. Z is for the isotope value and W gives us a sense of over the course of the year. And the curve here is a representation of isotopic variation throughout the year, according to this temperature data set that we have and there's noise in that curve. So if you take a lot aligned through it at your temperature value or your isotope value, then you get regions of sort of increased likelihood. And if you take not only one, but several, for example, five growth band measurements, then you can start to build up a sharper estimate of when the shell is harvested. So this is backwards here, but besides point. So, moving on. plausibility. This is plausibility analysis, do our conclusions make sense remembering that this is a very small sample size only 40 shells, and those shells are split across three different sites and five different sort of chunks of radio carbon time. So the things that we want to look to here are other archaeological data sets and archaeological models that have been developed for this region and its time period the central coast over the over the mid to late Holocene. We want to look at historical accounts both oral and written. We know already about the way that harvesting was done on this part of the coast that's really important information, and especially as we get into building a proper Bayesian model this is something that we can incorporate as prior prior information. That's a technical term in Bayesian stuff but basically means what do you already know about the phenomenon you're interested in. And now you're doing an experiment and adding your your data analysis to that. And the thing that we'll look at today as well as the robustness of our conclusions to changing our assumptions, which is another nice thing that automation lets us do. I'll just briefly say, since I am, I am not truly a Californianist the way I'm a Californian. But with no, no real expertise about shell fishing alas, they taste good. And I'm hoping that this this mathematical analysis will will send me on the way to really securing a reliable source of of muscles in my life. So that's that's perhaps, you know, ulterior motive that I should reveal. Anyway, so I'll note very briefly that talking to the real Californianists in the room. The trends that we see taking place over the over the Holocene, based on that analysis that I showed you these these plots earlier. This plot, this plot scares people a little bit because it it scares but raises questions. Because it shows sort of a year round harvesting signature in the, in the middle of the scene and that sort of changes in the late halls and that's not really what we expect necessarily to see happen for various reasons. I'll, you know, and maybe tempered by various things but I'll let the experts talk about that. But I think what's important to mention here is that we know we're making some pretty big assumptions about growth rate and about the proper sort of reference data set to compare things with. And so those are things that I can vary in my nicely automated analysis so that's what I'll do now. We changed the growth rate. I said one and a half millimeters per month. Well, here are some comparisons of the conclusions and how they change for each shell if you assume one millimeter a month or two millimeters a month which are both totally plausible growth rates with very much within the likely range of variation. And the changes are not huge. In fact, it shows up a match for the shell number two which is kind of nice. But, you know, there are some changes here that are that are relevant by you know a couple of months here and there that we need to account for and certainly changes in sort of the the sharpness of the of the conclusion how how uncertain we are about it. Now, of course, things shouldn't get more uncertain when you vary the parameters so in fact what you want to do and this is why again a Bayesian approach will be fun is you can build this into the model, you can say well we think the growth rate is centered at one and a half millimeters a month but there's a bell curve variation around it perhaps. And then you can have that show up explicitly in your posterior distributions as it's called sort of what you think you know already your data, what you learned in your experiment experiment and posterior information which is what you think you know, based on what you knew before and now the experiment that you've done. So, anyway you can get this stuff into the into the sort of final density distribution. And here's the effect of growth rates on the on the distributions by site. Again, not huge but but not not negligible. So this is something to think about. Now, another important thing to consider is our sea surface temperature data set. And we used Pacific Grove, which is, as I said at the southern end of Monterey Bay, Granite Canyon is another recording station that doesn't have quite as long a data set which is why we didn't use it to begin with. It's only 1975 onward. So, but but that station is a little bit further south along the coast. So further from our archaeological sites but perhaps in a more similar marine environment. And so it might actually be that that Granite Canyon is a more appropriate comparison. But here we see this can affect the seasonality estimations by a great deal, almost in every case, it's a couple of months. And again, this is something we would want to build into the analysis like how likely do we think, you know, we can sort of break down what we think the SST curve sea surface temperature curve was the whole region that we think gathering may have been going on this shelfish harvesting activity for a given site right there's lots of things that we have to think about in terms of like, how are people interacting with this specific piece of archaeological evidence that in the past, and making those things explicit in our models for for this sort of analysis is is really helpful. So this is something that that we could potentially build in. But just to give a sense of the overall effect. Pacific Grove, this confusingly is different colors specific Grove now the, the Hopkins South Monterey Bay location that we use for our first analysis. That's in pink. The Granite Canyon data that is the reanalysis is in blue. And as we see there, this gives us a quite a different picture. And, and makes us think, well, it's actually probably pretty important to get a handle on this, you know, what the particular temperature environment was that that these muscles were being harvested in because it significantly affects our conclusions. And, and especially given that we have this other archaeological evidence in these archaeological models that say, you know, we should expect a different signature than the one we got from Pacific Grove, then that's really something that we need to bring in again sort of as prior information and that in that Bayesian terminology. So this is, you know, this is sort of the reason for taking this explicit uncertainty modeling approach and it allows you to integrate with other data sets. So that brings me to, you know, a discussion here of where we're going with this and in, in all cases that could be termed discussed in terms of collaborations. And, you know, we've got to do some experiments to understand muscle, California muscle growth rates on the Central Coast better. There's a pretty thin literature on that right now. Most of the studies have been done either the Northwest Coast, or in the Southern California, but south of the, the bite south of Santa Barbara, which are manifestly different, different ocean conditions. So we got to do some, do some experiments. And, and we hope to really connect those experiments with collaborators, both in the, the Amamutza and tribal band working specifically so that the experiments really like tie into ongoing stewardship work that they're, they're developing around, around seascape stewardship. And again, we'll have Alec and Mike on to talk a bit about that. And, and also hopefully to, to work on, work on these questions in association with the Hopkins Marine Station, and the good folks from, from Stanford who staff it, and, and know a lot about these, these organisms in these ecosystems and, and also have sort of a historical ecological interest in this stuff. But have to my knowledge at least not had the chance to work with archaeological data sets. And not yet had the chance to work with the knowledge that the Amamutza have about these, about these practices and, and, and how they tie in with sort of these social and ecological questions that go together. And, yeah, traditional resource management methods, and that sort of thing. And another element of this further approach is modeling. I mentioned my friend Gabriel Lewis who's a statistician over at the University of Massachusetts Amherst, and a very fine Bayesian has been working on on this problem, trying to develop that more explicit mathematical functional model that I was sketching briefly. And there are a lot of, you know, all of the things that we talked about as important assumptions that we need to, you know, sort of balance here, those things all need to go into the model. And that is the wonderful potential of Bayesian statistics is to really allow us to, to, to make, to make clear to the world to admit our full uncertainty without just becoming lost in limbo of not being able to conclude anything so hoping to, I hope that Nicholas Shackleton will, will not come down to, to put this Mac on my, my analysis here. And then thinking is always good. This is particularly thinking with archaeologists about the nature of the sites and thinking with our tribal partners about the nature and function of these sites and places and landscape. And how that relates to our quantitative analysis here, we want to make sure that we are sampling not only the shells in a sensible way but also the sites and the regions and thinking about the sort of statistical properties of the, the analysis that we're trying to do and the data itself and its properties and, and what we're trying to say about it. So, so these are some, some important questions that now sort of having a proof of concept I hope of this, of this method that that we can get into and try to build upon it. So, finally, I brought a picture from our very own micron here. And really the goal here is to, you know, to use this, this knowledge from, from the past, and, and for the moments and travel ban from ancestors to, to nuance and inform the way that, that, that all of us folks who live along the coast here interact with, with these ecosystems and which, with these organisms and, and with ourselves in relationship to, to, to our, our landscape and our environments, and having these sort of detailed archives in which these important environmental variables are recorded, and then more sort of more elaborately having these detailed archaeological archives that were, you know, constructed very intentionally by the indigenous folks, native Californians living along these, along these coastlines, as, you know, very clear testaments to a particular way of interacting with these, these coastal ecosystems and resources, and managing them and reacting to changes in, you know, in sort of external perhaps ocean climatic variables, and also conditioning those changes, buffering them and all sorts of things like that. So, this is what we're sort of all interested in, and getting at here. And, and that's very much a collaborative project so in light of that I will now bring on my, my special guests who are very much at the core of this project that I am somewhat of an interloper and so I just want to thank everybody at the, at the here at Berkeley, and, and of course can like but for, for occasioning this sort of broader research, research theme. And, and also, of course, the, the collaborators at the Amemites and Tribal brand and, and land trust for, for, you know, being willing to entrust these, these archives of, of ancestral practices to, to, to us to work with. And, and yeah, I hope for a lot of a lot more fruitful collaboration in the future. Thank you all so much. Thank you, Jordan. That was an excellent presentation and really impressive, just how you've, you know, taken on the problem of isotopes and archaeological assemblages from a nuanced perspective and, you know, contributing to a broader body of work about seasonal site use and slow patterns and traditional resource harvesting practices, very novel approaches and, as you mentioned, very useful for the ongoing efforts of the tribal band. There's an active project right now that I'm working on as a consultant with the tribe that's integrating archaeological information such as this into efforts of restoring coastal stewardship from for people who were largely removed from their traditional coastlines and shorelines. So, in a lot of cases, traditional resource management practice knowledge and traditional ecological knowledge like timing of harvest locations, extent size preference and just harvesting profiles in general are largely lost in the canon of ethnobiological information that the tribe holds. So this sort of analysis can provide these key links for cultural resource practitioners who are trying to restore traditional food ways and just relationships to the coast that span and but also have a very kind of seasonally prescribed means of interacting so I think really great stuff and exciting and very useful. I'm excited to see, you know, developing projects. More sampling with a larger more robust data set to just see how more however fine we can get these interpretations might pass it over to Alec Apodaka on that because he's a, you know, he did a lot of sampling for these, these individual shells and see what ways to refine this method. Thanks, Mike. And yeah, I'll just follow up on that I think that, you know, Jordan's call for some, some more experimental studies is really warranted and with you know the beauty about that is that if with the Alma Mutes and Land Trust and Alma Mutes and Tribal Band, going now and having controlled experiments of intertidal plot to open the door for just a lot more other things. And, you know, right now the Alma Mutes and Tribal Band is really interested in monitoring the effects of climate change, ocean acidification and pollutants that are affecting a lot of the organisms on the coast. And I feel that, you know, just starting with monitoring the growth rates of muscle shells over the course of let's say one year or two years or three years. This is going to provide a lot of information into how we interpret the isotopic results. Once we apply it to archaeology. So, you know, I think that, you know, it's, it's fruitful. We can probably sit back and feel that isotopes are, are fraught with a lot of uncertainty. And we can, you know, feel that we need to tread lightly and we don't want to make interpretations but I think the future is very bright. And we're headed in that direction. I know we don't have a whole lot of time left but folks have questions. Be happy to answer them, either for me or for Alec or Mike. Kent Lightfoot, our very own, has a question in the chat, which I will ask. Can you make any interpretations about muscle harvesting for the archaeological site at this time? And what is explicitly your next step besides eating more muscles? I'm glad Kent caught the primary aim of this talk. So, I mean, what I can speak to is what the analysis says. So, you know, I'll go back to one of the plots here. I'll bother to do this interview just so you can see it a little more clearly. So, this is sort of the basic conclusion by site and also by time period sort of chunked up into where the radiocarbon dates associated with this context cluster. So, what this says to me is that, you know, we're seeing during most of the time that through most of the context and SCR7 that we looked at, distinct likelihood of harvesting throughout the year. At the oldest time period, maybe a more pronounced sort of winter spring peak. And then looking at the late Holocene sites, perhaps this is just a single well, perhaps two shells, so I wouldn't give it a whole lot of credence. But this plot B shows SCR14, which is an inland site, you know, a couple of kilometers inland from the other, from the coast. It shows a winter, harvesting emphasis. And then the coastal site from the same time period, but further north on the coast, some tens of kilometers, not too much more than that, has a spring emphasis. And so, mostly I think what this shows us is that we're making some strong assumptions, because of course if we go down to the graph here where we use a different temperature data set that changes the conclusions pretty significantly. SCR7 in particular, no longer is throughout the year as much, much lower harvesting densities in winter, and much higher sort of concentrated in spring and summer. So I think the main conclusion to be drawn here is that, you know, we probably need larger samples, that's no surprise, we only have 40 shells and only like 15 from our biggest sites and that's put up between these different time periods that are separated by, you know, almost 1000 years each. So we wouldn't expect to like get the whole signature of the site in that. So we know that assuming our temperature data sets is the data set that we choose is important and has some pretty big leverage on the conclusion. So that's that's where I would leave it. Christine also has a question Christine Hastorf. You partially answered it. She asks, it, can you say if they were harvested year round or seasonally, and then specifically asks what about the red tide. Yes. So, the red tide is harmful algal blooms going on. The red tide is a natural phenomenon or that's what it's meant to distinguish from, from what I've heard, the termed habs, HABs, harmful algal blooms which are, which are not necessarily a sort of an on a natural cycle. And that we absolutely think is, is a, is a likely reason that people would be avoiding shellfish at certain times during the year as they become toxic to eat. And, and folks would have certainly been aware of these, these red tides. The thing to consider is that they are significantly associated with not only sort of solar for things in terms of kicking up photosynthesis but also upwelling and that's something that can vary through time and so that's, that's part of the reason that I showed the, the carbon isotopes because that's the way of starting to get at that. And there are also other geochemical signatures sclera chemical signatures in these shells that that might offer potential avenues towards that so that's a very good question. And then, yeah, I remembered to can't ask for, for next steps and I would say that both, I'd say two next steps that come immediately to mind. One is to apply these analyses which are all nice and automated to existing isotope data that's already in the there's thousands of isotope determinations, thousands of shells from various sites that have been recorded in a way that is absolutely digestible by this program it just will take somebody to tabulate them and clean the data. But that would be a very quick way to bump up sample size, not necessarily specific at these sites but but over the region as a whole. And could give us a sense of what those broader patterns are, and then as a twin to that is the stuff that Alex talking about of, you know, working with the tribe and developing these, these experimental plots that can also work as, you know, a sort of learning teaching learning testing ground for these traditional resource management techniques and and acknowledges that that folks have and and want, you know, a venue for. So, so yeah so I think those are kind of the two, two places to go to go next. Great. Well, those are the only questions I'm seeing right now we're already a little over time. But yeah, thank you all so much for a fun talk about yeah place that I think many of us are personally acquainted with makes a little bit more relatable. Because we can't necessarily travel there at the moment.