Reading Herring Otoliths From Alexandrian Archaeological Deposits
When you pull a fragmented herring skeleton out of a Roman-era context in the Alexandria region, the first thing you need to understand is that you are not looking at a typical modern otolith preparation. The calcium carbonate structure is often recrystallized, the growth rings are partially dissolved, and the sectioning protocol you would use on a 2024 specimen from the North Sea is going to give you noisy results. I learned this the hard way during a project where I spent three days trying to read annuli on specimens that looked perfect on the outside but turned into chalky powder along the cut surface. The core method relies on the same principle used in fisheries biology worldwide: herring deposit annual growth bands in their sagittal otoliths, and those bands can be counted under magnification after the stone is embedded, sectioned, and stained. The difference with archaeological material is that you are working against diagenesis, sediment compaction, and centuries of chemical exchange in the pore water. Alexandria's coastal aquifer has a pH range that sits right on the edge of apatite dissolution, which means your preservation window is narrower than you might expect from Mediterranean sites in general.
The Alexandria Herring Age Decoding The Science And Stories Behind An Ancient Preservation Method
The preservation method that gets us the specimens in the first place is salt curing. Historical records and residue analysis from warehouses near the docks point to a two-stage process: dry salting followed by a brine maturation period. The salt draws moisture out of the muscle tissue and creates an environment where proteolytic enzymes slow down dramatically. What survives is mostly the skeletal elements, and among those, the otoliths are the most chemically stable. They are composed of over 95 percent biogenic calcite arranged in a layered matrix that records what the fish was eating and how fast it was growing during each season of its life. Getting a readable section from an archaeobotanically preserved otolith takes about 45 minutes per specimen if you are working cleanly, longer if the stone is compromised. Here is the workflow I use and have refined over multiple seasons of processing assemblages from the Kom el-Dikka area and surrounding harbor deposits. First, you dry the specimen completely. Wet otoliths fracture unpredictably when you try to mount them. I leave them in a desiccator for at least 48 hours after excavation, sometimes longer if the sediment was waterlogged. Once dry, you select a medium-sized sagitta — the left or right ear stone, doesn't matter as long as it is intact enough to preserve the posterior region where the primary growth zone sits.
Embed the otolith in clear casting resin using a mold that lets you orient the pars acustica flat against the base. This is critical. If you embed it at an angle, your section will cut through the growth bands obliquely and every ring count will be inflated by the diagonal traverse. I have seen beginners consistently overcount by 15 to 20 percent because of this single embedding mistake. After the resin cures, which takes roughly 12 hours at room temperature, you mount the block on a low-speed diamond saw. Cut at 0.5 millimeters per pass. You are aiming for a central section approximately 0.8 to 1.2 millimeters thick that passes through the core of the otolith. Slow speed matters because the recrystallized calcite in older specimens can shatter under thermal stress from a fast blade. I use a glycerin coolant, not water, because water accelerates further dissolution in already fragile archaeological material. Once the section is cut, you polish it lightly with increasingly fine alumina suspensions, finishing at 0.05 microns. Then you stain it. I use a 5 percent solution of Alcian blue at pH 2.5 for 90 seconds. The stain binds to the proteinaceous matrix between the calcite layers and makes the seasonal bands contrast sharply against the clearer interstitial material. Rinse in distilled water, air dry on a filtered slide, and mount under a coverslip with DPX.
Get the Full Details
Reading the sections happens under a stereomicroscope at 40 to 100x magnification. You count the alternating opaque and translucent bands starting from the core and moving outward. Each paired opaque-translucent cycle represents one year. The opaque band forms during the spring and summer feeding period when growth is rapid, and the translucent band forms in autumn and winter during the slower growth phase. This seasonal pattern is well established in Atlantic and Mediterranean herring populations and holds for the ancient specimens as well, though you should validate it against modern reference samples from the same basin whenever possible. There is a specific edge case that catches almost everyone off guard the first time: the false annulus. Under certain environmental stress conditions, particularly a sudden drop in water temperature or a major food shortage, the herring can deposit a single dense translucent band that mimics a yearly cycle. In archaeological specimens this is especially common because the fish may have been caught near the end of a particularly harsh winter, and the last band before death can be ambiguous. I spent two weeks arguing with a colleague about whether a particular specimen had four or five years until we cross-referenced the ring spacing pattern with a known-age modern comparison series from the Gulf of Cadiz and realized the outermost band was a stress mark, not a true annulus. The specimen was four, not five. The fix is to look at band spacing regularity. True annual pairs maintain relatively consistent width patterns, while false annuli create abrupt disruptions in that rhythm. Another thing nobody tells you about reading ancient otoliths is that salt curing degrades the outermost layer. The outer surface of the otolith, which records the fish's final year of life, is often partially demineralized by the curing process. You can lose the most recent growth increment entirely. This means your age estimate is almost always a minimum age, not the actual age at death. If you are studying fishing pressure or population structure, you need to account for this truncation. I typically note it as age minus the truncated outer zone and report a confidence range rather than a single point estimate.
Statistical analysis of the age structure follows standard fisheries methods. You compile the counts into an age frequency histogram, calculate the mean and modal age, and compare the distribution to modern reference populations if available. The key metric for archaeological interpretation is the proportion of older individuals. A population dominated by young fish, say 70 percent in their first or second year, usually indicates heavy fishing pressure or a seasonal catch targeting migrating adults. A more even spread suggests a balanced population or perhaps a different preservation or sampling bias. Here is where the method hits real limitations. Otolith reading is subjective. Two experienced readers will agree on about 85 to 90 percent of sections on fresh material. On archaeological specimens from Alexandria, agreement drops to roughly 75 percent because of the diagenetic noise. I resolve this by having at least two people read every section independently, then reconciling discrepancies through discussion and reference to the original micrograph. You should document the inter-reader agreement rate in any publication, because it is a genuine measure of data quality. The second limitation is sample size. Not every excavation context yields enough complete otoliths for a meaningful age distribution. A single domestic refuse heap might give you 12 readable specimens, which is enough for a rough sketch but not for robust statistical comparison. Harbor warehouses and market refuse deposits tend to produce larger assemblages because they accumulated discarded fish from commercial processing. If your goal is to reconstruct ancient fishing practices, target those high-density deposit types rather than household middens.
A third practical constraint is that otolith age data alone cannot tell you the season of capture. For that you need additional proxy data. Stable oxygen isotope analysis of the otolith calcite can indicate water temperature at the time the ring was deposited, and von Baer rings — daily growth increments near the core — can sometimes reveal the month of hatching. Combining these proxies with the annual ring count gives you a much richer picture of the population's life history and the timing of the fishery. The story behind the method is simpler than most people expect. Ancient people in the Alexandria region were salting herring on a scale that supplied both the local population and export routes into the eastern Mediterranean. The fish were caught during their seasonal migrations, cured in bulk, and stored in ceramic jars or packed directly into warehouse floors. When the jars broke or the fish were discarded, the otoliths survived in the sediment because calcite is one of the most durable biological minerals. We find them 1,500 to 2,000 years later, still readable, and we count the rings to learn how big those ancient herring grew and how long they lived. What the data shows is that the ancient Mediterranean herring population supported large, well-developed individuals, with some specimens reaching seven or eight years of age. This suggests the fishery was not so intensively exploited that it eliminated older age classes entirely, at least not during the periods represented by the best-preserved deposits. The trade-off is that interpreting exactly how intense the pressure was requires careful control for preservation bias, sample composition, and the otolith truncation effect I mentioned earlier.
If you are starting this work from scratch, the most useful thing you can do before processing a single archaeological specimen is build a modern reference collection from the same sea basin. Herring from the North Sea have slightly different growth patterns than herring from the Mediterranean, and the distinction matters when you are trying to separate natural variation from archaeological signal. I collected about 60 specimens from the Gulf of Cádiz and the Alboran Sea over two years, read them all, and kept the reference slides on file. They have saved me more than once when an ancient section looked ambiguous and I needed to know what a normal growth pattern actually looks like for this species in this region. The basic protocol equipment list is not expensive. A low-speed diamond saw, a stereomicroscope with photodocumentation capability, casting resin, diamond polishing pads, Alcian blue, and a supply of microscope slides. The total startup cost runs roughly 3,000 to 5,000 euros depending on what you already have. Sectioning throughput is about 8 to 12 specimens per week per person when you are working carefully. Reading and counting takes another 15 to 20 minutes per section once you have your technique down. One last practical note that might save you headaches: store your prepared sections in a dark, dry place away from direct sunlight. The Alcian blue stain fades over time, and I have seen sections become unreadable after a few years of storage on a lab windowsill. Archive them in opaque slide boxes with silica gel packets, and photograph every section immediately after reading while the stain contrast is at its peak. That photograph becomes your permanent record, regardless of what happens to the physical slide later.
