There's a photograph in the catalogue from a 2019 expedition off the Azores. It shows what appears to be a small octopod—probably Grimpoteuthis genus, one of the so-called "Dumbo octopuses"—hovering above a rock outcrop at 2,847 meters depth.
The image is technically perfect. Good lighting, sharp focus, excellent color balance. The organism is clearly visible, its ear-like fins extended, its arms trailing below. It's a beautiful capture.
But the observation notes say almost nothing useful.
"Octopod observed at survey station AZ-19-07, depth 2847m, substrate: rocky. Single individual. No behavioral notes recorded."
That's it. That's the entire context we have for this organism. We know where it was and how deep. We know it was alone. We know there were rocks.
We don't know what it was doing before the ROV arrived. We don't know how it reacted to the lights and machinery. We don't know if this area had other octopods nearby or if this was an isolated individual. We don't know what the surrounding environment looked like beyond "rocky." We don't know the water temperature, the current speed, the oxygen levels, or any of a dozen other factors that might explain why this animal was in this particular spot.
The photograph exists. The organism was documented. But most of the actual observation is missing.
This happens more often than you'd think. A researcher is focused on their primary survey target—maybe they're mapping geological features or counting fish populations—and they encounter something unexpected. They capture an image because it's interesting or unusual. But they don't record detailed observational notes because that's not what they're there for.
So we end up with photographs that raise more questions than they answer.
The Azores octopod is one example among hundreds. There's a sea spider from the Mariana Trench with legs that seem disproportionately long, but we have no measurements and no notes about whether this was unusual for that species. There's a cusk eel photographed near a hydrothermal vent in the Indian Ocean, but the image doesn't show the vent itself and the notes don't mention how close the animal was to the vent field or whether it was actively feeding.
There's a stunning photograph of what might be a new species of stalked crinoid from the South Pacific, but the ROV was moving when the image was taken and there's no second photo, no close-up, no tissue sample, nothing to work with except that single blurred but tantalizing image.
Incomplete observations.
The hard part about cataloguing them is deciding what to do with the gaps. Do you note that information is missing? Do you make reasonable inferences based on what similar species typically do? Do you flag the entry as incomplete and hope that someday someone will survey the same area again with better documentation?
The conservative approach is to record only what's explicitly documented and make no assumptions. The octopod was observed at 2,847 meters depth on a rocky substrate. Period. Don't speculate about behavior or environmental conditions or anything else not in the notes.
This is scientifically rigorous. But it also means the catalogue entry is barely more informative than the raw data. A researcher looking for information about Grimpoteuthis behavior or habitat preferences won't find anything useful here.
The alternative is to add context based on what we know about related species and typical deep-sea ecosystems at that depth and location. Note that octopods at this depth typically hunt small crustaceans and polychaetes. Mention that the rocky substrate probably indicates a hard-bottom environment rather than abyssal plain. Infer that water temperature was likely between 3-5°C based on regional oceanographic data.
This makes the catalogue entry more useful. But it also mixes documented observations with inferred context. And inference, no matter how reasonable, introduces the possibility of error.
I usually split the difference. Record exactly what was documented in the primary observation section. Then add a contextual notes section that makes clear distinctions between verified data and reasonable inference based on related species and environmental knowledge.
"Observed at rocky substrate, depth 2847m. Regional oceanographic data indicates typical water temperature 3-4°C, oxygen concentration ~5.5 mL/L. Congeners typically hunt benthic prey (small crustaceans, polychaetes). Dumbo octopuses are generally solitary; single individual observation consistent with typical behavior. Note: environmental parameters inferred from regional data, not measured at observation site."
It's honest about what we know versus what we're inferring. It provides useful context without pretending the observation was more complete than it actually was.
But it's also more work. And it requires judgment calls about what inferences are reasonable versus what's speculation.
The Azores octopod got the full treatment because someone specifically asked me about Grimpoteuthis habitat data and I went back through every observation we had, including the incomplete ones. Most entries don't get that level of attention.
Which means the catalogue is full of photographs with minimal context. Organisms observed but not fully documented. Moments captured but not explained.
It bothers me more than it probably should. Each of those photographs represents an encounter between human technology and a living organism in one of the most remote environments on Earth. Each one is a small miracle of engineering and exploration and timing.
And we have so little information about what we were actually seeing.
Part of the problem is practical. ROV time is expensive and survey schedules are tight. Researchers can't spend twenty minutes documenting every interesting organism they encounter when they have specific mission objectives to complete. Sometimes a quick photograph is all there's time for.
Part of it is human attention. When you're four hours into an eight-hour dive and you've already documented dozens of specimens and your eyes are tired from staring at screens, it's easy to see something unusual, snap a photo, and move on without recording detailed notes.
And part of it is that we often don't know what details will matter until later. The researcher who photographed that octopod probably didn't think anyone would care about the exact current velocity or substrate composition at that spot. They captured the image. That seemed sufficient.
But six years later, when I'm trying to build a comprehensive understanding of Grimpoteuthis habitat preferences across different ocean basins, those missing details matter a lot.
The cruel thing is that the information existed at the moment of observation. The ROV had sensors measuring temperature and depth and salinity. The pilot could have panned the camera to show the surrounding environment. The researcher could have spent two more minutes recording notes about what else was visible in the area.
It would have been easy to capture. But nobody did. And now that context is gone forever.
You can't go back and re-observe something that happened in 2019 at a specific spot on the seafloor. That moment is past. What wasn't recorded is lost.
This is one of the fundamental tensions in observational science. You have to decide in real-time what's worth documenting, but you won't know what will turn out to be important until much later when you're analyzing the data and trying to answer questions nobody had thought to ask yet.
So you either document everything exhaustively, which is usually impossible given time and resource constraints, or you make judgment calls about what seems important and accept that you'll inevitably miss some details that later turn out to matter.
Most expeditions lean toward the second approach. Document the primary mission objectives thoroughly. Capture opportunistic observations of other interesting things. Move on.
Which means the catalogue accumulates incomplete observations. Fragments of encounters. Pieces of a story we can't fully reconstruct.
I've been thinking about whether there's value in flagging the incomplete observations explicitly. Not just noting that some details are missing, but actively marking entries that have significant gaps in documentation.
"Observation incomplete: no behavioral notes, environmental parameters inferred, no tissue sample collected."
It would be honest. It would make clear to researchers that this entry shouldn't be relied on too heavily. It would acknowledge the limitations of what we know.
It would also make obvious just how many catalogue entries are incomplete. Which might not be great for the catalogue's credibility.
But maybe that's okay. Maybe being honest about the gaps is better than pretending we have more complete knowledge than we actually do.
The alternative is that researchers using the catalogue might not realize they're working with fragmentary data. They might build hypotheses or draw conclusions based on observations that are missing crucial context. That seems worse than acknowledging incompleteness upfront.
So maybe the incomplete observations need explicit flags. "Limited documentation available" or something similar. A clear signal that there's less here than you might hope.
It won't fix the missing context. But it will at least make clear that the context is missing.
The Azores octopod observation is still incomplete even with my added contextual notes. I can't retroactively document what the researcher didn't record. I can only work with what exists and acknowledge what doesn't.
There are 1,847 observations in the current catalogue. I've gone through maybe 300 of them in enough detail to assess documentation completeness. Of those, approximately 70% have some kind of significant gap—missing environmental data, minimal behavioral notes, unclear taxonomic identification, single photograph with no follow-up.
Seventy percent. Most of our observations are incomplete to some degree.
That's not a failure exactly. It's the reality of deep-sea research. You work with what you can capture in the moment, knowing it won't be perfect.
But it does mean that the catalogue is less a complete record of deep-sea life and more a collection of fragments. Pieces of encounters. Glimpses that raise as many questions as they answer.
The scientist in me wishes every observation was exhaustive. The pragmatist in me knows that will never happen. The cataloguer in me just tries to document what exists and be honest about what doesn't.
Missing context is a permanent condition. The best I can do is make the gaps visible so researchers know what they're working with.
A photograph of an octopod hovering above rocks. Depth recorded, location noted, behavior unknown. One moment captured, most of the story missing.
That's the job. Preserve what was documented. Acknowledge what wasn't. Try not to let the gaps feel like failures.
The deep ocean is vast and strange and mostly unobserved. Every photograph is valuable even when it's incomplete. Every documented moment adds something to our understanding even when crucial details are missing.
It just also means we're building knowledge on partial information. Fragments of observations. Incomplete encounters.
The octopod in the photograph might have been hunting or resting or moving between feeding areas. We'll never know. What we know is that it existed in that place at that moment and someone cared enough to capture the image.
Sometimes that has to be enough. Even when it isn't.
Document what exists. Note what's missing. Move to the next entry.
The catalogue grows. The gaps remain. The work continues.