Extinctions and niches

 Watched a video on mass extinctions, discussing whether the idea of the "top" five is the right way to see it.  Leaving that aside, the showed a chart of the known variety of life, giving the impressions that the extinctions were relative small dips (ten percent or so) in an overall increasing level of variety.  The program didn't address that, but I feel the diagram misleading.  While number of variety may not have dropped much, number of organisms dropped sharply and the variety is not really increasing so much as, as we get closer to the present, we have identified more variety.

Your skepticism is well founded. There are a few different quantities that often get mixed together in popular presentations of extinction history:

    Diversity (number of species, genera, or families)

    Abundance (number of individual organisms)

    Biomass (total living mass)

    Completeness of the fossil record

A graph showing "variety of life" is usually plotting diversity, not abundance. During a mass extinction, abundance can collapse catastrophically while diversity appears to decline less dramatically.

For example, imagine a world with 1,000 species and a trillion individual organisms. If 100 species disappear but 90% of all individuals die, the diversity graph shows only a 10% drop while the ecological reality is devastating.

The end-Permian extinction, about 252 million years ago, is the classic case. Estimates suggest that roughly 80–90% of marine species disappeared and ecosystems were shattered. Even after the extinction pulse ended, Earth remained biologically impoverished for millions of years. A simple diversity curve often fails to convey just how empty the oceans became.

There is another issue with those charts. The apparent long-term increase in diversity is partly real and partly an artifact.

Real increase:
Life genuinely became more diverse over geological time. The Cambrian seas were impressive, but they did not contain flowering plants, birds, mammals, grasslands, coral reef fish communities, and countless other later innovations. Evolution has created new ecological opportunities and generally increased biodiversity over billions of years.

Artifact:
The closer we get to the present, the better the fossil record tends to be. Younger rocks are more abundant, less altered, and easier to study. Tiny organisms, delicate organisms, and localized species are much more likely to be recognized in recent periods than in ancient ones. Paleontologists have spent decades developing statistical methods to correct for this "pull of the recent."

So when you see a smooth upward trend, some of that rise is probably due to improved preservation and sampling rather than an actual explosion of diversity.

There's also a visual problem. Many popular diagrams use a very large vertical scale. If diversity rises from, say, 1,000 to 10,000 groups over hundreds of millions of years, then even a catastrophe that removes half the world's species can look like a modest notch in the curve. It's a bit like plotting the Great Depression on a graph of world GDP from the year 1000 to today—the depression is enormous in human terms but may appear as a small dent in a long upward trend.

In fact, one reason paleontologists call the "Big Five" extinctions exceptional is that they stand out despite the coarse nature of the fossil record. If they are visible as dips on a graph that averages millions of years at a time, the actual biological crises were likely even more dramatic than the graph suggests.

So I would say your criticism is essentially that the graph is conflating "how many kinds of organisms are known" with "how healthy and populated the biosphere was." Those are very different things. A diversity curve is useful, but it can easily understate the severity of a mass extinction if viewers interpret it as showing the amount of life on Earth rather than the number of distinct lineages.

That's actually quite close to modern thinking, with a few important refinements.

The old picture was often presented as: life invades an empty world, rapidly fills all available niches, and then reaches something like an equilibrium. In that view, diversity should level off fairly quickly after the Cambrian.

What paleontologists have increasingly found is that niches themselves evolve.

The Cambrian seas lacked forests, grasslands, coral reefs in their modern form, flowering plants, birds, whales, bats, and countless other ecological opportunities. When a major innovation appears, it can create entirely new ways of making a living.

Consider just a few examples:

    Land plants create forests.

    Forests create niches for insects, fungi, birds, and tree-dwelling animals.

    Flowering plants create niches for pollinators.

    Pollinators drive diversification of flowers.

    Grasses create grasslands.

    Grasslands create niches for grazing mammals.

    Grazers create niches for specialized predators and scavengers.

The number of ecological opportunities is not fixed. Life is constantly modifying the environment and creating new opportunities for other life.

One ecologist described this as organisms being not merely occupants of niches but "ecosystem engineers." Beavers are a familiar modern example, but on geological timescales entire groups of organisms reshape the planet.

That said, there is still debate about whether diversity can increase indefinitely. Some researchers have argued that Earth's ecosystems have carrying capacities and that diversity should eventually approach a ceiling. Others see little evidence of a hard upper limit over the Phanerozoic.

What makes the question difficult is that every time diversity seems to approach a plateau, something changes the game: land is colonized, forests arise, flowering plants evolve, mammals radiate, and so on.

Your observation about niche division is also important. Once broad niches are occupied, much subsequent evolution is indeed subdivision. A general seed-eating bird species may split into several species specializing on different seed sizes, habitats, or feeding times. The niche space becomes more finely partitioned.

An interesting question is whether the Cambrian world really had "all available niches" filled. Looking backward from today, the answer seems to be no. Many of the niches we take for granted literally did not exist yet because the organisms that create them had not evolved.

So the modern view is something like:

    Empty niches are rapidly filled.

    Competition encourages specialization and niche splitting.

    Evolutionary innovations create entirely new niches.

    The total amount of niche space therefore tends to expand over time, though not necessarily smoothly.

That produces exactly the sort of long-term upward trend in diversity that those graphs often show.

The caveat is that the graph's slope may exaggerate the reality because of preservation and sampling biases. The general upward direction is probably real; the exact steepness of the climb is much less certain.

There's also a delicious irony here. Mass extinctions, terrible as they are, often create huge numbers of empty niches. Mammals remained relatively minor players for over 100 million years while dinosaurs dominated most large-animal roles. After the end-Cretaceous extinction, those vacant niches helped fuel the extraordinary diversification of mammals. In a strange way, some of the upward steps in the diversity curve owe their existence to the very crashes that appear as dips on the graph.

I'm inclined to make the point that a niche is a classificatory device we have invented to put order into what is reality a bunch of constantly moving ranges.

That's a very defensible position, and many ecologists would sympathize with it, although they might phrase it a bit differently.

A niche is not a physical thing out there in nature like a tree or a river. It's a concept we use to describe how an organism interacts with its environment: what it eats, where it lives, what eats it, when it reproduces, what temperatures it tolerates, and so on.

The classic textbook picture tends to make niches look like neatly bounded boxes:

    Species A occupies niche A.

    Species B occupies niche B.

    Species C occupies niche C.

Reality is much messier. Most species occupy a range of conditions, and those ranges shift over time. The boundaries are fuzzy. Two species may overlap extensively, compete in some circumstances, cooperate in others, and change behavior when conditions change.

Modern ecology often thinks in terms of a multidimensional "niche space." The niche of a species is not a box but a cloud in a space with many dimensions: temperature, humidity, prey size, nesting location, activity time, and dozens of other variables. The cloud expands, contracts, and moves.

Even then, the niche is still a model—a useful abstraction.

Your wording about "constantly moving ranges" reminds me more of population thinking than typological thinking. Instead of seeing nature as made up of fixed categories, you see distributions that shift and overlap. Darwin himself pushed biology in that direction. Species are populations, not immutable types; perhaps niches are better thought of as statistical tendencies rather than fixed ecological job descriptions.

One illustration is the coyote. A century ago, a North American ecologist might have assigned coyotes a fairly specific niche. Today coyotes occupy deserts, forests, suburbs, and city centers, eat everything from rodents to fruit to garbage, and show remarkable behavioral flexibility. Did the niche change, or did our description of the niche change? Arguably both.

From that perspective, saying that "all niches are filled" can be misleading. It suggests a finite set of predefined slots waiting to be occupied, as if nature were a giant cabinet with labeled drawers. In reality, the drawers are being redesigned while the organisms are evolving.

That doesn't mean the niche concept is useless. It remains one of ecology's most powerful tools. But it is probably best viewed as a map rather than the territory—a way of organizing an extraordinarily fluid reality rather than a set of objectively existing compartments in nature.

That may be why you reacted to the extinction graph. Graphs, niches, species counts, diversity indices—they are all abstractions layered on top of a messy biosphere. Useful abstractions, often indispensable ones, but abstractions nonetheless. The trick is remembering what has been left out when the picture starts looking too tidy.

Specialists are in narrow niches, generalists in broader ones or even more than one.

Exactly. That's one of the reasons many ecologists think of niches as regions rather than points.

A specialist occupies a relatively narrow range of conditions and resources. A giant panda is the textbook example: it can survive in only a limited set of habitats and relies heavily on bamboo. If the bamboo disappears, the panda is in trouble.

A generalist occupies a much broader range. Rats, crows, cockroaches, and humans are famous examples. They can exploit many foods, many habitats, and many environmental conditions.

One way to picture it is not as a niche being a box, but as a territory on a map. A specialist occupies a small territory; a generalist occupies a large one. Some territories overlap. Some barely touch. Some shift with changing conditions.

It gets even more interesting because the same species can be specialist and generalist depending on which dimension you're examining.

A bird might:

    Eat many kinds of food (generalist diet).

    Nest only in one type of tree (specialist habitat).

    Tolerate a wide temperature range (generalist climate).

    Breed only during a narrow seasonal window (specialist reproduction).

So asking "Is this species a specialist or a generalist?" can oversimplify matters.

Your earlier comment about niches being classificatory devices fits well here. The distinction between specialist and generalist is often a statement about the size and shape of the ecological range we've chosen to measure. Nature doesn't contain little signs saying "specialist" and "generalist"; those are labels we apply to patterns we observe.

There's also an extinction angle. Specialists often do very well in stable environments because they become extraordinarily efficient at exploiting a particular way of life. Generalists tend to be less efficient in any one setting but more resilient when conditions change.

That is one reason mass extinctions can be so disruptive. When the environment changes rapidly, many finely tuned specialists lose the conditions they depend on. Generalists and adaptable species often survive disproportionately well. Afterward, the survivors diversify and new specialists evolve.

In a sense, evolution repeatedly runs a cycle:

    Generalists colonize opportunities.

    Some populations become specialists.

    Environmental disruption wipes out many specialists.

    The surviving generalists expand again.

It's not a universal rule, but it's a pattern that appears often enough that paleontologists and ecologists both recognize it. The history of life can sometimes look less like a steady filling of niches and more like a continual reshaping of a very fluid landscape.