The question itself is predicated upon an assumption that what humans do must be superior for some reason. Since many, maybe even most, macroscopic lifeforms die after a single reproduction event, a better question might be why us weirdo humans bother to live on after reproducing. It’s not like what we do is the norm or the most successful reproductive strategy. Heck, it isn’t even universally true of mammals.
Coming at it from another angle, the octopus has two choices. It can devote all its resources to the eggs, or it can devote only some of them and reserve some for another round of reproduction. Which tactic is superior depends upon the environment. Let’s look at some hypothetical scenarios.
If the octopus reserves energy for survival, it will need to lay down fat reserves. In order to do that it will have to do one of two things: either it will have to mate later, because the energy that could have gone into maturing earlier has instead gone into fat reserves. Or if it does want to mate as soon, it will need to hunt more.
Let’s assume it takes an extra 3 months to lay down the necessary fat reserves with normal hunting activity. And let’s assume that an octopus has a 5% chance each month of being killed during normal activity. What that means is that the initial delay results in a ~15% chance of producing no offspring at all. Additionally, if we assume that an octopus can breed just once a year, then between the time when it *could *have bred and the *next *opportunity to breed there will be a 51% chance of being eaten. As a result, the octopus that reserves energy for survival has a 15% chance of never producing at all, and only 49% chance of reproducing more than another octopus that doesn’t reserve energy.
IOW, over half the time a perrenial octopus will produce fewer offspring than an annual octopus that dies after breeding. If you have two competing populations of octopus, then the perrennial form will extinct within just a few hundred generations.
And that assumes a constant rate of death. If the chances of death following breeding, during “winter”, increases to 7%/month, then the strain that survives breeding will produce just 1/3 as many offspring as the one that dies.
If the octopus wants to hunt more to mature as fast, then it will have to contend with the fact that a hunting octopus has much more chance of being eaten or injured by prey than an octopus resting in a burrow.
So, let’s assume our octopus instead opts to hunt a total of 10 days more before its initial breeding to produce enough fat to survive reproduction and still mature at the same rate. And let’s assume that an octopus just a 1% greater chance of dying each for day that it spends hunting. Now our perennial octopus has a 10% greater chance of dying before it breeds even once. Moreover, after breeding it will need to feed enough to both make up for the weight lost during breeding, but also to lay down as much fat in 12 months as it laid down in the past 5 years. So we can assume it will need to hunt for an additional 20 days to be able to breed again next year. So that means a 17% chance of dying before it has a chance of breeding a second time. But it still has to survive the *normal *rigours of an octopus year, with that 45% chance of dying. That means that a perennial octopus has a 63% chance of producing less young than an annual octopus.
This isn’t a situation that requires much explanation. Nor is it a situation with no apparent survival benefit. The benefits of such a life history is potentially huge. All that is required is that the chance of dying before the opportunity to breed again be greater than 50%. Wherever that is true, the most sensible strategy is going to be to devote all your resources to a single clutch of offspring.
That’s why this lifestyle is so common amongst invertebrates and plants in highly seasonal environments. Insects such as mayflies or bumblebees, plants of desert of cold temperate regions, mammals of the deserts. They breed just once because the chances of surviving to breed again next year are low whether reserve resources to do so or not. Far better to maximise the number of offspring that will be produced during a known successful mating than to gamble on the short odds of ever being able to mate again.