Hacker News Reader

Comment by xg15 | Hacker News Reader

I mean, they at least claim that solving this problem constitutes the actual secret sauce of their project:

> "The niche really lies in our biotech, which is the support system for the seed once it's on the ground," says Walker.

“It protects the seed from different types of wildlife, but also supports the seed once it germinates and really helps deliver all of those nutrients and mineral sources that it needs, along with some probiotics to really boost early-stage growth."

No idea how credible this is of course.

View on HN

It’s not credible. I know people that work at this kind of company. None of the countermeasures have been successful and the germination rate is less than 1%.

I don't know what the germination rate is when you plant the seeds by hand, but it seems that aerial seeding requires 1.5 to 2 times more seeds. https://efotg.sc.egov.usda.gov/references/public/IL/Agronomy...

This does not seem excessive in regards of the time it saves.

Cover crops aren't trees. To elaborate...

Cover crop seeds (basically grasses) and tree seeds are apples and oranges. You can spray grass seed on the surface of the dirt and it will grow into a meadow. The same can not be said for trees. They will be eaten by birds and mice and digested, not dispersed. They also don't germinate well when left on the surface of the soil.

It's also a matter of numbers - tree seeds are by comparison, incredibly expensive to harvest. There's an industry of people who will camp out in forests, stalk squirrels, and see where they're storing their nuts. Or they climb trees to manually harvest fresh pinecones. It's labor intensive and you need a permit to do it.

> There's an industry of people who will camp out in forests, stalk squirrels, and see where they're storing their nuts. Or they climb trees to manually harvest fresh pinecones. It's labor intensive and you need a permit to do it.

Given we don't have squirrels in Australia, I'd imagine the knowledge you're providing here around gathering and permits are not accurate to the company in discussion.

What if a drone were to shoot the seeds into the ground with such force that they would be buried 4-5 inches deep (which is, I would guess, the necessary depth for successful propagation)?

That's probably too deep for most plants. 1-2 inches is probably more realistic, with 2 being close to the max.

what does this mean?

I've added more context to the comment

You dont plant tree seeds by hand. You plant saplings, small trees well past germination.

Well presumably someone is planting the seeds for the saplings unless they are all clones.

That's called a tree nursery. Seeds are planted and germinated in controlled conditions on a few acres.

Once they're baby trees, then you go plant them by hand in areas that need reforestation.

It's more like 1 in 3, an order of magnitude better. And they'd be sown under cover in more controlled conditions before planting out; not just tossed in the ground.

Can you site any sources supporting your claim?

No, this is hearsay, and I'm only passing on information I've heard from people that work at companies in this space and in silviculture (lots of companies are trying to crack the "plant trees with drones" nut).

They claim 80% germination rate at this company.

400 trees a day planted by drones seems ok. That’s 146,000 per year. Per drone. Seems ok to me.

Where did you see anything about a drone planting 400 trees a day? This would truly be remarkable if it existed.

40k/day Ad * 1% success rate criticism = 400 trees/day. So the criticism is a claim it is remarkable?

Except these drones are not planting trees is my hang up on the comment.

These drones are randomly throwing out seeds. The 1% is a seed germination rate. A germinated seed does not equate to a tree being planted.

I think a sapling planter plants more than their salary in saplings, so a comparison to the human sapling planter is false equivalence. I also don't see the difficulty in waiting until a seed has germinated in a nutrition disc before releasing it if you want to focus on getting 80% germination.

Regardless of what choices are made it seems likely to me that someone's automation in this area is already capable of doing about as well as careful management in some places with high labor costs. From equality, an automated system can easily double it's yield every few years while a managed system is going to be lucky with 20% improvement?

Really improving seed performance and dispersion by orders of magnitude from natural with automation never has to get anywhere near as good as a manual system of trying to raise every tree.

Even 0.5% of 40000 seeds is 200 trees per day.

That's nothing, and in addition to that, it's a huge waste of seeds (these companies are buying up all the available seed for certain species of trees, which means net fewer trees planted). Manually planting already germinated baby trees is more efficient both in terms of hours and labor spent and has a higher germination rate.

200 trees is about how many an experienced treeplanter will plant before thier first coffee break.

Yes but do they use AI and machine learning? Do they hover in place and look cool? They don’t even have a cool name like “drone.” I guess you probably also can’t mass produce and automate experienced tree planters too. While they’re swilling coffee they could be launching hundreds or even thousands of AI powered drones shooting fancy seed pods everywhere. You can’t stop progress John Henry.

Depending on the seed that might be enough. The same way just adding another machine to a cloud might be more economic than employing 2 engineers for a year in order to make a system more efficient.

A coffee bean weights 132.5 milligrams. 40k of those is ~5.3 kgs of beans. Sure it would be inefficient to employ all of those beans if only 200 of them end up being viable. But if those "wasted" 5 kg beans saves labour, then it might make economical sense. And if the alternative is running some sort of heavier planter machine, it might end up producing less emissions.

For the record I have not done real numbers here and my gut instinct is that the drone thing is a gimmick. I'm just trying to keep an open mind.

I don't think emissions should be a concern at all for a reforestation project, unless orders of magnitude more waste is emitted compared to that absorbed by a tree over, say, 10-20 years. Forests pay long term dividends beyond carbon capture.

Reply on news.ycombinator.com