Hacker News new | past | comments | ask | show | jobs | submit login
“Unprecedented” locust invasion threatens crops in East Africa (scientificamerican.com)
169 points by LinuxBender 17 days ago | hide | past | web | favorite | 125 comments

If anyone wonders what happens if a proper airplane tries to fly through a swarm of locust: http://avherald.com/h?article=4d1de8cc&opt=0

> The windscreen wipers were not able to clear the windshield anymore. The crew went around, climbed to 8500 feet, depressurized the aircraft, opened the cockpit side window and cleaned the windscreen by hand. The same happened on second approach to Dire Dawa. The crew again climbed to 8500 feet, cleaned the windscreen by hand again and diverted to Addis Ababa.

"Locusts are the swarming phase of certain species of short-horned grasshoppers in the family Acrididae. These insects are usually solitary, but under certain circumstances become more abundant and change their behaviour and habits, becoming gregarious" https://en.wikipedia.org/wiki/Locust

So, in normal circumstances, grasshoppers are benign but under certain conditions, they can form swarms.

Yes, been studied and shown it is overcrowding that causes this and you can by rubing the back legs of a grasshopper for about 30 mins(iirc maybe shooter time) and it will start to develop wings and change.

The rubbing of back legs in nature happens when you have high dense populations - overcrowding due to rushing for limited food in an area. Which evolving wings and flying elsewhere would be nature at play.

Real trick would be to curtail grasshopper numbers before they swarm and perhaps a more balanced approach to grasshoppers natural enemies and encouraging habits for those will do well. Chemicals for me are a desperate solution to a problem that got overlooked. Much the same way that ant do wonders with many pests for plants, yet production methods end up curtailing environments that see them thrive in area's that would benefit. So we end up with chemicals for things like aphids when ants will eat them all up for you.

I'm not pro-chemical, but it's hard to imagine other solutions that are actionable in the timeframe of a weather prediction (this timeframe is critical because these swarms were spawned due to weather anomalies). Indeed, according to TFA it sounds like the UN isn't even trying to curtail the locusts at all but are just preparing for disaster recovery.

Crop insurance seems like a more tractable solution.

I think the problem is rather that they destroy crops than insurance; farmers probably can't grow crops in time before another swarm destroy them.

As for stopping them, I would try covering a wide enough area with a giant electric net (just like those insects killers) also placing additional screening nets on the upper and lower side so that the screens would keep birds away, and the actual electric net would have its holes large enough so that smaller insects would remain mostly untouched. The idea is that if they're stupid enough, they will eat all crops in vicinity then eventually die attempting to attack the only remaining field.

That doesn't make any sense in the real world.

Why it shouldn't? The purpose is to attract and exterminate locusts when it becomes the only field in vicinity with intact crops. Build a strip of nets large enough to cover some area where the locusts will travel during migration, and they'll attempt to feed from the fields underneath. Sure it will be costly, but we're talking about decades old technology and cheap materials.

Grasshoppers already HAVE wings. Touch one you find somewhere and watch it fly away. They hop and fly.


Locusts and Grasshoppers are different insects.

That page says at the top that locusts are grasshoppers.

"Locusts are a collection of certain species of short-horned grasshoppers in the family Acrididae that have a swarming phase."

They are grasshoppers that have decided to change behaviour and physical characteristics for some reason. They used to be considered different species.

The very first sentence in the article says that locusts are a type of grasshopper.

> So we end up with chemicals for things like aphids when ants will eat them all up for you.

The ants species I know of "milk" aphids for their sugary excretion, and as a counter service, protect aphids from their predators (such as ladybugs). It's... complicated. I agree with you that chemicals shouldn't be the first thing to reach for. It's interesting to try to find out how we got into this situation in the first place. The book Seeing like a State (https://en.wikipedia.org/wiki/Seeing_Like_a_State) has some interesting (but opiniated, in my opinion) notes on disastrous (on many axes - socially, economically, ecologically, ...) modernist agricultural reform programs that took place in Tanzania, Kenya and Ethiopia in the second half of the 20th century — a region much affected by these locust clouds.

> you can by rubing the back legs of a grasshopper for about 30 mins(iirc maybe shooter time) and it will start to develop wings and change.

If true, this seems very effective for economic terrorism. I wonder why thats not more common.

A production rate of 32 locusts per person per 16-hour workday isn't enough to matter unless you have a small army of people rubbing grasshopper legs together all day, I'd think. Though it is a pretty funny image.

The grasshoppers self-activate as locusts under crowded conditions. Once they enter the swarming phase, the behaviour is self-perpetuating, generally until foid supplies are exhausted. With high mobility, this can take time and strip tremendous areas of land.

By stocking a sufficient population early on, in sufficient confinement,, it might be possible to create a perpetuating, expanding, swarm.

Just keep enough of them in a small enclosure. Be prepared to increase enclosure size rapidly for a few days or weeks.

Looks like grasshoppers generally live 50ish days as an adult. That's gonna put some serious limits on your max per-grasshopper-leg-rubber total accumulated grasshoppers.

The grasshoppers rub each other.

Most humans have two hands so you're off by at least a factor of 2.

If you implement "middle out" grasshopper rubbing, you can triple production

The rubbing only increased serotonin levels and 4-3 hours later after that boost, they transform. So if you was to increase their serotonin levels via other means, you would still only turn grasshoppers into locusts. Yes locusts metabolism in faster, so they are more veracious.

As a form of terrorism, well you would still need to bread and or/find a huge swarm of grasshoppers and if you found such a size large enough, they would already naturally be swarming. Though the prospect of some terrorist group breaking grasshoppers has an edge of Chris Morris about it that bemuses me.

Fascinating that it is related to serotonin! I wonder if an antagonist would prevent the behaviour? Might be challenging to dose large numbers of locusts.

Maybe target them with commercials for pro-depressants? Feeling restless, gregarious, like you just might take ravenous flight? Ask your locust doctor if pro-depressants are right for you.

Then we just need some sexually attractive locusts to lobby the locust doctors and problem solved!

What do you mean by develop wings? Don’t they already have them?

I was very surprised when I first learned this fact a few weeks ago. It's hard to think of another species that acts so differently in a pack, very similar to humans in my view.

Swarm behaviour is a field of study, across many species and phyla:


The Indian Ocean Dipole event did nobody much good. It's been attributed for the very low rainfall (and bushfires) across Australia in 2019 as well as the higher than usual rainfall (and locusts) in east Africa.

Also an impressive feat of weather modelling, however. This is from May 2019, apparently before the strong IOD event had fully formed, and before the bushfire crisis hit. https://www.abc.net.au/news/2019-05-16/positive-indian-ocean...

This is from last month and shows what happened with it https://www.abc.net.au/news/2019-12-11/indian-ocean-dipole-v... And they were right about the effects on rainfall while it was in place.

Thank you. Some scientific background, instead of just event reporting.

There're some entertaining control measures being suggested in the comments. To my knowledge, the best control measure is pesticide application programs that use modeling to suggest optimal treatment locations. If you can develop a better locust population model, FAO would likely be interested.

I think genetic modifications could help but it's probably a long shot.

Locusts are a solved problem in most parts of the world. Insecticides do work. I understand that this is HN and people are interested in drones, lasers, genetic engineering, etc. but in cases like this it's better to use proven solutions rather than experiment.

We only need a way to move Glyphosat from Western Europe to Africa.

A giant net and you could at least eat them after they feast on your crops:


There is no more ferocious predator than humans. If they became a delicacy all over the world they wouldn't stand a chance in hell.

According to the wikipedia article: "Locusts yield about five times more edible protein per unit of fodder than cattle, and produce lower levels of greenhouse gases in the process."

I always wondered why they're not collected and fed to chickens. Seems like a flock of fowl of any kind would gorge on these juicy bugs. Even a small inset net should harvest a good amount.

Yeah, a good source of protein, with the huge numbers a challenge would be drying and storage

You probably already have equipment to dry and store the grain.

Reuse of the same equipment to dry the insects might be doable.

And... you have Wuhan v2...

Wow, these stuff are beasts ! They look enormous !

There was also pictorial from The Guardian a few days ago, "Billions of locusts swarm through Kenya"[1] regarding this.


Oh my. One swarm is ~2400 square kilometers/~930 square miles, and could contain up to 200 billion locusts. And moving at 150km per day. Madness. You can understand why various cultures have viewed them as a punishment from god.

Here's this quote from https://www.nationalgeographic.com/animals/invertebrates/gro...: "Each locust can eat its weight in plants each day, so a swarm of such size would eat 423 million pounds of plants every day." This is describing a swarm a "mere" 460 sq miles in size.

> 200 billion locusts

For context - there are 4 billion IPv4 addresses in total if that helps with scale of things for some.

Talk about the internet of things...

That's unbelieveable... I can't imagine that there's any real way to control swarms that large, and the idea of burning the entire country clean to starve them out seems a bit counterproductive.

Aside from pesticides, nets. Some places where bugs eat produce if not protected have nets covering and enclosing the whole farm.

If the swarm is 960 square miles, and travels quickly, how much netting do you need?

There's a good scene depicting a locust swarm in Things Fall Apart.

The traditional village that is the subject of the (fictional) story sees swarms as a blessing -- they may eat crops, but you can harvest and eat the locusts themselves, and there are so many.

"The swarms increase in size twentyfold with each successive generation and could reach India by June."

India? Am I missing something or is this is an error in the text?

No this is not an error. The desert locusts Wikipedia says the locusts can fly 100km a day. June is four months away so that's plenty of time.


Looks like they're currently in Ethiopia (NE Africa) so India seems correct.

Looks like they rarely fly over water though: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1626356/

At 150 kilometers per day, they could cover the 3,500 km between Egypt and India in 20 days. Of course they're not covering that much distance every day, but they can jump between distant sources of food very quickly, so smaller gaps between crops or across deserts won't stop them.

There was a horrendous locust swarms few months ago in Punjab already before the winter.

The same megaswarm was in Saudi Arabia a year ago.

It seems that the swarm has since split, with one swarm going back to Africa

Getting real 40k Tyranid vibes

apparently they can move 90 MILES a day. I guess if they have food and no predators they can move and take over.

Having predators isn't an issue. There will be plenty, just there are so many locusts. So many that even after all the predators are filled to bursting, there won't be a noticeable drop in the numbers.

Fungi might be able to replicate fast enough.

At 150km a day, seems possible.

Or maybe they decide to go to Europe?

> The U.S. Agency for International Development says it released $800,000 to support FAO’s response in eastern Africa.

That's enough to send a couple diplomats down there to give the locals the middle finger, not including airfare.

The kenyan government is unprepared to handle this incident despite earlier warnings, the local minister cited a lack of an effective chemical in the local market, adding that the fight against the insect has been slowed down by the long procurement and import bureaucracies.


When the article says its the worst in 25 years, does that mean the historical record only goes back 25 years or that a worse invasion happened in 1995? From what I can see Wikipedia has an article for an infestation in 2004 and mentions 1989 as a previously notably bad situation in western Africa.

So my guess for deterring these pests would be a setup with high-intensity mid-ultraviolet LEDs (200-300 nm) pointed upwards, which should render large proportions of the exposed population sterile if not dead (sterile might be better than dead if it inhibits effective mating). 300 nm is about the limit for ionizing DNA. Humans can be advised to avoid the lamps.

The inverse square law being what it is, I think the power requirements alone of such a solution would render it impossible, and the LEDs would need to be insanely high intensity (I doubt they actually exist).

Given Locusts have an exoskeleton, would some kinds of focused sound weapon have a viable use in bringing down such swarms?

Imagine the source of protein you'd get! While probably not super handy (as it only happens during swarming times) and maybe even ecosystem-disruptive, turning them in to food would be useful.

Wouldn't look at this as a farming avenue, more a way to respond to such scaled issues using tools that may very well already exist or be modified to cater for such needs.

Certainly aware of crowd control sound weapons, I'd be supprised if the militaries of the world didn't have there own more powerful flavours begging for a PR test opportunity like this.

Now I wonder if there is a single resonant frequency that would work on all of them. The individual differences alone would be hard. Perhaps a general microwave type of weapon/harvester would work better?

Wow, we're approaching "Mars Attacks!" territory here. That could be a viable tactic though

Phased array microwave beams could cause enough instantaneous local heating to kill them, could optical detection to steer the beam.

And, if you keep a history of where the beam went, you know where go to to harvest the carcasses for protein.

At the densities you get, rainfall radar will pick them up.

Nice. Looks like frequencies from 10 to 20 khz may have a locust size wavelength. You might want to use half that frequency, and a very high power. Wonder if that would do anything to them

The swarms cover hundreds to thousands of square kilometres, and contain test to hundreds of billions of locusts. So no. There's no way to build enough sound weapons or focus them on billions of insects that move up to 150km per day.

30kw microwave radar would probably work too. You can buy the magnetrons on alibaba for a couple hundred bucks.

Yip - rainfall radar would pick up swarms of this scale - many instances of insect swarms being picked up upon weather rain radar.

I'm thinking cook their little brains in flight without damaging crops.

So much protein flying through air.. If only we have a way to capture them and turn into bars.

The swarms follow the wind, seek food, and signal necessary movement with pheromones.

Seems curbing food downwind (tillage, fire?), prompting movement, and starving the swarm, might be a control method.

Something like how wildfire is controlled.

Synthesize the pheromones, carpet spray them, and they'll keep moving non-stop until they die of exhaustion!

(Might have to do some due diligence eco-research to check, for instance, if the synthetic pheromone is specific enough)

Something like that. Or maneuver them into a narrow valley or canyon or to a too-long water crossing (shades of lemmings).

Or use the Mormon defence: seagulls.

Use drones with nets to catch them and turn them into food.

On that note, grasshoppers were a primary staple for many Native American tribes:



I'm having trouble finding articles, but basically they used to weave large nets and string them along places where insects gathered. It was basically land fishing.

I think people might be surprised to learn just how easy it was to thrive in hunter-gatherer societies. Salmon and shellfish were effectively unlimited, so if you knew which native plants to eat (like wild onions, carrots and especially camas where I'm from), you could meet all of your daily macronutrient and vitamin/mineral needs on 2 hours of work per day.

Edit: here's also a concise list of edible plants and herbs with their effects:


Build a jet UAV that is resilient against locusts and just fly them into the swarm.

Eat locusts instead of grain then xD they contain high amounts of protein and water and taste like chicken when fried. Add lots of butter if you can

Maybe we should all take up eating locusts? Creating more demand may help these farmers leverage this otherwise bad situation.

Anyone else up for a locust protein bar?

Couldn’t they do something like stick up big electrified nets at the crop areas?

Locusts fly into them and get buzzed? Or something similar

Not sure where to being with this, but locust swarms like this are a force of nature. Maybe you could save your backyard garden with nets or cover, maybe. But not acres and acres of farmland.

Monsanto would like to disagree.

What does Monsanto have to do with covering farmland with nets?

>>Couldn’t they do something like stick up big electrified nets at the crop areas?

Over millions and tens of millions of acres? Not doable. Maybe a genetic tweak and then release them to neutralize the rest. Poison will probably kill other beneficial or not so dangerous insects. Either way, a locust party will ruin this year crops for the unfortunate farmer being visited. This is even worst since Africa is already poor

the outer exoskeleton is quite tough and doesn't seem to conduct electricity

With enough volts, anything conducts...

We just need to find a way to eat them. Locust cakes could be a worldwide smash and great source of protein.

Why don't people eat the locusts? I mean... good protein, right?

Historically, they absolutely do. Not like one would have much of a choice after they've eaten all the crops.

Locusts are the only insect which is kosher, and this is most likely why.

Sure, would you like to be the first?

Lots of people already do. Think of it as a shrimp on land.

Will fuel-air incendiary work on insect swarms?

Probably as well as on all the vegetation and populace...

Surely. As would a nuclear bomb.

I wonder how drones would fair flying thru this, as on scale a locust in relation to the engines upon a drone is large in relation of a bird to an aeroplane engine.

Not well, there are documented cases of locust swarms bringing down small planes, and causing emergencies on large planes.

...as recently as 2 weeks ago, no less https://www.bbc.com/news/world-africa-51098209

Thank you, suspected as much but just didn't know. Was pondering drone fishing fleets harvesting such swarms, back to the mental drawing board then for me.

I understand that, we can make 'high protein flour' out of crickets[1] it seems like we should be able to process Locusts into some sort of protein source for human consumption.

But it is really hard to actually visualize/imagine these swarms without actually being there in one. The amount of infrastructure you would have to deploy essentially on a short term basis[2] is really infeasible with current technology.

[1] https://www.cricketflours.com/

[2] Swarms are a function of environmental conditions, they grow exponentially but they die quickly when conditions change, it is the series of monsoons at the right time that have made this one so big. In "regular" years they are much smaller.

That would mitigate the problem, but I don't think that could be used as a food source to replace what they destroyed, they eat up to their own weight per day, however insect digestion is pretty inefficient, they leave a lot of nutrients behind, so eating them up would only partially replace the losses.

Also catching a substantial fraction of a swarm that is spread over a very very large area and moves by up to 150 km a day would be very hard, and is largely impossible currently.

Exactly. The thought though was that if you created systems for processing locusts into flour around the area then that might incentivize people to catch and bring them locusts. If that became a regular thing it would keep populations down so that even in ideal swarm conditions the growth would not be as severe.

And it could provide an indigenous economic stream, always a good thing in Africa.

Locust are crickets so yes you would be able to

Honestly I think battery-powered drones with hardened or screened props would have the best shot against locusts. One of the main vulnerabilities for aircraft is the need to intake fresh air for combustion. This doesn't apply to drones. They also navigate using RF sensors and fly slow enough to avoid much collision damage. The locusts might start landing on them which could easily become incompatible with controlled flight, but that's the only thing that comes to mind.

It seems a bit tone-deaf to worry about drones when this could trigger severe food shortages.

Thank you for asking that and my apologies for not initially providing a context and by that poor omission opening up the very valid and fair possibility that I was being callous and uncaring. Also thank you for taking the time to raise that question instead of blindly judging and running with the worst case presumption, thank you.

I was thinking along the lines of fishing them as a food resources and drones with nets, but before I ran with that was wondering how well a drone would cope with such swarms and not something that the net did not offerup as a clear answer, either.

[EDIT ADD first paragraph ]

Ah, I see. I thought it was a random musing. My apologies then.

That being said regardless of the efficiency of such a solution I suspect that the scale of the invasion would make it hard to make a difference lest you have an army of drones at your disposal. The area to cover is insanely large.

No apology needed, and thank you for taking the time to comment your thoughts as without such feedback I would never of been able to clear things up.

[EDIT ADD] Indeed, if anything I should apologize as I'd left no context behind the question and leaving that open left the possibility of callous unthoughtfulness and you most kindly and rightly so, raised that aspect. Many would of just gone with a downvote - presumed the worst and ran with it in sillence, so again thank you for taking the time to question that. If I could mod you up more for that, I would.

[EDIT ADD - I was able to edit my initial reply as did seem you was being unfairly judged and I hope I've addressed that]

I think something more like a kite rig https://en.wikipedia.org/wiki/Kite_rig, but a net would be good. if they could be "thrown" up in the air quickly, have a sort of sticky inside so locust would be caught better, as they filled the net the net would become heavy and be brought down. The net should then be easily detachable somehow, attach another, throw up into the swarm.

Swarm passes by, the kite netters have a couple tons of locusts netted, a good harvest.

on edit: personally like because has such a 50s sci-fi feel to it.

I like the idea of kites, was thinking rockets and large net.

As for the Sci-fi, soon as you mentioned that I oddly thought of Dune, probably the rare resource mining aspect of such swarms and climate comparison going on there.

That sounds like industrial production of kungu.


Big fabric rotors, like those used on old windmills or a VAWT could scoop them out of the sky.

I just had a dig into what altitude they fly at and 2,000 meter ceiling would make things more complicated.

Hot air balloons dropping nets, or some weather balloons - again tethered. Though would probably want to run with hydrogen for costs, just safety and training go up, so greater initial costs there. So many details.

Kinda gets down to sky versions of submarine nets in predictable flight paths into crop area's as being the best spots to set such trapping nets up.

Though if it turns out you just need to lay down football sized green carpet and that attracts them - well that opens up other avenues in trapping or indeed - zapping them once they touch down.

How so? They didn’t say the safety of drones are more important than the people of East Africa. Drones could very well be used to help combat this crisis. Have you considered not being upset just for the fun of it?

This has already been addressed by Zenst, but this is a forum with a lot of engineers. Of course zenst was thinking about whether it's viable to use drones as a locust prevention measure, but you first have to start with whether drones can even fly in a locust swarm.

Get out of here with your pre-judgement

obligatory locust economy link (2013) https://www.ribbonfarm.com/2013/04/03/the-locust-economy/

Return the slab, or suffer my curse.

Napalm needed.

Applications are open for YC Summer 2020

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact