What shipped
Newest first. Short entries — what you can now do, and whether it fully works yet. The longer stories about how any of it went live on the blog, and the how-does-this-work pages are in help.
-
Fusarium yellows turned out to be two diseases as well
Same shape as the celery blights earlier today, caught before either description was written rather than after. Fusarium yellows on celery and on cabbage-family crops are two different fungi that do not cross, so they now have two entries. The celery one comes with the test that settles it: cut a stalk across and the rings inside are stained orange-brown, which is what tells it apart from aster yellows — the other thing that turns celery yellow, and one we already describe. Ask about yellowing celery and you get both, because you genuinely cannot tell them apart without cutting one open. The cabbage-family entry carries something useful too: the source says cauliflower, broccoli and brussels sprouts are not known to get it, so a yellowing brassica bed is not automatically all of it.
-
Two screens disagreed about whether your pruning window was open
Open "What did I miss?" on New Year’s Day and a peach’s dormant prune sat under "Gone for this year — nothing you do now reopens it". It was open, and stayed open until late March, and the daily list said so on the same day about the same tree. The cause was two ways of writing down a window that crosses New Year. The part of the app that works out your jobs writes that window as running past day 365; the catch-up screen expects it written the other way round, and neither knew the other existed, so it asked whether 1 January falls between mid-November and late March, decided it does not, and filed a live job as gone. Fixed at the join. Checked across every once-a-year job we hold, on every day of the year: the two screens now agree everywhere. The worst affected before this was wrong on 128 days out of 365.
-
Tiny holes in your leaves: a new page on when flea beetles matter and when they don’t
Flea beetles chew small round holes clean through a leaf, and every gardening site shows you the damage and then reaches for a spray. The thing none of them tells you is which case you are in. On a seedling that damage can kill the plant outright. On an established plant the identical holes are cosmetic and the plant grows through them. Same insect, opposite answers, and the difference is what the plant is doing rather than how many holes you can count. So the new page says how long the dangerous stretch actually lasts, using the growth-stage data we already hold for the 37 crops that get this pest. Arugula and radish are seedlings from day 7 to day 21 — fourteen days and they are through it. Cabbage runs day 8 to day 49. The median across the 27 crops we can put day numbers on is eighteen days. Count from the day you sowed, not from the day you noticed. Inside the window, cover it; outside it, leave it alone. Two things the page deliberately does not do. It does not name a spray or a rate, because flea beetles are a group of species with different hosts and different state guidance, and a product name from us would be a number we had not checked for where you live. And when it gets to what eats them, it says we have not sourced that and will not invent it — then gives the thing we can stand behind, which is that no predator arrives fast enough to save a seedling in a two-week window.
-
Your harvests now add up
The summary above a plant’s history used to say how many times you had picked something. It now says what those pickings came to — 3 harvests, 4 lb. Two units stay two facts: four pounds and six fruit are shown as both, because adding them would mean guessing what one of your tomatoes weighs and then showing you the guess as a measurement. And if some of those harvests had no amount on them, it says “at least”. A total that quietly leaves out the ones you skipped is worse than no total, because nobody goes back and re-checks a number that looks finished.
-
In the ground is not a pot
If you marked a plant as going straight in the ground, or into a raised bed, but the spot it sits in was named a pot or container, the app handed that plant the pot watering advice — check often, water little. That is written for a five-gallon pot drying from every side, and open ground holds moisture for a week. It came from reading “in the ground” as though you had said nothing, so the name of the spot filled a silence that was not there. What you pick now genuinely wins. Nothing to redo: the advice is worked out fresh each time you look, so it is already right.
-
It stopped saying “Do the carrot have proper leaves yet?”
The app asks about your plants by name, and one of those questions had the grammar wrong: "Do the carrot have proper leaves yet?" It was not a typo in one place — the name is dropped into a fixed sentence, so every plant reaching that stage got it. Pluralising it would only move the problem to anyone with a single plant, so the question is rewritten to work either way: "Proper leaves on the carrot yet?" Checking the rest turned up the same fault pointing the other way — "Is the tomatoes filling out?" — which is fixed too.
-
Aphids: the whole internet says kill them, and Minnesota says mostly do not
Search how to get rid of aphids and you get thirteen results telling you how — remove them this instant, eliminate them, nine natural options, once and for all. Not one of them prints the sentence University of Minnesota Extension actually publishes: "In most cases, aphids cause little to no damage to plants and can be ignored." Our new page leads with that. It holds because something is already eating them — lady beetles, lacewing larvae, syrphid fly larvae and parasitic wasps are named by all three extension services we used — and because aphids are weak insects that a good rainstorm knocks off, which is why a dry summer seems to bring them. The page also gives the identification everybody skips. Small green bugs describes fifty insects. Every aphid has two tiny tailpipes sticking out of its back end and nothing else on your plants does, which two universities state independently. When you do need to act, start with a hose, and Colorado State explains why that is not a token gesture: very few aphids knocked to the ground ever find their way back onto the plant. The honest limit gets its own heading rather than a footnote — aphids carry plant viruses, the aphid leaves and the virus does not, and that is why "ignore them" is a starting point rather than a rule, especially on seedlings and on anything you are growing to eat.
-
Your week stops filing “no rush” jobs under today
Two changes to the same screen. Jobs marked no rush have their own section now instead of sitting in today’s list — on a twenty-five plant garden that is eighteen rows of forty-nine moved out from under a heading that said the opposite of what they were. Nothing is hidden: they keep a heading and a count, because a screen whose promise is showing you everything in one moment cannot answer a request to declutter by dropping a third of it. And Today now shows the weather: tonight’s low first, because that is the number that decides whether a tray comes in, then the next few days. That was the other half of a complaint about a card saying "hot weather in three days" on an app that had never shown a temperature anywhere. The card carries its number now and the line above lets you see the shape of the week it came from.
-
"Nothing yet" is an answer, and now it is on the home screen too
When the app asked whether your seeds were up, the home screen gave you one button: "Yes, it is up". If you had gone out, looked at the row and seen bare soil, there was nothing there you could honestly press — so the only options were to say something untrue or to ignore it, and ignoring it brought the same question back the next day. Both answers are now on both screens. Saying "Nothing yet" is a real answer that gets recorded: the app stops asking, and speaks up again on its own if the seeds go past the point where that crop should have shown. Before this, saying no threw the answer away.
-
Parsley gets its own leaf spot, instead of a tomato’s
The last of four plants that were being handed a description of a disease they do not get. Septoria leaf spot on a tomato and on parsley are two different fungi, and we only had the tomato one written up — so a parsley grower got "dark spots on the leaves", taken from a page about tomatoes. Parsley now has its own entry, written from the University of California page for that exact disease, and it describes what you would actually be looking at: small sunken tan-to-brown spots, often ringed in red-brown, going gray in the middle as they age and coming up peppered with tiny black dots. The stalks get them too. It also records something the source is careful about — that page covers cilantro as well, but the fungus only affects parsley, so we say so rather than letting the title imply more than the page does.
-
Profile went quiet about limits instead of answering
Somebody asked how much of their allowance they had used and got a line describing what a free account "comes with" — a ceiling that had already been removed. We fixed the wrong number by hiding the line, which is better than stating something false and is still not an answer to the question. Profile now says the true thing: there is no limit at the moment, and we will say so there first if that ever changes. If the server reports a number for your account, that number is what you see instead — a fact about your account beats a general statement about the tier. The rule deciding all of this used to be a condition written inside the screen itself, where nothing could test it, and it had been wrong twice. It is one function now, with a test covering all four states.
-
You could tell us the tree was grown from a seed, and nothing listened
A tree grown from a pip has no rootstock, so it has nothing to sucker — and "Remove suckers" should never have been on its list. You have been able to say so on the plant page for a while, and we said on the help page that saying so was enough. It was not. The answer was saved and then dropped three separate times on its way to the part of the app that decides what to show you: the rule that reads it was never given the crop it needed, the year-ahead list never asked the rule at all, and the object handed to the engine was rebuilt by hand without the answer on it. Each one alone was enough to keep the job on screen. Fixed and checked on the real page in both directions — pick "grown from seed / cutting" and the job goes, unpick it and it comes back. Anyone who already answered does not need to answer again. The question above those buttons also read "What size tree is it? (on the nursery label)", which is a strange thing to ask somebody holding a tree they grew from a seed; it now asks how the tree started.
-
When to plant potatoes: the one crop that does not wait for the frost date
Every spring planting guide is organized around the last frost, because a freezing night kills a tomato outright. A potato is not a tomato. Maryland records it as a "Half-hardy annual" that can withstand light frosts, and Utah State says outright that potatoes should be planted in the garden two to three weeks before the last frost. Before it. The fact that makes that safe rather than reckless comes from Iowa State, and most pages quote the first half and stop: freezing temperatures below 32°F damage potato shoots, from blackened leaf margins up to death of all aboveground growth — and then, "Fortunately, severely damaged potatoes will send up new growth within 10 to 14 days. There is no need to replant the potatoes." A frost does not end a potato crop. It costs it two weeks. What the crop is actually waiting for is soil at 50°F, which is a thermometer rather than a calendar and certainly not a hardiness zone — a zone describes your coldest winter night, which is the one thing this crop is not waiting on. The page also explains something no other page seems to: why the planting depth you are given keeps changing. Three to five inches early, five to six later, and Utah State gives four to six across the board.
-
The frost check on hardening off now covers every plant, not 38 of them
Hardening off ends with a night left outside, and it is the one step in those two weeks that can kill a tray rather than just slow it down. For the 38 crops where we have a sourced temperature floor, the app already refused on a cold night and said why. For the other 114 it did not check at all — it told you to check the forecast yourself, which is odd, because the app has the forecast. Now nothing goes out below 35 degrees, whatever it is. That number is not a claim about your particular plant: it is freezing, plus a margin, because a forecast is a forecast and your patio is not the weather station. Where we do have a sourced figure for a crop it still wins — basil stops at 50. And a night with no forecast still stops nothing, because a warning that fires on a guess is one you learn to tap through.
-
When to harvest garlic: the leaves are the wrappers
Every page about harvesting garlic tells you to wait until the lower leaves yellow. Ohio State Extension says why, in a sentence almost nobody quotes: "The leaves are actually connected to the garlic bulb wrappers below." Each leaf you can see is a wrapper you cannot. That single fact turns the usual warning into arithmetic — waiting until the leaves turn brown, in Ohio State’s words, "will result in rotted or missing wrappers", and wrappers are what keeps garlic through the winter. The new page also carries the test that settles it, which is not the leaves at all. Minnesota: pull a bulb, cut it across, and if the cloves fill their skins the bed is ready. One bulb, one minute, done. And it names what each mistake costs: too early gives small bulbs, too late gives cloves popping out of the bulb, and a bulb with split wrappers will not store. On curing, we show the disagreement instead of averaging it — Utah State says one to two weeks, Minnesota says three to four, and those two do not overlap. The finish line is not a date. It is skins papery and roots dry. This is the first page in a new section for harvest timing, which exists because of something in our own data: "lower leaves going yellow" shows up at 33 different crop-and-stage combinations in our records, and it means go and dig on fourteen of them and absolutely nothing on nineteen.
-
The last night of hardening off is finally offered — and checked against the forecast
Hardening off is ten days, and the tenth is the night the plant stays out. The app never asked for it. It counted to ten, decided the ramp was over on the day the last step was due, and jumped straight to "ready to plant" — so the one step that matters most was skipped every single time. It is offered now, on the right day, and the frost check we added this morning finally has a card to sit on: below your plant’s own temperature limit the app will not suggest it at all. Two other things came out of the same look. If the plant is still in a seed tray, the app now tells you that the extension services genuinely disagree about leaving those out overnight, and that none of them mention how little soil a tray holds — so you decide, with both sides in front of you. And finishing the ramp is no longer written like a deadline: a hardened plant can live outside, so it waits there until you have a weekend, rather than the app implying you are already late.
-
Your fruit tree page said “2 to 4 years” without saying which kind of tree
Almost every fruit tree sold in a garden center is grafted — a named variety joined onto someone else’s roots — and every years-to-first-harvest figure published by an extension service is measured on one of those. A tree grown from a seed is a different plant on a different schedule. Our data has known that since May, and the app has used it: tell the app your tree is on its own roots and it stops nagging a three-year-old orange seedling to produce fruit. The plant pages on this site never said any of it. Orange read “2 to 4 years” flat, while we held a sourced 7 to 10 years for a tree grown from a pip. Thirteen fruit pages now say which tree the number is for — on the summary strip, in the opening paragraph, and beside the figure itself. Pomegranate goes the other way and is worth reading: the everyday one is struck from a cutting and already on its own roots, so its 2 to 3 years is the honest number for what you would buy. Apricot, peach, plum and cherry gained the label too, each checked against the sentence it came from rather than the crop’s reputation — UConn’s table is titled “after planting” and Minnesota’s says “if you plant a 1 or 2-year-old tree”, which is nursery stock either way. What we still do not have is a trustworthy figure for a peach or apple grown from seed. Two extension pages were read and neither gives one, so those pages say nothing rather than guessing.
-
How to start seeds indoors: the number nobody prints is two inches
Every seed-starting guide tells you to give seedlings adequate light, and then photographs a tray on a windowsill. University of Minnesota Extension does neither. It heads a section "Avoid the windowsill" — a sill is the coldest place in the house at night and the hottest by day — and it gives the number the rest of the internet rounds off: "Keep lights no more than 4 inches above the tops of your seedlings: as close as 2 inches is ideal." Twelve to sixteen hours a day, and not continuously, because most plants need a dark period. It also names the cause of the thing every beginner photographs and asks about: "Lack of light is the major cause of elongated, skinny stems." That is the leggy seedling, diagnosed in one sentence — not a watering problem, not bad seed, just a light source too far away. The honest part is that this is a prevention answer rather than a cure: a stem that has stretched does not un-stretch, and the tomato trick of planting deep is one crop rather than a general fix. And the comparison that makes the number make sense is the least intuitive line on the page: natural light is relatively weak next to artificial light kept close to the plants, even in a greenhouse. This is the how. The when already had its own page and this one links to it rather than repeating it.
-
Your photo guesses the pot size, and you just say yes
Asking a pot for gallons instead of width and length was the right question, but it still arrived as an empty box. Nobody knows their pot size by heart, and the app has usually already seen the pot — it is standing in every photo you take of the plant in it. So when a picture has told us roughly how big a pot is, that pot now shows you the number with one button: "That is the size". Tap it and it is saved. The reason behind the guess is shown with it, so you can tell whether it looked at the right thing. It is never typed into the box for you and nothing is stored until you agree — a guess sitting in a field looks exactly like a number you gave us, and that is worse than an empty box. If it is wrong, say "Not right" or just type yours over it; either way we keep a record of what we guessed and what you made it.
-
Twenty-three crop pages now show the autumn planting window as dates, where a source publishes one
Most of our planting dates are counted from your own first frost — "sow 70 to 84 days before it" — because that sentence is true in Maine and in Georgia and you supply the frost date. A lot of extension services publish the same fact as plain calendar dates instead, and until now we held those and showed them to nobody. Twenty-three crop pages now carry them, each window labeled with the place it was published for. Spinach is the clearest case: Clemson gives South Carolina two answers, August 1 to September 30 in the Piedmont and August 15 to October 15 on the Coastal Plain — six weeks apart, same crop, same state. Be honest with what that means for you: right now every one of these windows is Clemson’s, so unless you garden in South Carolina they are a demonstration rather than an answer. The page says so, and it will say "two states" or "six states" on its own as we add them, because that sentence is counted from the data rather than typed. What we will not do is convert them. Turning "sow in July" into a number of days before your frost means assuming a frost date, which means assuming a zone, and handing that back to you with a citation on it would make a guess look like a source. One more thing worth saying plainly, corrected the same day we shipped it: the chart tells you WHEN, not whether to put seed in the ground or set out plants. Clemson only names the route where the route is the whole crop — it has separate rows for onion sets and onion plants and a plain row for everything else. Snap beans always go in as seed; fall broccoli in South Carolina normally goes in as a transplant. Both sit in the same column. We called it a sowing window at first, which claimed more than the source did, and the pages now say planting.
-
Sweet potato slips: the sprouts are the plant, and why the last one never sprouted
A gardener asked our app on 13 August whether he could plant the sweet potatoes in his cupboard that had little roots and shoots coming off them. The chat said we had not written it up. Now we have. Those shoots have a name — slips — and they are how the entire crop is grown. Mississippi State: sweetpotatoes are produced from vegetative stem tip cuttings, or slips, and slips come from sprouted storage roots saved from last year. The sweet potato is not a seed; it is a nursery. And the exact thing he asked about is answered in one line: slips may or may not have roots when they are cut. Either way they grow. The page also carries the sentence missing from nearly every article about putting a sweet potato in a jar of water. Maryland: "Wash any purchased roots to remove anti-sprouting chemicals." Supermarket sweet potatoes are treated so they will not sprout on the shelf, which is why the jar trick works for the person writing about it and not for the person reading. And it does the arithmetic out loud instead of stopping at the encouraging part: six weeks to raise slips, then soil at 65°F, then 90 to 120 days of growing, all finished before the soil drops below 60°F. For most of the country a sweet potato sprouting in August is next spring’s crop, and starting it now in a warm room is a good way to have slips ready in time.
-
Septoria leaf spot: the thing that identifies it is inside the spot
Plenty of things put spots on a tomato leaf, and the two that get confused most often are Septoria leaf spot and early blight. They start in the same place — the lowest, oldest leaves — and both work their way up the plant, so where it starts tells you nothing. The spot itself does. Missouri Botanical Garden describes Septoria spots as circular, an sixteenth to a quarter inch across, with dark brown margins and "tan to gray centers with small black fruiting structures". Those tiny black dots inside the pale center are the fungus’s own spore cases, and early blight does not make them — its spots enlarge into concentric rings with a yellow area around them. Count them too: Septoria makes many spots per leaf, all about the same size. The new page also carries the control caveat almost nobody prints. Every page tells you to remove the infected leaves; Missouri Botanical is the one that says where to stop, because "removing leaves above where the fruit has formed will weaken the plant and expose the fruit to sunscald". Strip from the bottom, stop at the fruit. And the overwintering surprise: the fungus survives on old debris, on nightshade weeds, and on your stakes and cages.
-
Japanese beetles: the trap you were about to buy makes it worse
Metallic green head, coppery wings, leaves eaten down to lace. If that is what is happening in your garden, the thing most people do next is hang a trap — and two extension services say that is the mistake. Minnesota: "Don’t use Japanese beetle traps. Hanging a trap in a home garden is not an effective way to protect plants," because research at the University of Kentucky found more beetles fly toward the traps than get caught. Colorado State says traps often increase damage by drawing in larger numbers than they capture. The bag really does fill up, which is exactly why people trust them — you can see what was caught and not what was called in. The new page also answers the question people actually have, which is where to hang the one they already own: the serious trapping research is on commercial orchards, it rings the whole perimeter, and even it warns against putting a trap in the middle of the thing you are protecting, which is where a garden trap goes. What to do instead is knocking them into soapy water while they are still cool — early morning or evening — and every beetle you remove is a few more that never get invited, because chewed leaves give off chemicals that call others in. Fifteen of our crops carry this pest; the page says which ones Minnesota names as favorites and which it does not.
-
A heat warning you could not check
The card said "Hot weather in 3 days" and left it there. The app has never shown a temperature anywhere — weather decides when we warn you about heat, when a tray should come in and whether it is worth planting this week, and none of it was visible — so a warning about the weather was something you could only take on trust. A warning you cannot check is one you stop believing the first time it looks wrong. It now carries the number that raised it: "Around 99°F, and nights near 70°F." Which figure leads depends on what triggered the warning, because a warning about warm nights that opened with the daytime high would point you at the wrong thing — a night that never cools is harder on a plant than a hot afternoon. The numbers were already being worked out to decide whether to warn you at all; they were simply never shown. A place to see the next few days at a glance is still to come.
-
Three cabbage caterpillars re-checked, and one thing we had wrong
The three caterpillars that eat cabbage-family plants all shared one source page, and that page no longer exists — so all three were rewritten from University of Wisconsin pages, one each. Two corrections came out of it. We had said the imported cabbageworm works its way into the middle of a cabbage head; the source says plainly that it does not bore into heads at all, and the damage stays on the leaves. And we had said the diamondback moth caterpillar eats the inside of the leaf; it actually feeds on the underside and leaves the top skin intact, which is what makes the little windowpanes. Same result, different place to look — and turning a leaf over is the accurate instruction. The descriptions also picked up the fastest way to tell the three apart: the looper humps its back as it walks, the cabbageworm just crawls, and the diamondback thrashes and drops off the plant the moment you touch it.
-
Two citrus mites, and the reassuring half of one of them
Citrus red mite bleaches leaves and fruit into a pale speckle that runs together into dry brown patches when it gets bad. The useful detail is what it does not do: it makes little or no webbing, which is how you tell it from the spider mite you were probably about to blame. Citrus rust mite is stranger — it turns the skin of the fruit silvery on a lemon, rust brown on a colored orange, black on a green one, and it is far too small to see even with a hand lens. It also comes with the best news in the entry: the damage is skin deep and the fruit inside is fine, so it is a question of how your oranges look rather than whether you can eat them.
-
When to harvest potatoes: two crops, one plant, and two weeks of doing nothing
A potato plant grows two crops and they are ready months apart, which is why the advice you find never quite matches the plant in front of you. Minnesota puts both in two consecutive sentences: you can dig new potatoes about seven to eight weeks after planting, and harvest mature tubers after plants have dried. The new page asks which one you want before it answers when. For storage potatoes it carries the step almost nobody leads with, from Maryland: "If you plan to store potatoes, leave tubers in the ground for 2 weeks after plants die back. This will allow them to develop a thickened skin." Two weeks of doing nothing, underground, after the plant is already dead — and that is the difference between potatoes that keep until spring and potatoes that go soft by November. It is also why the same page says not to dig in wet soil and not to wash them before storing: all three instructions are protecting the same surface. On curing, three services give three answers, from ten days to three weeks at three different temperatures, so we show the spread rather than inventing an average. And one fact worth carrying around: Minnesota says moisture stress while the tubers are enlarging causes knobby or hollow potatoes. If you have ever cut into a potato and found a cavity, that was a dry spell months earlier, with nothing to see at the time.
-
Two more pests you can now look up: plum curculio and carrot rust fly
Both were on the list of problems our plant data names but we had never described, and both are the kind where knowing what to look for is most of the battle. Plum curculio leaves a curved scar like a thumbnail print on small green plums, peaches and apricots, often with a bead of amber sap weeping out of it. Carrot rust fly is the opposite problem: there is nothing to see above ground at all, which is worth knowing on its own — the tops look fine while shallow tunnels packed with rusty mush run over the root. That takes us to 103 problems we can describe, covering four in five of the times a plant in our database names one.
-
When to plant carrots: the sowing temperature is not the germination temperature
Almost every page on this gives you one number. Utah State gives two in the same breath, and the gap between them is where carrot rows are lost: "Carrots can be sown after soils reach 40°F. Seeds germinate best at 55-65°F and require 14-21 days to emerge." Sown and germinate are doing different jobs in that sentence. Sowing at the bottom of the range is allowed; it means asking a seed to sit in cold soil for three weeks before it shows. And the depth is what makes that hard — a quarter of an inch, which is the exact layer that dries first. So the practical version of "when to plant carrots" is not a temperature at all. It is: when can you keep the top half-inch damp, every day, for three weeks? The page also carries the fix two extension services give independently for the same problem, which sounds like folklore and is not. Sow a few radish seeds in the row with the carrots. Utah State says it helps minimize the crusting that stops carrot seedlings emerging and it identifies where the rows are; Maryland suggests the same for row-marking, with the caveat to harvest the radishes before they compete. And there is a summer gap at the far end that most pages skip: above 80°F germination drops, and high summer temperatures cause bitter or off-flavors to develop.
-
Spider mites: everyone publishes one test, and the second one matters more
Spider mites are about one fiftieth of an inch long, so you never diagnose the animal — you diagnose the leaf. Fine pale speckling, like it has been sanded, then bronzing, then leaf drop. Webbing is the late sign, not the early one, which half the internet has backwards. Every page on this subject gives you the same test: shake a leaf over white paper and watch for specks that walk. Almost nobody gives you the second one, and it is the one that matters. Ohio State: wipe the specks with a fingertip and look at the smear — green is a plant feeder, the pest, while yellow or orange is most likely a predator that eats them. That distinction is the whole page, because two extension services say that killing the predators is how a small mite problem becomes a bad one. Minnesota warns that long-lasting insecticides "also kill natural enemies and could possibly make infestations worse in the long run." Colorado State goes further than anyone else we found and names products — carbaryl, Malathion, imidacloprid — as things that have contributed to outbreaks. And Ohio State adds the part nobody mentions on the shelf: ordinary insecticides usually do not kill mites at all. What does work starts with a hose. The page also explains why hot dry spells bring them: dry air lets the mite feed harder while stressing the predators that would have eaten it.