Watering Globes vs Terracotta Spikes: Which Works Better?

01 · THE SHORT ANSWER

Watering globes vs terracotta spikes: which works better?

For actual unattended watering — a week away, a plant you keep forgetting, a room you visit once a fortnight — terracotta spikes work better, and the reason is mechanical rather than marketing. Clay delivers water in proportion to how thirsty the soil is. A globe delivers water in proportion to how easily air can get into its stem, which has nothing to do with the plant at all.

That is not the same as saying globes are useless. They are cheap, widely available, often genuinely attractive, and perfectly adequate for what they actually are: a slow top-up that buys you a few extra days on a plant you would otherwise have watered by hand. If you use one that way — as a supplement rather than a system — you will probably be happy with it. Problems begin when a globe is asked to guarantee a fortnight, because it cannot, and its failure modes are silent.

The rest of this article explains why. It is worth understanding rather than taking on trust, because once the two mechanisms are clear, every practical question — how long will it last, what happens in a chunky aroid mix, why did mine empty overnight, why did mine not empty at all — answers itself.

02 · HOW GLOBES WORK

Watering globes are pressure devices, not demand devices

A watering globe is a sealed glass or plastic bulb with a single narrow opening at the end of a long stem, which you push into the soil. Fill it, invert it, insert it. Water starts to run out of the stem, which lowers the pressure of the trapped air in the bulb. Flow continues until that reduced internal pressure, plus the weight of the water column, balances against atmospheric pressure at the stem opening. At that point flow stops — until a bubble of air manages to travel up the stem into the bulb, raising the internal pressure again and letting another slug of water out. The device is a slow, irregular oscillation between those two states.

Everything about the release rate therefore depends on conditions at that one small opening. How tightly soil particles pack against it. Whether the soil is coarse enough for air to reach it easily. Whether the pot gets knocked, or a door slams, or a floorboard flexes — vibration dislodges a bubble and triggers a release. Whether the room warms up in the afternoon, expanding the trapped air and pushing water out. None of these variables is the plant's water requirement. The globe has no way of sensing that, and no mechanism through which it could respond even if it did.

This produces the two opposite complaints you see in every review section, both of which are correct. In a loose, chunky, bark-and-perlite mix, air reaches the opening far too easily, bubbles rise continuously, and a globe can empty in a single day — dumping its whole reservoir into a pot that did not need it. In a fine, dense, peat-heavy mix, the opposite happens: silt and fine particles pack into the narrow stem, capillary tension in the fine pores holds water back, and the globe sits there for three weeks having released almost nothing while the plant dries out. Same product, same instructions, opposite outcomes, decided entirely by the mix.

There is also a straightforwardly practical issue: glass globes break. They break when you push them into compacted soil, when a cat knocks the pot, and when you pull them out to refill them. The thin stem is the weakest point and it usually breaks below the soil line, which means glass fragments in the rootball. Plastic versions solve that but are more prone to being lifted by the buoyancy of the shallow stem in a wet mix.

03 · HOW TERRACOTTA WORKS

Unglazed clay releases water down a potential gradient

Unglazed terracotta is a porous ceramic: a continuous network of pore spaces small enough to hold water by capillarity but connected all the way through the wall. That structure makes it behave, in soil-physics terms, like a piece of soil rather than like a container^[12].

Water in the reservoir, in direct contact with the inner face of the clay, sits at approximately zero water potential — free water, held by nothing. The soil outside the clay is a different matter. As roots take up water and the surface evaporates, the remaining water in the mix retreats into progressively smaller pores, where it is held with progressively greater tension. Its matric potential becomes increasingly negative^[13][15]. Water always moves from higher potential to lower, so it moves through the saturated clay wall and out into the soil.

The consequence is the whole argument. The rate of that movement is proportional to the size of the gradient. Dry soil creates a steep gradient and pulls water out of the clay quickly. Moist soil creates a shallow gradient and pulls slowly. Fully rewetted soil creates essentially no gradient, and outflow drops to near zero — the clay simply stops delivering, without any valve, timer or moving part. Then, as the soil dries again, the gradient re-establishes and flow resumes. The device is genuinely coupled to the plant's water use, because plant water use is exactly what dries the soil.

This is not a novel idea. Buried unglazed clay pot irrigation — the olla — is one of the oldest documented irrigation methods, and it has been characterised in modern agricultural research. Bainbridge's review in Agricultural Water Management documents buried clay pot irrigation achieving water-use efficiency far above surface irrigation in arid conditions, precisely because delivery is driven by soil suction and almost nothing is lost to evaporation or deep percolation^[11]. Guidance on efficient irrigation makes the same general point: matching delivery to demand is what separates efficient systems from wasteful ones^[10].

04 · HEAD TO HEAD

Four criteria that separate them

Strip away appearance and price and four things actually matter: whether delivery self-regulates, whether it is repeatable, how long it lasts, and what it costs you in upkeep.

Note what is not in that list: capacity in millilitres, which manufacturers emphasise most. Capacity only tells you the ceiling. A globe with a large reservoir that discharges it all in eighteen hours has not watered your plant for a week; it has watered it once, heavily, and then left it dry for six days. Delivery profile matters more than volume, and delivery profile is exactly what the mechanism determines.

05 · FAILURE MODES

How each one fails, and how loudly

Every watering aid fails eventually. What matters is whether the failure is obvious and whether it is recoverable. Globes fail in four characteristic ways, and three of them are silent.

• Runaway discharge in a coarse mix. Bark, perlite and pumice leave large air-filled channels right up to the stem opening. Air enters continuously, the globe empties in twelve to twenty-four hours, and the pot is saturated for days — the classic route to the oxygen-starved roots that precede rot.
• Complete blockage in a fine mix. Silt and fine peat particles migrate into the narrow stem and pack. The globe stays visibly full, which reads as “still working,” while the plant dries to the point of wilting. This is the most dangerous failure because it looks like success.
• Temperature-driven surges. The trapped air in the bulb expands as a room warms through the afternoon and contracts overnight, pushing water out in pulses unrelated to demand. A sunny windowsill in summer amplifies this considerably.
• Breakage at the stem. Glass stems snap where they meet the soil, most often while being inserted into firm mix or withdrawn for refilling, leaving sharp fragments buried in the rootball where you will meet them at the next repot.

Terracotta's failures are different in character: they are gradual, visible and reversible. A spike that has lost soil contact leaves a visible gap and simply stops emptying. A spike whose pores have mineralised empties more slowly than it used to, which you notice at refill time. Neither dumps a reservoir into the pot overnight, and neither produces the false reassurance of a full globe attached to a dying plant. That asymmetry — fails-loud versus fails-silent — matters more than any spec sheet when you are away from home.

Neither device, incidentally, protects a plant that is already in trouble. If a pot has no drainage, if the mix is compacted and airless, or if the plant is well past the point covered in how long houseplants can go without water, adding a reservoir will not fix it and may accelerate the decline.

06 · CAPACITY AND COST

Realistic duration, cost and longevity

Duration is arithmetic, not marketing. A plant's water use is set by leaf area, light, temperature and humidity, and in a warm bright room a medium foliage plant in a six-inch pot can transpire well over 100 ml a day, while the same plant in a cool dim corner in winter might use a fifth of that^[9][13]. Divide reservoir volume by that daily rate and you have your honest ceiling. A roughly 500 ml reservoir covers about five days of heavy use or two to three weeks of light use, and no mechanism changes that division — what a demand-driven system changes is that it does not waste volume on days the plant did not need it.

The practical implication is to measure rather than guess. Water the plant fully, weigh the pot, weigh it again a week later, and the difference in grams is very close to the millilitres used per week. Then size the reservoir, or the number of spikes, to that figure with margin. This is the single most useful preparation step before a trip, and it is covered in more detail in how to water plants while you are away.

On cost and longevity, the comparison is closer than the mechanism debate suggests. Both are inexpensive. A glass globe's life expectancy is limited mainly by breakage, which is a matter of when rather than whether. Terracotta does not break in normal handling, but fired clay is brittle under freeze–thaw: water inside the pores expands when it freezes and can crack the body, so any spike used outdoors must be emptied and brought in before frost. Indoors, a terracotta spike is effectively permanent, needing only an occasional descale.

Pot size and mix decide which device even makes sense. In small pots up to about 6–8 inches, a single spike sits close enough to the whole rootball to matter. In larger pots the wetted zone around one spike is a small fraction of the volume, so you need two or more, or a larger buried reservoir. This is why the olla format — a bigger buried clay vessel — exists at all: it scales the same physics to bigger containers and outdoor beds. For a broader look at reservoir systems in general, including their drawbacks, see whether self-watering pots are good for plants.

07 · THE HONEST LIMITS

What terracotta does not do

A comparison that only lists the winner's strengths is an advertisement. Terracotta has four real limitations and you should plan around them.

Capacity still bounds duration. Self-regulation improves how efficiently a volume of water is spent; it does not create water. A spike holding roughly half a litre cannot cover a thirsty plant for a month, and any claim that it can is arithmetic that does not survive contact with a warm July.

It needs firm soil contact to work at all. The water potential gradient is continuous only if clay and soil are physically touching. An air gap breaks the connection, and outflow stops almost entirely. Insert the spike into moist, settled mix, press the soil gently around it, and check after a day or two that the level is actually dropping. If the mix shrinks away from the spike as it dries, re-firm it.

Hard water gradually reduces porosity. Dissolved calcium and magnesium carbonates precipitate inside the pore network as water passes through, and over months in a hard-water area the flow rate drops noticeably^[8]. This is fully reversible: soak the spike in a solution of roughly one part white vinegar to three or four parts water for a few hours, rinse thoroughly, and the pores clear. Doing this two or three times a year keeps performance flat.

It does nothing useful in soil that is already saturated. With no potential gradient there is no flow, so a spike in a waterlogged pot is inert — which is a safety feature, but it also means the device cannot rescue a plant from a badly chosen pot or mix. Drainage holes, a free-draining medium and adequate light still do the heavy lifting. And clay can crack if water freezes inside it, so bring outdoor spikes in for winter.

08 · THE VERDICT

Which to use, by use case

Use a watering globe if you want a decorative object that also buys a plant a few extra days, you are away for a long weekend rather than a fortnight, and you can accept that the actual delivery is unpredictable. Test it at home first — fill it, insert it, and watch the level for three days — so you learn how your particular mix behaves before you rely on it.

Use terracotta spikes for anything you actually need to work: a one to two week absence, a plant in a room you rarely enter, or the plants you simply keep forgetting — the same category discussed in the best houseplants for people who forget to water. Match the number of spikes to pot size, insert into moist settled mix, and check the drop rate over the first two days.

Use an olla once the pot gets large, or you move outdoors. A single spike wets only the soil immediately around it, so in a floor-standing planter, a raised bed or a container garden you want the larger buried clay reservoir instead. Same physics, more volume, and it is the format the agricultural research on buried clay pot irrigation is actually based on^[11].

And the honest caveat that applies to all three: no reservoir device substitutes for the right pot, the right mix and enough light. A plant in a drainage-free pot of dense compost in a dark corner will decline whatever you bury next to it. Get the fundamentals right, and then use a demand-driven reservoir to smooth out the gaps in your attention.