How does my critter communicate across the liquid/air barrier?

I've been working on this creature some more and have hit another snag.

If my critters communicate through the use of pheromones and chemical trails, how do they communicate the message that it is safe to follow across the liquid/air barrier?

---

In my question on 'why my little critters were afraid of water', the answer by @AndyD273 helped me clarify that they can,

...communicate through chemicals in the air

When they are inside a gooey host or victim,

Any chemicals that they release to communicate with each other and keep their hive mind like link going are going to be contained in their victims body.

And the effect of large bodies of fresh or sea water,

... would dilute and wash away all of the communication chemicals cutting them off from the rest ...of the swarm.

@Josh King's answer correctly identified why they are afraid of water bodies and moving water

Over time they have developed an innate fear of all large concentrations of water, but especially indications of fast moving water. A misting of water, especially a directional spray, often corresponds to a large moving body of water, like a waterfall or river rapids. This signals one of their primal fears, causing them to contract as a group and retreat from the source.

And as @Renan pointed out that due to osmosis their bodies, or possibly just the ovisacs they carry,

...can't regulate the salt in their bodies like we do. The range of salinity they can tolerate is narrower.

---

Information relevant to new question

I decided that Mistraille (my critters), are in constant chemical contact with one another. When exposed to water, any water, the signal back to the swarm will be lost. The entire swarm retreats from perceived danger.

Water vs Victim.

(1) When one or more of them are 'accidently' immersed in liquid that is not a living organism, there is nothing to trigger their transformation from 'travel' mode to 'attack' mode. They end up stuck in the fresh/salt water with no way back out. They end up dying grotesque deaths by either exploding in fresh water, or withering into a desiccated husk in salt water (thanks @Renan). No signal is ever returned to the swarm outside the water. The swarm correctly perceives this as a danger (an evolutionary trait) and retreats.

I figured they are attracted by the chemical 'emission' of open wounds in animals and plants. If they couldn't get any scarier, I had to go make them SMELL BLOOD!

(2) When one or more of them are immersed in the liquid of an animal or plant, the biological matter provides the ability to transform but again, the 'viscous' nature of water compared to air means that they can't leave their 'watery' host. Under my current design, no signal can be returned to the airborne swarm and they incorrectly perceive this as a danger and retreat.

Here in lies the conundrum. How do they communicate back to the airborne swarm that there is 'no' danger? That the swarm shouldn't retreat but should in fact advance!

'Food this way!!!'

The swarm would have been attracted by the smell of 'blood', so they may not retreat as fast or very far from the direction where the contact was lost. But I still need a way to figure out how they realise fairly quickly (a few minutes max) that the loss of contact is not an immediate threat.

I don't just want them divebombing because of 'blood' + 'loss of contact' = 'go go go!'

Why? Because there could quite possibly be a non-viable water source right next to the location where the smell of blood is coming from. While they think they are entering a safe haven, the loss of contact was actually from an accidental immersion into a neighbouring pool of water. They are instead committing self-inflicted genocide!

---

Further Mistraille characteristics, that might help influence answers

I have worked out that the infected organisms immune system can fight Mistraille off in this first period of infection. Animal white blood cells and plants cells are very 'similar' size-scale wise. I'm thinking that the 'travel' Mistraille would only be maybe a couple hundred µm. The ovisacs even smaller (each holding say 10-50 dormant eggs each). The various white blood cells are max 70-80µm. Plant cells are about max 100µm. This both works against the Mistraille and in their favour.

The travel Mistraille inside a victim are capable of some consumption of biological material. They use this, not to eat, but to coat the dormant fertilised ovisacs, with the plant's or animal's own cell matter. This protects the eggs from the victim's immune response giving them time to incubate and hatch. The larger travel Mistraille 'activiate' the 'dormant' ovisacs in some manner before being overwhelmed by the immune system and destroyed, leaving the eggs behind. MISTRAILLE ARE NOT INVULNERABLE!

In animals, the immune response to the 'initial' invasion is huge, and the white blood cells are depleted, at least locally. In humans it takes about 5 days for new white blood cells to be replaced. So either the travel Mistraille can last long enough or maintain enough of a threatening presence by shear numbers that there is a period of time when the immune response no longer works. When the travel Mistraille are 'destroyed' and the remaining immune system 'believes' it has won as it does not perceive the eggs as foreign bodies (seeing as they are covered in masticated cells of the victims own body). The eggs hatch a short time later and the newly hatched attack larvae are then free to develop for several days in relative peace.

In plants, there are no travelling immune cells but rather signals are sent once an invasion is detected warning other cells. These neighbouring cells then either commit suicide or shut down their cell receptors. This is no problem to Mistraille. Dead biological matter is still viable biological matter. Mistraille have a much easier time with plants than with animals. Some plants kill off entire limbs to try stop infections. Some create 'solid' barriers to try limit the spread of infections. Again this won't help unless the plant actual severs the infected section. This is because the travel Mistraille and ovisacs will have practically unimpeded time to transform from travel mode to attack mode. They will then be able to just burrow through any desiccated flesh or 'solid' barriers.

There is more, but by this stage there should be enough Mistraille infecting the victim that communication with the previous remaining airborne swarm isn't necessary.

I need to figure out a way that these tiny critters can send a message back out to the airborne swarm IMMEDIATELY after entering viable hosts!

HA, could it be that the breaking or dissolving of the ovisac, indicating successful activation of the previously 'dormant' eggs, releases a potent chemical signature that can be detected out in the air? No, that would take more than a couple minutes.

No, wait! The Pus.