Ecologists emphasize the natural, rejuvenating effects of fire in Fynbos when lay persons lament the destruction wrought on landscapes. Fire is part of the ecosystem!

Wildlife is for the most part well-adapted to or even dependent on fire, they say. It is true that there are a myriad of fascinating adaptations and strategies for dealing with fire in Fynbos and other fire-prone systems – things that have evolved over millions of years.

Isn’t it incredible how the genetic result of millions of years of adaptation history has led to Protea and Leucadendron seeds blowing around a scorched landscape attached to cute feathery parachutes? We know this game, they say, let it land where it may! It is obvious that their fresh start after fire is far from a blank slate.

These are some of the thoughts that go through a technician’s mind as they walk a familiar route to inspect what is left of their field equipment, hoping to see a little flash of red light that indicates their data logger is still alive and they have a chance to download the valuable data.

Wednesday 24 February 2021 – 10:24am:

On a visit to inspect a weir-cleaning progress in Jonkershoek, Fynbos Node Technician Abri de Buys reports this image to the Fynbos Node WhatsApp group.

Concerned about our instrument network, the Fynbos Node team starts gathering fire status information from as many sources as possible and Node Manager Dr Jasper Slingsby gets in touch with Cape Nature and other local authorities. We begin considering the possibility of evacuating equipment, until we are informed that there is a high likelihood that the fire will enter Jonkershoek. At this point, on the night of the 24^th, an evacuation plan is made for 25 February, targeting high-value instruments.

At 7 am on the morning of the 25^th, the team of five people assemble at the Node offices and proceed to Jonkershoek armed with two vehicles, storage crates, backpacks, safety gear for working at heights and multiple sets of the necessary tools. With one eye on the smoke rising over the mountains to the north of the valley, we manage to dismantle and retrieve four weather stations and most of our eddy covariance system over the course of the day, making sure to keep track of which sensors belong to which stations.

We return home anxious and uncertain whether evacuation was the right decision, given the data gaps that this creates, and sharing jokes about the possibility of having to do this a few times this fire season.

Thursday 25 February 2021 – midnight:

After recording a temperature of 52 degrees Celsius and unusually high outgoing radiation (heat/energy from below the net radiometer), Dwarsberg weather station goes offline.

Daily fluctuations in evapotranspiration from the catchment are responsible for the sinusoidal pattern before the fire as the vegetation “breathes out” (transpires) water during daily carbon dioxide uptake before the fire. After the fire, most of the plant parts that transpire have been removed or damaged, so the typical daily drop in water level associated with a vegetation covered catchment is much less pronounced. In the case of this catchment (Lambrechtbos B), the vegetation was mostly pine plantations.

Maintaining resilience in long-term monitoring infrastructure

From a technician and monitoring platform manager’s point of view, the clock starts ticking when your site burns. Until each monitoring station has been assessed and/or repaired, data are being lost or at risk of being lost.

Our ability to quickly and effectively respond to a crisis such as a site burning down, or some other large-scale disturbance, depends on the stock of equipment available for immediate deployment. It also depends on the necessary labour power to assess damage and deploy replacement equipment quickly.

Having a full set of redundant equipment ready at the Node for postfire deployment can mean millions of rand sitting idle in storage. While this would allow for the most effective reduction of data gaps, it is arguably not the wisest approach. This approach requires difficult calculations like weighing up the cost of retaining spare instruments against the value of the data.

An alternative approach is ensuring there is enough redundancy in the organisation as a whole to replace instruments at any particular site on short notice. This allows for flexibility and is arguably a better value for money “insurance policy” against data loss.

A potential shortcoming of this approach is that instruments need to be shipped to where they are needed, which can be risky. A problem during transfer can lead to extended data gaps, especially with high-value instruments, and this is accompanied by a tricky administrative workload – keeping track of assets as they move around.

It is also tempting to call on the backup instruments even when there is no crisis, to avoid cumbersome, time-consuming procurement procedures, or the potential that multiple Nodes may reach for the same resource at the same time.

If anything is clear from our fire experience in 2015, responding to a crisis by procuring new equipment or having equipment repaired is NOT the way to minimise data gaps. Where we relied on quick shipments from our colleagues across the country, we rapidly managed to deploy new instruments in 2015. Where we had to procure replacements or have instruments repaired, data gaps extended for months.

A valuable and financially prudent tool for emergency response would be contracts with instrument suppliers, in place permanently or with minimal interruption, so that we have access to new equipment and spare parts that can be shipped anywhere in the country within days after a damage assessment.

We are realising the value of longer-term contracts with suppliers as a way of reducing duplication of effort across the organisation and time spent in procurement processes. The reduction in the administrative workload can be substantial and this approach can limit investments in equipment sitting idle.

Considerations or risk areas with this approach may be the contracted supplier’s ability to access instruments from manufacturers overseas or international shipping in cases where they are a local branch of a foreign company, and/or fluctuations in prices.

As I write this, we managed to ensure the entire Jonkershoek historic network (rain gauges and weirs) was operational within a week after the February 2021 fire. We did get off lightly this time in terms of damage, which made the task easier, but it would not have been possible without having new equipment at hand.

Having got off somewhat lightly also meant it was possible for one technician to restore normal operations (or close to it) in this network. Reinstalling all the evacuated equipment is a different matter entirely. Having the equipment at hand is only one part of the equation that determines the length of data gaps.

Our decision to evacuate high-value instruments turned out to be the right one. Not only did we save >R1.5 million of taxpayer investment (today’s replacement cost), but the instruments can be redeployed immediately after fire or as soon as site access is gained, and team members are available.

Where evacuation can be done safely, it is potentially a solution for landscapes where fire prevention is not possible, as is often the case where we do not own the sites, the vegetation type does not allow it, or we do not have the ability or resources to manage the risk involved in burning controlled fire breaks. Fire suppression is also arguably not always desirable when trying to understand how fire-adapted systems respond to environmental change, even if it were possible in the long term.

Like the environment we work in, we need to adapt to fire and other disturbances so that a postfire fresh start is not on a blank slate. We can learn from case studies and previous experiences and align our approach and processes with our needs.

This should allow us to save what we can and replace what we need to, and to do so quickly to maximise the data we produce and the value we get out of our equipment.