With the advent of firefighting techniques, I have to assume that when a forest fire broke out in the past, it spread uncontrolably. Are there any records of forests that were wiped out entirely, or severely reduced in size, due to a large fire? If so, what effect, if any, would a fire of that size have had on the surrounding enviroments?
I'm going to answer this question with reference to the US mostly, and Canada to some degree because that is where I have the most knowledge. Australia is the other place where massive fires are famous and common, but wildfire exists everywhere that has any concentration of fuel to support it, from rain forests to the arctic boreal forests.
Massive forest fires are a phenomenon that goes back millions of years. However, I think you may have a bit of misunderstanding on the effects fire has on the ecosystem. Every type of forest has a historical "fire interval." These intervals range anywhere from 3-5 years for the ponderosa pine forests in the western US to every 100-200 years for spruce and fir forests at higher elevations. The more frequent fires are less intense, and generally tend to burn litter and small plants in the understory while killing very few of the larger trees. The infrequent fires that occur a century or more apart are commonly referred to as "stand replacement fires." In a human lifetime these huge stand replacement fires may look like a dead forest after being nuked out, but the succession is just starting all over again. It can take a couple hundred years for the forest to get back to the point you may be used to seeing it, but in the interim various species can thrive. Grasses, woody brush species, and trees like aspen all thrive in the years after an intense wildfire. So it's not entirely accurate to describe a forest as being destroyed after a wildfire.
Fire is a natural process that is critical to balancing out forest growth. It just doesn't seem that way to people a lot of the time since it takes so long for forests to return to the way you saw them before the fire. In addition, very severe fires are a more common sight in the past few decades due to the effects of logging, fire suppression, and grazing. The quantity and severity of large fires is unprecedented for the period of time we can measure.
Now, let's take a look at some fires to answer the rest of your questions:
The 1910 "Big Burn" in Idaho and Montana. One of the most notorious fire complexes in US history, and a watershed moment for the fledgling Forest Service. Fire fighting tactics were in their adolescent years at this point. Most of the fire never saw any human influence and it burned around 3 million acres. Pictures after the fire went out are pretty depressing, but visit there today and you'd have a hard time telling there was a fire. Some dead white pines are still standing from 1910, and you can find some burned stumps, but the forest has come back. It's in an earlier stage of succession; lodgepole pine and grand fir now live where western red cedar and white pine lived before, but it isn't a ruined moonscape anymore.
The 1871 Peshtigo Fire in Michigan. The largest and deadliest fire in US history. This fire is good evidence of what a completely uncontrolled fire can do when everything lines up to create extreme fire behavior. Logging and human sprawl caused the fire and contributed to its severity and the area hasn't grown back the same for those reasons.
The 2010 Schultz Fire and 1977 Radio Fire in Arizona. More recent fires, but an example of some of the more long term effects a fire can cause. The Schultz Fire burned a comparatively small 15,000 acres, but was very active and killed most of the trees in its area. The Radio Fire was even smaller, but occurred earlier on the same mountain as the Schultz. In the months and years after the fires much of the soil slid off the side of the mountain the fires burned, destroying homes below and leaving the mountainside stripped of soil. Now, large tracts of land will take thousands or millions of years to recover and rebuild the soil layer needed to support the same vegetation that was present prior to burning.
Finally, it's worth noting that modern firefighting techniques are largely ineffective at stopping fires showing extreme fire behavior. Fires such as the 2015 Motorway Complex in Idaho, most of the 2015 fires Alaska, and the Big Burn in 1910 were stopped by precipitation and/or the fires running out of available fuels. You can see this right now as the city of Fort McMurray burns, even though Canada is throwing every fire resource it can find at the fire.
Hopefully this provides some answers for you, OP. I've had a few beers, so I hope it's not too scatterbrained to make sense. A lot of my "sources" come from my experience as a wildland firefighter and as a student of forestry. However, if anyone wants some better sources, I would be happy to edit them in.
Further reading:
-Firestorm at Peshtigo, by Denise Gess.
-The Big Burn, by Timothy Egan.
-Smokechasing, by Stephen J. Pyne
I discuss this a bit in a previous answer about the Western U.S. (Rocky-Mountain U.S.) prior to European contact.
The below is a modified version of that answer.
The invention of modern firefighting actually has contributed to the frequency of devastating wildfires in certain contexts, especially the Rocky Mountain West. Periodic burning of Ponderosa-Juniper woodlands is a healthy part of the ecosystem that clears out undergrowth. Fire suppression in the U.S. has prevented these smaller-scale, periodic fires, allowing for dead undergrowth to build up. Consequently, fires end up being larger and more devastating because they have so much kindling built up over time.
These periodic fires in the past also increased the spacing and height of trees, particularly in ponderosa forests. Many ponderosa forests today are overcrowded with shorter, stunted trees. This allows for easier spread of crown-to-crown fires (because of the spacing of the trees) and promotes these crown fires because of the short height of the trees. Larger trees are less likely to catch fire, the bark being scorched but leaving the tree largely unharmed.
There is also some evidence that human forestry management helped mitigate forest fires in the past. In New Mexico, management of forests by Pueblo people involved the creation of trails that may have acted as (unintentional) fire breaks. Furthermore, scavenging for firewood from the forest floor may have also helped to reduce the danger posed by fires by helping clear out undergrowth and other kindling in a way similar to smaller, periodic fires. General logging and other wood collecting activities may have also helped buffer against really devastating forest fires.
It seems Pueblo people were quite able to manage the impact of fires on settlements. To my knowledge there isn't a single archaeological example of a Pueblo village that burned down due to wildfires during the occupation of the site.
A very interesting article (Liebmann et al.) was recently published that used a lot of lines of evidence (including archaeology, paleoclimate data, and historical data) to look at population decline in the Jemez valley of New Mexico after European conquest and contact. They conclude that fire regimes changes following European contact and declines in Native populations, further bolstering the idea that indigenous fire management (both intentional and unintentional) helped reduce the risk of massive wildfires.
In summary, the ecosystem of the Rocky Mountain West didn't tend towards huge wildfires prior to the more modern practices of fire suppression. Additionally, the forestry management of Native people (based on the ways they exploited forests for resources) may have further dampened the likelihood of these massive fires.
Source: