Regime Shifts DataBase

Large persistent changes in ecosystem services

Reinforcing feedback loops:

R1. Soil temperature reinforcing feedback (local, well established):

• Regime 1: A low deciduous:coniferous abundance ratio i.e. black spruce dominated forest (Picea mariana),  prefers cold environments and thick layers of organic litteron top of the mineral soil. The deeper the soil organic layer is, the more insulated the mineral soil becomes, which in a cold environment such as interior Alaska means that the soil will stay cold despite warm summer temperatures. This in turn will prevent permafrost from melting. Permafrost limits the percolation of rainwater through the soil, increasing the soil moisture. This reinforces a coniferous dominated forest (Johnstone et al., 2010).
• Regime 2: The deciduous dominated forest tends to have a shallow soil organic layer, which makes the soil more susceptible to temperature shifts. In average, this means that the soil is warmer, which leads to permafrost melting and the soil becoming drier. The warmer, drier conditions favour deciduous tree regeneration (Johnstone et al., 2010).

Two important delays are present in R1 feedback, both related to the time required for the vegetation to reorganize and have a significant effect on their surrounding environment. It takes time for the trees to create their preferred soil conditions such as accumulation of required depth of the soil organic layer and soil moisture. At the same time, there is a delay between the soil characteristics present and an observable change in the deciduous:coniferous abundance ratio, since the trees need several decades to be established and exert their dominance.

R2. Decomposition reinforcing feedback (local, well established):

• Regime 1: The coniferous dominated forest promotes a deep soil organic layer which functions as an insulator limiting the increase of soil temperature during the summer months. The cold soil temperatures leads to low rates of decomposition of organic litter, that in turn will further deepen the soil organic layer (Johnstone et al., 2010).
• Regime 2: The deciduous dominated forest promotes a shallow soil organic layer which increases the soil’s sensitivity to atmospheric temperature. On average, this makes the soil warmer than in the coniferous dominated regime. The warm soil stimulates fast decomposition rates. Combined with the high organic litter production, this creates high nutrient turnover and a shallow soil organic layer, which are conditions that deciduous trees thrive in (Johnstone et al., 2010).

R3. Regime shift reinforcing feedback (local, well established):

• Regime 1: Coniferous trees like black spruce, have developed regeneration mechanisms that have made them competitively superior when a deep soil organic layer is present. At the same time, the black spruce have high production of organic litter to the forest floor, reinforcing its own regeneration (Johnstone et al., 2010 ; Hollingsworth et al., 2013).
• Regime 2: On the other hand, deciduous trees species are competitively superior when a shallow soil organic layer is present. A shallow soil organic layer is maintained by these species through the creation of soil conditions (i.e. temperature, moisture and nutrient turnover) which allows them to remain competitively superior (Johnstone et al., 2010 ; Hollingsworth et al., 2013).

The previous delay mentioned in R1 between the deciduous:coniferous abundance ratio and the depth of the soil organic layer, is also apparent in the opposite direction. It takes time for the forest community composition to adjust to the environmental conditions such as changes to the depth of the soil organic layer.

R4. Albedo reinforcing feedback (regional, speculative):

• Regime 1:Lower atmospheric temperatures shorten the growth season of the forest, maintaining the high albedo snow longer in the spring. A high albedo lowers the atmospheric temperature by reflecting more solar radiation.
• Regime 2: However, the previous scenario is highly unlikely. Instead, due to climate warming, the opposite reinforcing feedback is happening (Mann et al., 2012).

Balancing feedback loops:

The mechanisms driving the balancing feedback loops are the same for both regimes.

B1. Coniferous forest balancing feedback (regional, speculative):

Coniferous forests develop under moist and cold conditions, leading to low frequency of fires However, when fires occur they tend to be more severe due to the accumulated highly flammable biomass. Severe fires burn the soil organic layer, decreasing its depth, and opening the door for deciduous trees to dominate (i.e. a balancing loop). However, if the severity is not strong enough, the remaining depth of the soil organic layer will still be enough for the coniferous forest to reinforce itself (i.e. regime reinforcing loop, see R3) (Johnstone et al., 2010 ; Kelly et al., 2012).

B2. Forest Heterogeneity balancing feedback(regional, speculative):

Severe fires contribute to a more heterogeneous forest which limits the spread and severity of subsequent fires.This balancing loop has the potential to limit the speed at which the regime shift takes place (Johnstone et al., 2010 ; Kelly et al., 2012).

B3. Fuel quantity balancing feedback 1 (local, speculative):

An increase in fire severity will burn the soil organic layer, decreasing its depth, which is translated into a decrease in the availability of fuel for subsequent fires  (Johnstone et al., 2010 ; Kelly et al., 2012).

B4. Fuel quantity balancing feedback 2 (local, speculative):

See B3  (Johnstone et al., 2010 ; Kelly et al., 2012).

B5. Insect outbreak balancing feedback (local/regional, contested):

Fire severity has a direct impact on insect outbreaks by acting as a pest-control and influencing insect diversity. Insects affect forest productivity and fuel quantity by for example killing the trees. The fuel quantity has a direct impact on how severe fires can be (McCullough et al., 1998).

B6. Growth season balancing feedback(regional, speculative):

A coniferous dominated forest has a lower albedo, which decreases the atmospheric temperature. With colder temperatures, the growth season decreases, which decreases the   storm season. Since lightning is the most important cause of extensive wildfire ignition (Kasischke et al., in press), it increases the overall frequency of fires. From this point on, see B1. Deciduous-dominated forest, on the other hand, have a higher albedo, which decreases the atmospheric temperature, which makes the growing season longer, and so on.

B7. Albedo balancing feedback (regional, speculative):

A coniferous dominated forest (i.e. low deciduous:coniferous abundance ratio) has a lower albedo than a deciduous dominated forest (Rupp et al., 2002). A lower albedo will cause higher local atmospheric temperatures (Blok et al., 2011), increasing the soil temperature. This will melt the permafrost and decrease the soil moisture as water percolates down the soil (Wolken et al., 2011). A drier soil favours the thrive of deciduous forests. The strength of this balancing feedback depends on the regional extent of vegetation change affecting the albedo  (Johnstone et al., 2010; Mann et al., 2012).

Important shocks that contribute to the regime shift:

• Fire (local and regional, well established):Fire constitutes a natural, dramatic disturbance shaping the boreal forest landscape by abruptly influencing the vegetation composition and structure. If the fire is severe enough, the whole soil organic layer will be consumed, which will make the post-fire conditions favour regeneration of deciduous species, since they have been shown to be competitively superior for sprouting in mineral soil (Hollingsworth et al., 2013). The persistence of the new, deciduous dominated regime will depend on the climatic conditions - warmer drier conditions benefit deciduous trees, while colder, moist conditions could gradually shift the system back to a  coniferous dominated regime (Johnstone et al., 2010)
• Insect outbreaks (regional, speculative): Insect outbreaks have the potential to be a major shock and greatly affect forest structure in boreal forests (McCullough et al., 1998). However, in present conditions in Interior Alaska, the insect infestations only constitute localized disturbances (Rupp et al., 2002), which is why it has not been included as a disturbance in the present system analysis. In a future warmer climate, there is a risk that insect might become more common and therefore insect outbreaks have the potential to become a more dominant disturbance shock in the future.

The main external direct drivers that contribute to the shift include:

• Climate warming (regional, well established): The boreal forest is expected to be one of the biomes most sensitive to climate change (Lynch et al., 2004), and the link between climate warming and increase in fire is well established (Kelly et al., 2012). The increase in temperature is expected to increase soil temperature and lengthen the growing season (Mann et al., 2012). To date, observed temperature increase is believed to be the main driver behind the increase of fire frequency and severity in Interior Alaska. As a consequence of projected decreases in relative humidity, the occurrence and extent of fires will potentially increase by up to 60 percent by 2039 (McCoy and Burn, 2005). An increase in fires will greatly affect the composition and structure of the boreal forest, and the projected drier conditions will increase the probability of local-scale tree stands shifting from regime 1 to regime 2 after a severe fire, which in turn will lead to a gradual regional shift (Mann et al., 2012).
• Fire management and forestry practices (local, contested): Due to the large areal extent of Interior Alaska, the fires are at present not managed in the entire region. This means that fire management as such is not a major driver in all of Interior Alaska. However, where the risk for human life and property is high, an area which constitutes 17 percent of the total Interior Alaska, active fire suppression is practiced. There, fire management is decreasing the severity of wildfires in the forest, possibly counteracting the regime shift (Chapin et al., 2008). This partial fire suppression is possibly affecting the structure of the forest, slightly increasing its heterogeneity.

The main external indirect drivers that contribute to the shift include:

• Population increase (regional, speculative): The population in Alaska is projected to increase by 25 percent by 2030 relative to 2006  (Huntsinger et al., 2007). This population increase might potentially increase the area where active fire suppression is deemed necessary, which means that the impact of fire management would increase.
• Socio-economic global development and climate change-regulating policies (global, well-established): The global socio-economic and population development has a major effect on the extent of climate change, due to its importance for e.g. greenhouse gas emissions through land use and burning of fossil fuels. The nature of the international action taken, e.g. the kind of climate change-regulating policies instituted, will also have a major effect on the future global climate (IPCC, 2011).

Slow internal system changes that contribute to the regime shift include:

• Change in soil characteristics (local, well-established): The characteristics of the soil, mainly the depth of the soil organic layer and soil moisture, affect the species composition of the forest, e.g. if it is dominated by coniferous or deciduous trees (see feedback loop descriptions). Shocks, such as fire, and drivers, such as climate, can change these characteristics and thus cause internal system changes.

• Threshold 1: Depth of the soil organic layer: The depth of the soil organic layer is the most important soil characteristic defining the dominant vegetation in the boreal forest. If the soil organic layer reaches a certain depth (or lack of depth), the ecosystem will shift from one tree species type dominance to the other (Johnstone et al., 2010).
• Threshold 2: Atmospheric temperature: The atmospheric temperature is the most important climatic factor defining the direction of the regime shift. Since the forests have been shown to create self-regenerative conditions, the stickiness of the regimes mean that the gradual change in temperature will probably not lead to a gradual change in tree species dominance. However, when a sufficiently large change in temperature has occurred, a threshold will probably be crossed, leading to a switch from one set of self-reinforcing feedbacks to another (Johnstone et al., 2010; Beck et al., 2011). 

• Atmospheric temperature (global, certain):

Since climate warming is the strongest driver of regime 2, a decrease in temperature by reducing input of warming greenhouse gases in the atmosphere would have the largest impact for restoring regime 1 (Chapin et al., 2008)  . Lower temperature would imply a decrease in, soil moisture, growing season, humidity and insects outbreaks which in turn are conditions that favours the coniferous forest in regime 1 (see feedbacks). However, climate change is a global driver which is due to major global causes, hence not realistic to manage at a regional scale.

• Fire management and forestry practices (local, certain):

Another leverage point originates from fire management  practices, which is the most direct effect humans can have on fire regime dynamics. Through changes in fire suppression management, the fire severity and forest heterogeneity can be influenced . A decrease in fire severity implies deeper soil organic layer which favours the coniferous trees (see B1). Forestry practices could enhance the heterogeneity of the forest which limits the spread and severity of the fires (see B2 feedback.) (Johnstone et al., 2010 ; Kelly et al., 2012). Allowing distant fires (i.e. fires not too close to communities) to burn under controlled conditions conditions could have a cooling effect at regional scale in interior Alaska due to higher albedo of post-fire deciduous species compared to previous coniferous forest (Chapin et al.,2008). 

It is unclear if the observed recent increase in fire frequency and severity will be a permanent change in the fire regime in Interior Alaska, or if it is part of the period of regime shift. It is possible that the high accumulation of organic litter in the coniferous dominated forest, combined with the new, warmer conditions, is promoting more frequent and more severe fires. However, once the soil organic layer has burned and the forest has properly shifted from a coniferous to a deciduous dominated system, the fires will possibly decrease in frequency again due to the smaller amounts of organic litter on the forest floor. Mature deciduous forests tend to have lower fire frequency than coniferous forests (Johnstone et al., 2010). However, different models predict that due to the warmer, drier conditions and longer summer season, fires might keep on being as frequent as now, or even increase until 2020, when the fire regime might stabilize (Mann et al., 2012).

Since the main driver behind the regime shift is climate warming, the inherent weaknesses of climate models translate into the predictions of how vegetation cover might change. Especially precipitation patterns tend to be hard to model with high certainty (McCoy and Burn, 2005). Since the change in relative humidity is of great importance for the shift from coniferous to deciduous trees as well as for the fire regime in Interior Alaska, the projections presented in this report share the same uncertainties as the climate models that have been used in our references.