Estimating LDN Baseline
(Kust et al. 2018) noted there are three limitations to establishing the LDN baseline as an average value of 15 years prior to the adoption of the SDG Target 15.3, namely: irregular land cover and land productivity dynamics as a result of drastic socio-economic changes and contradictory land reforms since early 1990s; large areas of forests with a longer period of restoration or formation of new plant communities, reaching hundreds of years; long-term ”background” processes of natural evolution of landscapes and adaptation to climate change. Since this publication, a large amount of data has been collected, which confirms these conclusions on the basis of retrospective analysis of space imagery and scientific archives, studies of evolution and variability of landscapes and their components in different geographical locations, including responses to different natural and anthropogenic impacts. For key sites in forest and agricultural areas an LDN risks assessment against natural land potential was carried out, including forecasting vulnerability for existing land management. At a local level, studies were undertaken to compare results using different basic timelines, as well as using “timeless” baseline options, such as background landscapes or crop yield estimations. (Kust et al. 2019) indicated that the irregular abandonment and subsequent return of farmlands, which is reflected in the NDVI changes during this period, cannot be interpreted for definite. The NDVI time-series of agricultural regions in the southern regions of European Russia, with actively recovered abandoned farmlands in recent time, formally indicate “negative change” in land productivity between 2000 and 2015 as compared to natural vegetation growth during the same period. Studies of LDN dynamics of managed forests in the Russian North (RSF 2020) estimate that despite NDVI changes in forest land becoming noticeable within 15-20 years after felling and fires, especially in pine forests, full restoration of forests occurs only after 100-150 years, and restoration of ecosystem and species diversity requires at least 200-300 years or more (Figures 7, 8).
Even for grasslands a period of 15-20 years is not sufficient for baseline considering the full restoration of natural capital and ecosystem services. In steppe natural ecosystems regeneration takes at least 50-60 years (Kust et al, 2019), and restoration of natural plant communities and soil cover in disturbed tundra ecosystems takes 40-100 years (Sizov et al. 2020). Regeneration of once disturbed ecosystems is often inhibited by other constant and prolonged disturbances, such as forest and grassland fires, pest invasion, etc. In addition, natural forests are practically not restored to their original state: forested areas expand at the expense of pine and soft-wooded broadleaf species, while old areas of native forest consisting of more valuable coniferous and hardwood species have declined.
Complex processes of rapid depletion and degradation and subsequent long-term restoration of terrain ecosystems are reflected in the practice of forensic land use expertise, for which “significant harm”, “restoration to baseline” or “full restoration” are among the key concepts, and for which the LDN approach can be useful. (Kutuzova and Kust 2018) highlight that ”reclamation” actions to restore the full scope of ecosystem services and social significance of disturbed ecosystems are almost never sufficient to return an area to the “initial state”, due to its uncertainty. In fact, we are talking about diverse, albeit complementary, goals of restoration and qualities of disturbed and degraded systems by: (i) restoring initial productivity, (ii) restoring ecological functions, (iii) restoring ecological balance, (iv) eliminating negative impact , (v) termination of the negative processes/functions, (vi) elimination of certain negative qualities/properties, (vii) acquiring certain positive qualities/processes/regimes relative to the initial state, (viii) sustainable state. To reduce the uncertainty in expert decisions, authors propose using the concept of ”optimal state” instead of ”initial state” when determining the goal of restoration. This can ensure the implementation of a complex of ecosystem functions that best meet the criteria of environmental safety in site-specific conditions. This proposal fits the concept of “compensation for environmental damage” as stated in Russian legislation, as well as the definition of “restoration/rehabilitation” provided in (IPBES 2021). Using the term “optimal state” for LDN estimating purposes helps avoiding contradictions in setting a baseline for most cases.
As noted above, natural trends often dominate over human impacts, therefore land dynamics should consider their background for a baseline setting. Thus, studies of Holocene evolution of landscapes in southern Russia demonstrate that many LD processes are predetermined by the history of landscape development: anthropogenic impacts most likely trigger degradation processes (salinization, alkalinization, deflation) on sites where their natural analogs took place in the past, despite these old processes have been completed and their features being no longer visible. Human attempts to shift the direction of current degradation processes at a local scale (for example, irrigation against salinization) can temporarily decrease their intensity, but often are not able to reverse them with the technologies available (Andreeva and Kust 2019). Therefore, attempts to achieve LDN in such cases is doomed to failure with insufficient investment. Achieving LDN targets at a regional level occurs more rapidly and at a lower cost, especially where there are “positive” Holocene trends. On the contrary, the “genetic memory” of landscapes on negative “degradation” trends complicates the prospect of successfully achieving LDN.
A baseline can be established by selecting natural ”standards” or their human-supported simulations imitating natural systems while preserving ecosystem functions and natural capital (e.g., no-till technology, adaptive farming, pasture rotation, etc.) This hypothesis was tested in protected areas pilots in the Central Chernozem biosphere reserve and the Samarskaya Luka National Park. It was determined (Figure 3) that the PA territories can be considered as observatories for comparative assessments of LDN indicators for adjacent territories. For the studied PAs themselves, the baseline period of 15 years is sufficient to assess their dynamics in the absence of extreme impacts (Kust et al. 2021).
(Lobkovskiy et al. 2020) drew attention to the fact that, in addition to the temporal and ecosystem factors of baseline setting, the duration of baseline period also depends on the spatial and scale factors. For example, if the land quality of individual farmlands can be restored within several years, then it takes decades to recover a sustainable system of agricultural landscapes. If forest ecosystem restoration takes tens of years, then hundreds of years is needed to restore the integrity of fragmented intact forests. It was also demonstrated that the real land quality may not be adequately reflected, since the starting point of the LDN baseline period may not always coincide with the starting point of recovery. In this case averaging indicators provides an incorrect interpretation of LD processes. For example, the LDN estimate for highly degraded landscapes frequently shows stable or positive dynamics, although the functioning of such landscapes and the performance of important ecosystem functions could be unbalanced (RSF 2020). Depending on the LDN targets for different land use types it was proposed to consider a number of baseline states, contained in Russian regulatory documents: “reference state” as a zero level of loss of natural and economic land value, “initial state” as an undisturbed and non-degraded analogue before the start of impact, “basic state” reflecting the start of monitoring, “background (for local and regional purposes)” as a state formed within a complex of natural or human factors typical for given territory.
(RSF 2020) report concluded that in most cases the approaches to establishing the LDN baseline for practical application in Russia at regional and local levels should differ from the 15-year period recommended for global assessment. The selection of baseline parameters depends on the LDN monitoring objectives. For example, for forest lands, these parameters should consider the dynamics of forests assessed on the basis of typology of time series and changes reflecting typical impacts (felling, fires, etc.). Considering that the peak of the negative dynamics of the LDN indicators falls on 0-3 years after clearcutting, and the sustainable recovery of forest stand and biodiversity begins 10-15 years after felling, the periods for establishing LDN baseline indicators are estimated as follows: for carbon flux 15 years, for productivity and biomass stock - up to 100 years, for species diversity and ecosystem >200 years (Figure 6). For tundra vegetation impacted by wildfires in North-Western Siberia the minimum period for baseline is determined by a 60-year (Sizov et al. 2020)cycle.
When developing the topic of baseline establishment, both parameters for reference points and required time interval are proposed (Lobkovskiy et al. 2022) to distinct the groups of dynamic and target indicators. Dynamic indicators are useful to define baseline periods for different land use models (for example, 10-15 years or others). Target indicators (or State/Quality indicators) are recommended to determine land quality using the principle of ”maximizing” parameters, to demonstrate a sustainable/optimal state in terms of the closeness to (attainability) the “best” state (e.g., achieving a certain soil quality, yield potential of crops, allowable forest cut, sustainable functioning of PA ecosystems, etc.). The principle of “minimizing” parameters is also applicable, for example to demonstrate the rate of desertification as “remoteness” from the state of natural deserts or badlands. “Standardization” approach using national environmental norms and standards also can be applied, e.g., for soil pollution or fertility, state of pastures, forest lands, etc.