This section describes the work carried out under sub-task T.3.1.3: Climate risk and vulnerability assessment. It provides an overview of the methodology adopted, which is grounded in the analytical framework of impact chains. Impact chains are a tool applicable in vulnerability and risk analyses related to climate change, developed by Adelphi and Eurac on behalf of the German Corporation for International Cooperation in 2014 and subsequently updated based on the definitions and conceptual framework of the IPCC AR5 (Adelphi-Eurac, 2014; GIZ-Eurac, 2017).

To develop the impact chain we begin by identifying a specific RISK and then work backwards through its components to identify intermediate impacts, or potential damages, and explore the various factors that contribute to its severity:

• HAZARD: the triggering climatic event and its direct physical impacts.
• EXPOSURE: the elements of the socio-ecological system that suffer damage due to the direct physical impacts.
• VULNERABILITY: the combination of characteristics that increase susceptibility or influence the capacity to respond and adapt.

Within the BEEadapt project, this logical framework can be simplified with the following image that shows all the components used to create the risk map.

The risk considered by BEEadapt consists in the decline of insect populations, which stems from a scarcity of food sources – primarily the absence or reduction of flowers – and a shortage of suitable habitats for nesting and reproduction. The components of such risk can be modelled as follows:

Hazard

This component includes some of the characteristic phenomena of climate change: rising temperatures and decreasing precipitation (especially prolonged droughts) can cause phenological alterations that affect the availability of nutrients, while extreme weather events can lead to the sudden destruction of habitats.

In order to investigate this component, we chose to use a dataset produced by the WorldClim 2 database (worldclim.org), which performs a monthly spatial interpolation of climate data with global coverage at a spatial resolution of 1 km.

Among the available scenarios, two have been selected: one that describes a sustainable pathway with ambitious climate policies, compatible with the goal of keeping global temperature rise below 2°C (SSP1-2.6), and one with high emissions, representing a future with intensive use of fossil fuels and rapid economic growth (SSP5-8.5).

By analysing four bioclimatic variables, we obtained the final map of climate hazard for central Italy. The variables considered are:

BC4        Seasonality of temperature
BC5        Maximum temperature of the warmest month
BC14     Precipitation of the driest month
BC15     Seasonality of precipitation

Where BC4 and BC15 are attributable to Extreme weather events, ΔBC5 to Temperature increase, and BC14 to Prolonged droughts (Decreased precipitation).

Hazard factors

Exposure

The exposure analysis assesses to what extent the territory is influenced by the consequences of climate change on pollinator activities. As “exposed elements” we considered agricultural activities and populations of different pollinator species.

To create spatially explicit indicators of this vulnerability, we combined statistical data on Italian farms with abundance models for pollinator species (lepidoptera and apoidea).

For the human-related component, we used ISTAT data from the 2020 Agricultural Census, linking them to the geometries of the Land Use Map corresponding to agricultural areas (code 2). For the insect component, we utilized modelled abundance data for pollinator species specific to each project area.

Exposure factors

Vulnerability

Vulnerability, is the risk component which is most influenced by human activities, it refers to the system’s susceptibility to the adverse effects of climate change. It encompasses both the system’s sensitivity to these changes and its capacity to adapt and respond.

To assess vulnerability, we considered the following aspects and combined, by overlay mapping, the related dataset:

• Presence and vigour of vegetation (represented by NDVI index);
• Heterogeneity of the environmental mosaic (represented by RAO index);
• Presence of green infrastructure and ecological corridors (represented by Copernicus Small Woody Features – SWF);
• Land cover suitability for pollinators (represented by a reclassification of Land Cover data)

To account for the seasonal variations in pollinator activity, each input level was assessed at three distinct moments of the year, from the beginning of spring to the end of summer. Low values indicate high vulnerability, while high values represent low vulnerability, suggesting favourable locations for the presence of pollinators.

Vulnerability factors

The Normalized Difference Vegetation Index (NDVI) is an indicator used to monitor the health and vigor of vegetation in a specific area. It is calculated using measurements of light reflected in the visible red (RED) and near-infrared (NIR) bands of the electromagnetic spectrum. The use of this information allows for the attribution of a value to the presence of vegetation in urban and peri-urban areas, in agricultural settings, and in natural and forested environments.

The Rao’s quadratic entropy index (RAO) is useful for assessing the diversity of biological communities in a given geographical area and, consequently, the heterogeneity of the environmental landscape mosaic.

The use of this information allows for the evaluation of the variety of natural and semi-natural contexts in reference to the vegetation index. In agricultural settings, low levels of heterogeneity are associated with intensive crops, which are detrimental to pollinator biodiversity. Conversely, high levels of heterogeneity indicate agricultural practices that support the presence of green infrastructure and, consequently, favour the presence of diverse pollinator species. In natural environments, the same dynamic is observed: high levels of heterogeneity favour the abundant presence of pollinators, while dense and uniform forests do not provide a suitable habitat for their presence.

Urban environments constitute a special case: although intrinsically heterogeneous, they are not suitable for the presence of pollinators, which can influence the results of the RAO index. For this reason, they have been excluded.

The SWF factor indicates the presence of green infrastructure and ecological corridors. It has been included in the calculation of the territorial vulnerability level to stress the importance of green infrastructure already identified using the RAO index. It is one of the cartographic products provided by the Land Monitoring Service of the Copernicus project: the 2018 High Resolution Layer Small Woody Features (SWF), which provides detailed data on landscape characteristics related to small woody elements, such as isolated trees, hedges, rows of trees, and small woods. These infrastructures facilitate the movement of pollinators and their survival by creating green corridors between natural, agricultural, and urban environments.

Land cover suitability for pollinators (CUS) To improve our assessment of vulnerability, we included land use data (CUS) and evaluated its suitability for pollinators. Land use classes were therefore reclassified assigning each a score based on expert opinion from the project’s entomology team.

Risk

The risk of ecosystem service alteration was calculated as a function of its three components – hazard, exposure, and vulnerability – each of which was in turn calculated based on the relevant factors identified and described above. The formula used was:

Risk = Hazard * Exposure * Vulnerability

To allow for analysis at a larger spatial scale than 10 meters, the results were calculated for hexagonal areas of 10 hectares.

The summary table below reports, for each intervention area, the average values of all components of the impact chain, allowing an analysis of which components most influence.