DivMost

Objective:

The DivMoSt project (2024-2025) examined meadow orchards across Austria. The aims were to, (1) record the biodiversity of indicator animal groups (wild bees, butterflies, grasshoppers, birds and bats) in orchard meadows and to (2) develop a geodata-based method that enables the nationwide monitoring of orchard meadows.

Background:

Meadow orchards are extensively farmed areas, such as grassland with orchard trees (i.e. large-crowned fruit trees that stand freely without permanent support). They are considered high-quality, semi-open habitats that are very important habitats for many animal and plant species. The number of orchard trees in Austria was estimated at 35 million in 1930 and decreased to around 4.2 million by 2020. The current number of orchard trees is based on estimates, especially for non-agricultural areas. 

Method & results:

To automatically locate meadow orchard, 23 test areas (each ≥ 120 km²) across Austria were examined. Meadow orchards were mapped in the field by experts in two landscape sections per test area using a uniform methodology. The condition was documented using quality parameters and comparative analyses were carried out to assess the current state. A total of 1,170 orchard polygons with a total area of 231.9 ha were recorded. Further, around 5,900 orchard trees were located as individual trees (points) and identified to species level. Most of the mapped orchards were located on grassland (79%) or in domestic gardens (17%), the latter being at least regionally, of great importance for orchard cultivation. In 71% of the test areas, the age of the stands was between 20 and 60 years, and in 20% trees were predominantly over 60 years old. Overall, this shows insufficient replantings in Austrian meadow orchard cultivation to maintain at least the current state. When promoting maintenance and replanting, the preservation and creation of structural diversity need to be considered, even if fruit yield is the dominant purpose of the meadow orchard.

The orchard mapping results provided the basis to development the methods for the automated detection of meadow orchard stands based on digital geodata. The digitized field data were combined with a variety of geodata, such as elevation models, usage data and multi-seasonal satellite image time series from 2023 and 2024, as well as a classification procedure from the field of machine learning. The final model achieved satisfactory accuracy in detecting scattered fruit trees. Verification with independent geodata showed the potential of applying such methods and the extent to which scattered fruit trees can be distinguished from other woody plants. 

To survey indicator animal species, a 625 m × 625 m reference area was defined within each landscape section, in accordance with Austria's existing biodiversity monitoring programmes (ÖBM, BINATS, MOWI). The animal groups were surveyed based on species richness, abundance and/or activity. Based on the biodiversity data, characteristic species were defined for Austrian meadow orchards. 

A total of 458 animal species (321 insect, 23 bat and 114 bird species) and 230 entomophilous flowering plant species were recorded in just one survey year. For insects in detail, 31 grasshopper species (1,626 individuals), 65 butterfly species (2,251 individuals) and 225 bee species (5,196 individuals) were documented – 50% of the bee individuals were honeybees (2,719). In relation to the Austrian fauna, 24% of Austrian grasshopper species, 30% of butterfly species and 32% of bee species were found in the meadow orchards. Concerning vertebrates about 50% of the breeding bird species in Austria and 74% of the Austrian bat species were documented in the investigated meadow orchards. These proportions already indicate the importance of meadow orchards as habitats for vertebrates and invertebrates in agro-ecosystems.

Many more exciting results are presented here in the final report.