The Offset Design Process
38
BBOP Pilot Project Case Study – Ambatovy Project
The Ambatovy forest habitat loss scores were determined by merging data from the three existing habitat
types, since the selected forest attributes did not exhibit a statistical difference and faunal movement from
intensive fauna surveys in the pre-clearing perimeters showed the same occupancy pattern for the three
habitat types. Knowledge of the condition class is particularly important, since it reflects the biodiversity loss at
stake, thus highlighting the application of multipliers in degraded habitats that are subsequently subject to
negative impacts. The importance of multipliers was highlighted, for example, when considering degraded
azonal forests which constitute 44.6% of the azonal habitat loss, but equivalent to only 29.3 habitat hectares
of the total 620 habitat hectares score for the azonal habitat. High impact areas corresponded to 100% forest
clearance with earthworks, medium impacts corresponded to the potential edge effects on the forest
environmental buffer (50 m for linear features and 100 m for polygons), while low impacts did not apply to the
impact areas. The habitat hectares score provided above is for the ‘without post-impact’ mitigation scenario.
The habitat hectare gain for the proposed offset site has not yet been calculated, as detailed forest structure and
quantitative species attribute data are still being acquired, with field surveys for flora / forest structure and aquatics
planned in July 2009 and terrestrial fauna in November 2009. As the proposed Ankerana site is considered to be
an ‘
IN-KIND
’ offset (relative to the impact site), the same benchmark will be used to calculate gains.
Benchmark
Based on the BBOP definition of a benchmark, the following criteria were used to define and identify a
candidate site:
Forest habitats:
– Minimal critical size: in the Ambatovy forests, a benchmark is required that captures the progressive
structural changes between the existing three vegetation types, and the faunal movement between
them throughout the seasons. A contiguous area of quasi pristine ‘azonal’, ‘transitional’ and ‘zonal’
forests, each of which must be a minimum of 40 ha is proposed. This area of 120 ha includes the core
area of original habitat with a 100 m wide buffer around it and appears to be the minimum required as a
home range for lemurs and to represent an adequate assemblage of vegetation.
–
CONNECTIVITY
: the contiguous forest area (minimum size – 120 ha) must be connected to other forest
habitats.
– Human disturbance: there must be no sign that the site has ever been cleared by humans (in both core
and buffer areas). However other evidence including tree stumps, historical records, soil charcoal,
archaeological remains and signs of selective logging (defined as less than 12.5% of crown cover loss)
over the last 20 years do not exclude a site from consideration as a benchmark (such evidence is
widespread and unavoidable in the region). Any small degraded areas within the larger benchmark area
are mapped and excluded from the benchmark calculations (and surface area).
– Natural disturbances: a site that has experienced a natural fire in the last 20 years, at a level of 10% of
its surface area (in both its core and buffer area) is excluded. Also, no evidence of cyclonic events is
acceptable (above 10% loss of canopy crown cover in the last 20 years).
Streams and ephemeral pools:
– The benchmark must be the mostly pristine habitat.
– The stream locations where benchmark data were acquired must be surrounded by quasi-pristine
and natural habitat, unaffected by any major human-induced disturbance and under pristine forest
cover.
