Content uploaded by Philip Gibbons
Author content
All content in this area was uploaded by Philip Gibbons on Feb 11, 2018
Content may be subject to copyright.
PRIMARY RESEARCH ARTICLE
Outcomes from 10 years of biodiversity offsetting
Philip Gibbons
1
|
Andrew Macintosh
2
|
Amy Louise Constable
2
|
Kiichiro Hayashi
3
1
Fenner School of Environment and
Society, The Australian National University,
Canberra, ACT, Australia
2
College of Law, The Australian National
University, Canberra, ACT, Australia
3
Civil and Environmental Engineering
International Programs Office, Nagoya
University, Nagoya, Japan
Correspondence
Philip Gibbons, Fenner School of
Environment and Society, The Australian
National University, Canberra, ACT,
Australia.
Email: philip.gibbons@anu.edu.au
Abstract
We quantified net changes to the area and quality of native vegetation after the
introduction of biodiversity offsetting in New South Wales, Australia—a policy
intended to “prevent broad-scale clearing of native vegetation unless it improves or
maintains environmental values.”Over 10 years, a total of 21,928 ha of native veg-
etation was approved for clearing under this policy and 83,459 ha was established
as biodiversity offsets. We estimated that no net loss in the area of native vegeta-
tion under this policy will not occur for 146 years. This is because 82% of the total
area offset was obtained by averting losses to existing native vegetation and the
rate that these averted losses accrue was over-estimated in the policy. There were
predicted net gains in 10 of the 14 attributes used to assess the quality of habitat.
An overall net gain in the quality of habitat was assessed under this policy by sub-
stituting habitat attributes that are difficult to restore (e.g. mature trees) with habi-
tat attributes for which restoration is relatively easy (e.g. tree seedlings). Long-term
rates of annual deforestation did not significantly change across the study area after
biodiversity offsetting was introduced. Overall, the policy examined here provides
no net loss of biodiversity: (i) many generations into the future, which is not consis-
tent with intergenerational equity; and (ii) by substituting different habitat attributes,
so gains are not equivalent to losses. We recommend a number of changes to biodi-
versity offsetting policy to overcome these issues.
KEYWORDS
averted loss, avoided loss, biodiversity offsets, deforestation, land clearing, restoration
1
|
INTRODUCTION
Biodiversity offsetting (also known as compensatory habitat and mit-
igation banking) is a policy in which residual impacts of development
on biodiversity (i.e. after actions are taken to avoid and minimize
impacts) are compensated with actions elsewhere—typically with the
objective of achieving no net loss of biodiversity (BBOP, 2012b).
Offsetting is considered a flexible alternative to traditional com-
mand-and-control regulation because it theoretically allows greater
scope for development to continue without net loss of environmen-
tal values (Fromond, Simil€
a, & Suvantola, 2009; Gardner, Hase, &
Brownlie, 2013). Although introduced as policy in the 1970s in the
United States of America and Germany (Rundcrantz & Sk€
arb€
ack,
2003); and in the 1990s in Canada (Rubec & Hanson, 2009), the
Netherlands and Sweden (Tischew, Baasch, Conrad, & Kirmer, 2010),
there has been a surge in uptake of biodiversity offsetting in other
countries, companies and institutions in the last decade, possibly due
to its promotion by the Business and Biodiversity Offsets Pro-
gramme (BBOP) since 2004. As of 2015, biodiversity offsetting poli-
cies were employed in approximately 40 countries (Maron, 2015),
with an additional 29 countries in which the policy is in development
(Maron, Ives, & Kujala, 2016).
No net loss is considered an underlying principle of biodiversity
offsetting (BBOP, 2012b, Ten Kate, Pilgrim, & Brooks, 2014) and
most organizations that employ the policy cite no net loss of biodi-
versity as an objective (BBOP, 2012a, Jenner & Balmforth, 2015;
Mckenney & Kiesecker, 2010; Miller, Trezise, & Kraus, 2015; Qu
e-
tier, Regnery, & Levrel, 2014). However, outcomes from a long
Received: 4 July 2017
|
Accepted: 21 September 2017
DOI: 10.1111/gcb.13977
Glob Change Biol. 2017;1–12. wileyonlinelibrary.com/journal/gcb ©2017 John Wiley & Sons Ltd
|
1
history of ecological restoration has led to speculation that no net
loss can be achieved in a narrower range of scenarios than biodiver-
sity offsetting policy is typically applied (Curran, Hellweg, & Beck,
2014; Gibbons, Evans, & Maron, 2016; Hilderbrand, Watts, & Ran-
dle, 2005; Laitila, Moilanen, & Pouzols, 2014; Maron, Hobbs, &
Moilanen, 2012). This suggestion has tentative support from the
evaluation literature. That is, although there is evidence that biodi-
versity offsetting has delivered no net loss (or net gain) in biodiver-
sity for individual projects (Norton, 2009; Pickett et al., 2013), no
net loss of biodiversity has yet to be demonstrated at a programme
level. In a review of 68 wetland offsets (mitigation banks) established
in the USA, Brown and Lant (1999) estimate a net loss of 21,328
acres (8,631 hectares) of wetland. In a review of 70 mitigation pro-
jects in California, Sudol and Ambrose (2002) judge that 16% were
successful and in a review of 103 fish compensation projects in
Canada, Harper and Quigley (2005) report that 64% of projects
achieved no net loss.
Other studies point to the general absence of sufficient data to
evaluate whether biodiversity offsetting achieves no net loss. In a
review of wetland mitigation in Canada, which was introduced in the
1990s, Rubec and Hanson (2009) conclude that it remains unknown
if no net loss objectives are being achieved. In a review of German
compensatory habitat policy applied to road developments, Tischew
et al. (2010) could not evaluate 26 of 57 projects. May, Hobbs, and
Valentine (2017) noted inadequate reporting for 18% of offset pro-
jects implemented in Western Australia over 11 years. The combina-
tion of few evaluations and lack of suitable data for biodiversity
offsetting programmes means a consensus has yet to emerge as to
whether biodiversity offsetting achieves no net loss of biodiversity.
A key challenge when evaluating any policy intervention—and
particularly biodiversity offsetting—is defining the counterfactual, or
what would have occurred were the policy not implemented (Ferraro
& Pattanayak, 2006). Different counterfactuals are relevant when
evaluating whether offsetting achieves no net loss of biodiversity
(Bull, Gordon, Law, Suttle, & Milner-Gulland, 2014). The first coun-
terfactual relates to the biodiversity on an offset site in the absence
of a change to management. Because biodiversity offsets are pro-
cured partly or wholly by averting future losses of biodiversity, we
can only calculate the gain in biodiversity for averted loss offsets by
comparing the future trajectory of biodiversity with and without the
change in management that comes with establishing the offset (Gib-
bons, Briggs, & Ayers, 2009). A second counterfactual that is rele-
vant for evaluating biodiversity offsetting policy is the outcome for
biodiversity where it replaces another policy instrument. On one
hand, biodiversity offsetting places a price on biodiversity equivalent
to the cost of its replacement, thereby theoretically creating an eco-
nomic incentive for developers to avoid or minimize impacts on bio-
diversity. A counterargument is that biodiversity offsetting
incentivizes greater losses of biodiversity when introduced to replace
traditional command-and-control regulation, partly because it legit-
imizes biodiversity loss based on the (sometimes unrealistic) expecta-
tion that this loss can be adequately compensated (Apostolopoulou
& Adams, 2015; Gordon, Bull, Wilcox, & Maron, 2015; Moreno-
Mateos, Maris, B
echet, & Curran, 2015; Walker, Brower, Stephens,
& Lee, 2009).
As noted by Curran, Hellweg, and Beck (2015) in a response to
Qu
etier, Van Teeffelen, Pilgrim, Von Hase, and Ten Kate (2015), con-
sensus is unlikely to emerge on these issues until there is a clearer
picture of outcomes from biodiversity offsetting programmes. To
contribute to this evidence-base, we evaluated a biodiversity offset-
ting programme implemented for a decade in New South Wales,
Australia using the two counterfactuals discussed above. We asked:
(i) What is the net change in area and quality of native vegetation
under a biodiversity offsetting policy? And (ii) How did rates of
native vegetation loss change after the introduction of biodiversity
offsetting (i.e. did they increase or decrease)?
2
|
MATERIALS AND METHODS
2.1
|
Study area
We addressed these research questions using data from a biodiver-
sity offsetting programme implemented from 1 December 2005 to
20 November 2015 as part of legislation (Native Vegetation Act
2003) to regulate clearing of native vegetation for rural land uses in
the state of New South Wales, Australia—an area of 809,444 km
(over 3 times the size of United Kingdom). An objective of the
Native Vegetation Act is “to prevent broad-scale clearing unless it
improves or maintains environmental outcomes.”Biodiversity offset-
ting was one policy instrument used to meet this objective. That is,
developments which required offsets were only approved where the
predicted losses of habitat from development were compensated by
commensurate predicted gains in habitat at offset sites that are pro-
tected in perpetuity (New South Wales Government Office of Envi-
ronment and Heritage, 2013).
2.2
|
Net change in the area of native vegetation
We estimated net change in the area of native vegetation under this
biodiversity offset policy as the balance of losses of native vegeta-
tion from development and gains from offsets. All calculations of
area were based on the extent of the canopy cover of woody vege-
tation. We assumed that gains in the area of native vegetation on
offset sites accrued from: (i) restoration (i.e. creating new habitat)
and/or (ii) averted losses (i.e. removing threats that would have led
to future declines of habitat under the counterfactual). For this anal-
ysis we used data from developments approved with offsets
between 1 December 2005 and 20 November 2015 that were con-
tained in a public register of approvals (http://www.environment.
nsw.gov.au/vegetation/approvedclearing.htm). The area of habitat
lost due to development was summed across all development
approvals that required an offset. Where the clearing was of rem-
nant trees scattered throughout previously cleared land, the calcula-
tion of area was based on the net area of canopy cleared (not the
gross area impacted). Sites located in non-woody habitats (e.g. grass-
lands) were excluded from this analysis.
2
|
GIBBONS ET AL.
The area of native vegetation gained from offsets was estimated
using different methods depending on whether the offset was based
on averted loss or restoration. We located each offset site on Goo-
gle Earth and evaluated whether it fell into one of three categories:
(i) relatively intact overstorey, which we defined as offset sites domi-
nated by canopy cover similar to surrounding intact native vegeta-
tion (e.g. reserves, roadsides and large remnants; Figure 1a); (ii)
scattered trees, which were offset sites dominated by canopy cover
occurring at a lower density than surrounding areas containing intact
native vegetation (Figure 1b) and (iii) cleared, which were offset sites
dominated by continuous areas without trees or offset sites domi-
nated by recently planted seedlings (Figure 1c). Offset sites
dominated by relatively intact tree cover (Figure 1a) were assumed
to provide 100% of gains in the area of native vegetation by avert-
ing future losses. Offset sites dominated by scattered trees (Fig-
ure 1b) were assumed to provide 50% of gains in the area of native
vegetation by averting future losses and the other 50% of gains in
the area of native vegetation by restoration. And offsets dominated
by no tree cover or seedlings (Figure 1c) were assumed to provide
100% of gains in the area of native vegetation by restoration.
We estimated the area of native vegetation gained from averted
loss for offset sites over t=1tonyears as
Xn
t¼1ðOtPtrÞ;(1)
where O
t
is the total area of offsets established at tyears, P
t
is the
proportion of the total area of the offsets based on averted loss at t
years and ris the annual rate of loss of native vegetation under the
counterfactual. We did not apply time-discounting when estimating
the net present value of averted losses because the policy did not
provide scope for this.
We used different estimates for the annual rate of loss for native
vegetation on offset sites under the counterfactual (r) depending on
whether offset sites were dominated by relatively intact tree cover
or scattered trees. Areas where the canopy cover exceeds approxi-
mately 20% are reliably classified as continuous areas of tree cover
using Landsat satellite imagery (Lehmann, Wallace, Caccetta, Furby,
& Zdunic, 2013). Thus, we assumed that historic vegetation mapping
(which is based on Landsat imagery) can be used to estimate annual
rates of change for sites in which the tree cover was relatively
intact. For offsets established on sites dominated by a relatively
intact canopy cover we predicted rusing a similar method to Maron,
Bull, Evans, and Gordon (2015). That is we compared historic esti-
mates of native vegetation cover with contemporary estimates of
native vegetation cover excluding tenures not covered by this policy.
We estimated the annual loss of woody vegetation across the study
area under the counterfactual (r) for these sites as 0.13% based on
the formula
C=HðÞ
1=t1;(2)
where His the total area of woody native vegetation available for rural
land uses to which the policy applies (i.e. on freehold and leasehold
land) across the study area in 1998 as detected by Landsat satellite
imagery (14,012,000 ha) (Bureau of Rural Sciences, 1998) and Cis H
minus the total area cleared for rural land uses (cropping, pasture and
thinning) over t=13 years (i.e. 1998–99 to 2010–11) (234,900 ha) as
detected using Landsat satellite imagery (Office of Environment and
Heritage of the New South Wales Government, 2014).
Sites with scattered trees are not reliably detected as tree cover
from classifications of Landsat imagery (Gibbons et al., 2007). For
offsets established on sites dominated by scattered trees we esti-
mated the annual rate of loss for woody vegetation as 0.90%, which
is the average annual loss observed for scattered trees from aerial
photography in cultivated and grazed landscapes within the study
area over 30–34 years (Ozolins, Brack, & Freudenberger, 2001). This
FIGURE 1 Examples of three sites identified as offsets under the
policy examined here. All offset sites were classified as providing
gains in the area (not quality) of native vegetation as: (a) 100%
averted losses, (b) 50% averted losses and 50% restoration and (c)
100% restoration. Images (a) and (c) ©2017 CNES/Airbus and image
(b) ©2017 DigitalGlobe
GIBBONS ET AL.
|
3
figure is higher than the annual rate of loss for relatively intact
native vegetation (0.13%) because scattered trees in cultivated and
grazed landscapes are subject to higher rates of clearing than intact
native vegetation (Gibbons et al., 2009), and suffer from dieback
from a range of causes (Landsberg, 1993), which is exacerbated
because natural regeneration does not typically occur on cultivated
or intensively grazed sites supporting scattered trees (Weinberg, Gib-
bons, Briggs, & Bonser, 2011).
2.3
|
Net change in quality of native vegetation
We obtained data from the government regulator (New South Wales
Office of Environment and Heritage) on predicted losses (from
development) and gains (from offsets) in the quality of native vege-
tation from sites approved for development between 2008 and
2015. These assessments were undertaken by government employ-
ees using a combination of ground surveys and satellite images
based on methods described in Gibbons et al. (2009) and Depart-
ment of Environment, Climate Change and Water (2011). These data
represent predicted losses (from development) and gains (from off-
sets) over an undefined timeframe rather than actual changes in the
quality of vegetation. No data were available on actual changes of
vegetation quality that have occurred to sites considered in this pol-
icy. For this study we used current and predicted future values for
each of: (i) habitat quality at the patch scale, (ii) connectivity to adja-
cent native vegetation and (iii) the cover of native vegetation in the
landscape (Table 1). A full description of all methods used by the
regulator is described in Gibbons et al. (2009), Department of Envi-
ronment Climate Change and Water NSW (2011) and New South
Wales Government Office of Environment and Heritage (2013).
To estimate the initial habitat quality at the patch scale for each
proposal, each of 10 habitat variables (Table 1) were measured by field
assessors employed by the regulator in plots at each development site
and corresponding offset sites and these observations were then
expressed as a score of 0, 1, 2 or 3 based on the difference between
the mean observed value for the variable and a reference value for
that variable (the expected value for that variable were the site rela-
tively unmodified). For example if a mean of 20 native plant species
were measured in a plot and the reference value for that variable at
that location was 20 species then the score for that variable would be
3. A predicted score for each habitat variable on the site given the
proposed development (for development sites) or proposed offset (for
offset sites) was then estimated using a standard protocol based on
the starting condition of the site and the combination of management
actions employed. For example if the site above was to be cleared for
cultivation then the predic ted score for native plant species ric hness is
likely to change from a current score of 3 to a predicted score of 0.
The change in scores for each habitat variable (i.e. with development
or with offsetting) was then calculated by subtracting current scores
from the predicted scores. For the above example, the change in
native plant species richness on the development site would be 3.
Gibbons et al. (2009) specified that the timeframe for these predic-
tions should be no greater than a human generation, although the
timeframe over which gains are predicted to accrue are not explicit in
the policy. The predicted gain for some variables (e.g. total length of
fallen logs) is based on averted loss. That is, if the landholder agrees to
surrender their legal right to degrade the variable (e.g. removal of logs
for firewood is legally permitted) then gain is calculated as the differ-
ence between the current value of the variable and the value of this
variable if this legal right was exercised. The regulator combines scores
for the patch-scale habitat variables (Table 1) into a single score and
multiplies this score by area to arrive at a quantity 9quality metric for
each site. However, we could not calculate this score because we
could not match habitat quality data and area for each proposal and
we did not want to confound these variables by combining them.
Four landscape attributes were also used to assess changes in the
quality of vegetation by the regulator: connectivity and the cover of
native vegetation at three spatial scales (Table 1). Connectivity was
expressed as a score based on the number, quality and width of link-
ages to adjacent patches and the size and quality of those adjacent
patches (New South Wales Government Office of Environment and
Heritage, 2013). The amount of native vegetation in the landscape at
different scales around each development and offset site was mea-
sured as one of four classes (0%–10%, 11%–30%, 31%–70% and
>70%) or in 10% increments (i.e. the methodology changed over the
period examined here). These data were visually estimated by asses-
sors employed by the regulator using SPOT5 satellite imagery. Losses
for each attribute (Table 1) were based on the difference between
observed values (prior to development) and predicted values (assum-
ing the development would proceed) and gains for each variable were
based on the difference between observed values on proposed offset
sites and predicted values given the management actions agreed by
the proponent at offset sites over an undefined timeframe. We calcu-
lated all losses and gains by taking the median of each recorded %
value and averaged these across all sites.
TABLE 1 The variables used as surrogates to assess losses and
gains in the quality of native vegetation at development and offset
sites that were examined in this study
Scale Variable
Patch Native plant species richness
Native overstorey cover
Native midstorey cover
Native ground stratum cover (grasses)
Native ground stratum cover (shrubs)
Native ground stratum cover (other)
Exotic plant cover
Number of trees with hollows
Proportion of overstorey species regenerating
Total length of fallen logs
Landscape Connectivity score (width and quality of linkages and size
of adjacent patch)
Cover of native vegetation within 1.79 km (1,000 ha)
Cover of native vegetation within 0.55 km (100 ha)
Cover of native vegetation within 0.18 km (10 ha)
4
|
GIBBONS ET AL.
2.4
|
Net change in rates of clearing
We estimated net changes in rates of clearing for relatively intact
native vegetation for rural land uses to which the policy applies (pas-
ture, cropping and thinning) before, and after, the introduction of bio-
diversity offsetting using data on the change in area of woody
vegetation from 1988–89 to 2010–11 (Office of Environment and
Heritage of the New South Wales Government, 2014). These data
were estimated using supervised classifications of Landsat imagery
(Queensland Department of Science, Information, Technology and
Innovation, 2015) that detected woody vegetation approximately >2
metres tall and with ≥20 per cent crown cover (Office of Environment
and Heritage of the New South Wales Government, 2014). All esti-
mates of woody vegetation change after 2011 were made using a dif-
ferent methodology and thus were not comparable and could not be
used. We tested whether there was empirical support for a change in
the annual amount of native vegetation cleared for rural land uses
between 1988 and 2006 (before biodiversity offsets were employed)
and 2006 to 2011 (after biodiversity offsets were employed). To do
this we log
10
transformed the response variable (annual amount of
woody vegetation cleared) so it was normally distributed and fitted a
linear regression model with the following explanatory variables:
whether clearing occurred in a year before or after offsetting was
introduced and annual national farmers’terms of trade which repre-
sents an index of the balance of farmers’expenditure and receipts
(Australian Bureau of Agriculture and Resource Economic Sciences,
unpublished). Annual farmers’terms of trade is a variable that has been
correlated with land clearing rates in Australia (Macintosh, 2014) and
thus has potential to confound the influence of a change in policy on
the amount of woody vegetation cleared annually. Because there is
potential that the amount of native vegetation cleared between subse-
quent years is not independent, we determined the extent to which
the residuals from this regression model were correlated through time
using an autocorrelation function (ACF) plot. This plot indicated a sig-
nificant degree of correlation between the model residuals in adjacent
years. We therefore predicted the (log
10
) annual amount of native veg-
etation cleared using an autoregressive integrated moving average
(ARIMA) model with an MA(1) structure where the years in which
offsets were applied and annual national farmers’terms of trade were
fitted as explanatory variables. Model selection was undertaken with
Akaike’s Information Criterion (AIC) and diagnostic plots included stan-
dardized residuals fitted against time, a normal Q–Q plot of standard-
ized residuals and an ACF plot to determine if any correlation between
model residuals persisted. These analyses were undertaken using the
mgcv and astsa packages in R (R Development Core Team, 2010).
3
|
RESULTS
3.1
|
A description of sites approved for
development and offsets
The regulator approved a total of 497 developments with offsets
from 1 December 2005 until 20 November 2015. A total of
21,928 ha of native vegetation cleared for rural land uses was
assessed to “improve or maintain”environmental outcomes subject
to the establishment of 83,459 ha of biodiversity offsets. Overall,
3.8 ha of offset was established for every hectare of native vegeta-
tion approved for clearing to meet this standard. Thirty-eight per
cent of approvals included clearing of relatively intact native vegeta-
tion. A total of 13,171 ha of relatively intact native overstorey was
approved for clearing with a mean area of 34 ha per approval.
Seventy per cent of approvals included some clearing of scattered
native trees. A total of 71,371 scattered trees were approved for
clearing (mean 204 scattered trees per approval) with a stated total
area of 7,423 ha. Two per cent of all approvals that required an off-
set included some thinning of native vegetation. The area approved
for thinning totalled 1,294 ha, with a mean area of 162 ha per
approval.
Sites approved for clearing were, on average, poorer quality and
occurred in more fragmented landscapes than sites approved as off-
sets. The mean quality score (derived from 10 habitat variables)
(Table 1) recorded by the regulator at sites approved for clearing
(0.19 of a maximum of 1.0), was significantly less than the mean
quality score at sites approved as offsets (0.46 out of a maximum of
1.0) (p<.001, one-tailed t-test) (Figure 2a). There was greater per
cent cover of native vegetation within 1,000 ha (p=.053, one-tailed
t-test), 100 ha (p<.001, one-tailed t-test) and 10 ha (p<.001, one-
tailed t-test) of the centroid of offset sites compared with develop-
ment sites (Figure 2b).
3.2
|
Net change in the area of native vegetation
The net area of native vegetation approved for clearing (with offsets)
from 2005 to 2015 was 21,928 ha. Of the 83,459 ha established to
offset this loss, 0.9% (771 ha) was established on land predominantly
cleared of canopy cover, 7.8% (6520 ha) was established in remnant
vegetation dominated by scattered canopy cover and 91.3%
(76,168 ha) was in relatively intact remnant canopy cover. Using his-
toric annual losses of intact native vegetation (0.13%) and scattered
trees (0.90%) as the counterfactual, we estimated that no net loss in
the area of native vegetation under this policy will occur in the year
2150—a time-lag of 146 years from the introduction of the policy
(Figure 3). Estimated gains in the area of native vegetation from off-
sets at the end of the year 2150 comprised: averted loss of
13,783 ha of intact canopy cover (63%), averted loss of 4,129 ha of
modified canopy cover (19%) and restoration of 4,031 ha (18%).
3.3
|
Net change in the quality of native vegetation
Of the records provided by the regulator, n=94 approved develop-
ment proposals between 2008 and 2015 contained data for each of
the 10 patch scale variables in Table 1. Data for these 94 proposals
were reported in 132 separate zones (relatively homogenous strata)
for clearing sites and 172 zones for offset sites. We calculated the
% change in mean scores at the zone level across all development
and offset sites for each of the 10 patch-scale variables (Table 1).
GIBBONS ET AL.
|
5
There were predicted net gains in mean scores for 6 of the 10 habi-
tat surrogates used to assess habitat quality: regeneration of native
overstorey (+97%), cover of native grasses (+87%), total length of
fallen logs (+78%), cover of native ground cover plants (other than
grasses and shrubs) (+64%), reduction in exotic plant cover (+39%)
and cover of midstorey plants (+33%) (Figure 4). We calculated pre-
dicted net losses in mean scores for four habitat surrogates: native
overstorey cover (49%), numbers of trees with hollows (34%),
native plant species richness (32%) and native ground cover shrubs
(1%) (Figure 4).
The regulator also assessed changes in connectivity at develop-
ment and offset sites. Of n=170 development proposals that con-
tained data used to calculate a connectivity index (New South Wales
Government Office of Environment and Heritage, 2013), declines in
the connectivity index were predicted on 9% of development sites
and improvements in connectivity were predicted on 12% of offset
sites, suggesting an overall predicted net gain in connectivity using
this measure. As part of an assessment of landscape-scale impacts,
the regulator also collected data on changes in the % cover of native
vegetation within 1,000 ha, 100 ha and 10 ha for these 170 sites
(Table 1). At development sites there were declines in mean native
vegetation cover within 1,000 ha, 100 ha and 10 ha of 0%, 0.24%
and 1.76% respectively. At offset sites there were gains in mean
native vegetation cover at each of these scales of 0.29%, 0.85% and
2.2% respectively. Thus, there were overall net gains for mean cover
of native vegetation within 1,000 ha, 100 ha and 10 ha of 0.29%,
0.62% and 0.45% respectively.
3.4
|
Net change in rates of clearing after the
introduction of offsets
The total area of native vegetation cleared for rural land uses (crop-
ping, pasture, thinning) (which is the focus of this policy) (Office of
Environment and Heritage for New South Wales, 2014) across the
study area in the 5 years after the introduction of biodiversity off-
setting (2006–2011) (80,100 ha) was 28% less than clearing of
native vegetation recorded in the 5 years before the introduction of
biodiversity offsetting (2000–2005) (111,300 ha). Using the annual
area of native vegetation cleared for rural land uses (cropping, pas-
ture, thinning) from a longer time-series (1988–1989 to 2010–2011)
0
5
10
15
20
25
30
35
40
45
1,000 ha 100 ha 10 ha
% native vegetation cover
Area around site
(a)
Development sites Offset sites
0
0.1
0.2
0.3
0.4
0.5
Development sites Offset sites
Initial condition
(b)
FIGURE 2 The initial condition
(mean 95% confidence interval) of
development sites and offset sites scored
between 0 and 1 (a) and the initial % cover
of native vegetation (mean 95%
confidence interval) within 1000 ha,
100 ha and 10 ha of the centroid of
development sites and offset sites (b)
6
|
GIBBONS ET AL.
(Figure 5), we tested whether the annual rate of clearing (log
10
)
changed when biodiversity offsetting was introduced. The clearing
estimate for 2010–2011 (6,600 ha) (Figure 5) was identified as an
outlier in diagnostic plots for the fitted model and was therefore
excluded from this analysis. The ARIMA model with the lowest AIC
value was the null model, indicating that models including a variable
indicating whether clearing occurred before or after offsetting was
introduced, national farmers’terms of trade or both of these vari-
ables were no better than the null model containing only the inter-
cept (Table 2). That is, when we excluded the datum from 2010 to
11, there was no empirical evidence for a change in annual rates of
clearing when the offsetting policy was introduced, even when con-
trolling for annual farmers’terms of trade.
4
|
DISCUSSION
We used available data from a biodiversity offsetting programme
introduced in south-eastern Australia to answer the following ques-
tions: (i) What was the net change in area and quality of native
vegetation under this policy? And (ii) How did rates of native vegeta-
tion loss change after the introduction of biodiversity offsetting?
Overall, we found that this policy resulted in: (a) net loss in the area
of native vegetation for 146 years, (b) net gains for 10 of the 14
habitat surrogates used to assess the quality of native vegetation
and (c) was not associated with a change in annual rates of land
clearing compared with previous policies (that did not employ offset-
ting). We discuss each of these findings in greater detail below.
4.1
|
Net change in the area of native vegetation
When we defined the counterfactual on offset sites as habitat loss
continuing at historic rates, we estimated that net gain in the area of
native vegetation will not occur for 146 years. This is because 82%
of the area offset was obtained by averting losses to existing native
vegetation and the rate that averted losses accrue was overesti-
mated. Explicitly defining the counterfactual—what would happen
were the action not introduced—is critical when evaluating offset-
ting policy (Bull et al., 2014). Although the rate of native vegetation
loss under the counterfactual was not explicitly defined in the policy
–25000
–20000
–15000
–10000
–5000
0
5000
10000
15000
20000
2005
2015
2025
2035
2045
2055
2065
2075
2085
2095
2105
2115
2125
2135
2145
Area (ha)
Year
Native vegetation cleared
Averted loss (intact over-storey)
Averted loss (scattered trees)
Restoration
Net area of native vegetation
FIGURE 3 Cumulative losses (from
development) and gains (from offsets) of
native vegetation (ha) managed under the
biodiversity offset policy examined here
comprising: losses from development
between 2005 and 2015 (solid grey line),
averted losses of relatively intact remnant
native vegetation (long-dashed line),
averted losses of modified native
vegetation (i.e. scattered trees) (short-
dashed line), gains from restoration (dotted
line), and the net amount of native
vegetation based on adding losses and all
gains (solid black line). No net loss of
native vegetation was estimated to occur
from the year 2150
–60 –40 –20 0 20 40 60 80 100 120
Lack of exotic plant cover
Native ground cover (grasses)
Native ground cover (other)
Native ground cover (shrubs)
Native mid-storey cover
Native over-storey cover
Native plant species richness
Number of trees with hollows
Native over-storey regeneration
Total length of fallen logs
% change
FIGURE 4 The % change for mean
scores recorded for 10 attributes used to
assess habitat quality across all
development and offset sites. The data
represent the net change in scores for
each habitat attribute when % losses from
development and % gains from offsets are
summed across all sites
GIBBONS ET AL.
|
7
examined here, Maron et al. (2015) estimate this figure to be 0.5%
per annum. This is well above the rate we estimated for intact native
vegetation (0.13% per annum) which represented 91% of the total
area of offset sites. Gordon et al. (2015) suggest that biodiversity
offsetting creates a perverse incentive to overstate the counterfac-
tual. Thus, explicit estimates for rates of habitat loss under the coun-
terfactual are critical data that should be stated and used
transparently when making decisions about averted loss offsetting.
Our estimate that no net loss of native vegetation under this
policy will not occur for 146 years would be exacerbated if we
applied time-discounting to averted losses, which is recommended
by several authors (Gibbons et al., 2016; Laitila et al., 2014; Moila-
nen, Teeffelen, Ben-Haim, & Ferrier, 2009) and is employed in sev-
eral biodiversity offsetting policies (Miller et al., 2015; NOAA, 2002),
but not the policy examined here. Employing a discount rate of 3%,
Gibbons et al. (2016) concluded that averted loss offsets are only
likely to be feasible (i.e. offset ratios ≤10:1) where the rate of loss of
biodiversity under the counterfactual is ≥6% per annum, which is
well above our estimates of 0.13% (for intact vegetation) and 0.9%
(for scattered paddock trees). Our results suggested that averted loss
offsetting as employed in the policy observed here will only deliver
no net loss many generations into the future.
It could be argued that a counterfactual based on historic rates
of clearing, as we have employed in this study, is not appropriate
under a no net loss policy (Maron et al., 2015). That is if a site was
developed under the counterfactual then theoretically there should
be no net loss of biodiversity. In this case, averted losses could only
accrue on sites that can be developed outside a no net loss policy
and/or where losses of biodiversity are unrelated to development.
The majority of historic losses of native vegetation in our study area
were due to approved land clearing that occurred prior to the intro-
duction of a no net loss policy (Office of Environment and Heritage
of the New South Wales Government, 2014). If the counterfactual
on these sites is no net loss, then the majority of offsets secured
under the policy examined here will provide no gains in biodiversity
and thus there would be a considerable net loss in biodiversity under
this policy.
4.2
|
Net change in the quality of native vegetation
Although there were net losses in 4 of the 14 habitat surrogates,
the regulator calculated an overall net gain of habitat quality under
this policy. This is because individual attributes are combined to cre-
ate overall scores of habitat quality at the patch scale and landscape
scale. Combining variables in a metric—and therefore allowing these
variables to be substituted—leads to dissimilarities between losses at
development sites and gains at offset sites (Ten Kate et al., 2014).
We found that allowing substitution resulted in replacement of
patch-scale habitat attributes that are difficult to restore (e.g. native
plant species richness, mature trees) with habitat attributes for which
restoration or improvement is relatively easy (e.g. establishing tree
seedlings, improving the cover of plant life-forms for which propag-
ules are available) (Figure 4). Our results therefore suggested that
habitats that take a long time to develop (e.g. old-growth trees) or
habitats for which complete restoration is highly uncertain (e.g.
unmodified ecosystems) are likely to be lost or degraded if managed
using biodiversity offsetting. Substitution was also evident with
respect to changes in landscape-scale attributes. Development
0
5000
10000
15000
20000
25000
30000
35000
Area (ha)
Year
FIGURE 5 Annual clearing of native
woody vegetation (ha) for rural land uses
(crop, pasture, thinning) in New South
Wales before (solid bars) and after (white
bars) the introduction of biodiversity
offsetting. These data were estimated from
supervised classifications of Landsat
satellite imagery undertaken every 2 years
until 2005 and then annually from 2005-
11 (Source Office of Environment and
Heritage of the New South Wales
Government, 2014)
TABLE 2 ARIMA models used to predict the amount of annual
clearing of woody native vegetation fitted with the intercept (the
null model) (Model 1), a variable indicating whether the clearing
occurred before or after biodiversity offsetting was introduced
(Model 2) and/or a variable representing national farmers’terms of
trade (Model 3)
Model Intercept
Before or after
offsetting intro-
duced
Farmers’
Terms of
Trade AIC ΔAIC
1+45.65 0
2++ 43.67 1.98
3++ + 42.21 3.44
Models are arranged by ascending AIC values with differences in AIC val-
ues (ΔAIC) calculated relative to the “best”model.
8
|
GIBBONS ET AL.
tended to occur in more fragmented landscapes and offsets in more
intact landscapes (Figure 2b). While this may benefit some taxa, the
species-area curve indicates that the same amount of habitat loss (or
gain) has greater marginal impact on species richness overall in land-
scapes with less remnant vegetation (Cunningham, Lindenmayer, &
Crane, 2014; Radford, Bennett, & Cheers, 2005).
To reduce substitution, biodiversity offsetting metrics often
weight variables according to their importance for biota (Parkes,
Newell, & Cheal, 2003) or their relative difficulty to restore once lost
(Gibbons et al., 2009). However, this only exacerbates substitution.
For example giving more weight to mature trees because they are
difficult to replace will only result in greater use of other habitat
attributes (e.g. seedlings) to offset the loss of mature trees. One
alternative suggested by Gibbons et al. (2016) and Maseyk et al.
(2016) and adopted (in part) by the Australian (Miller et al., 2015)
and New Zealand (Department of Conservation, 2014) Governments
is to keep as many variables or components of biodiversity separate
when undertaking assessments, which limits opportunities for substi-
tution. However, one trade-off of this approach is that, as biodiver-
sity credits become more complicated they also become less
fungible and therefore difficult to trade in a common market (Habib,
Farr, Schneider, & Boutin, 2013).
4.3
|
Net change in rates of clearing after the
introduction of offsets
The introduction of biodiversity offsetting is variously predicted to:
(i) reduce rates of clearing because of the increased economic incen-
tive they place on developers to avoid impacts or; (ii) increase rates
of clearing because of a perception that development can occur in a
greater range of locations without net impact (Moreno-Mateos et al.,
2015; Walker et al., 2009). Although total clearing of native vegeta-
tion for rural land uses in the 5 years after introduction of the policy
we examined was 28% less than total clearing in the 5 years imme-
diately before introduction of the policy, rates of land clearing in
rural areas of Australia vary with other factors, such as farmers’
terms of trade (Macintosh, 2014). We therefore examined annual
rates of clearing over a longer time-frame (1988–2011). Controlling
for changes in farmers’terms of trade, we found no evidence that
annual rates of land clearing after the introduction of biodiversity
offsets in our study area were different from annual rates of land
clearing before introduction of the policy (Figure 5 and Table 2).
It is difficult to isolate the effects of biodiversity offsetting on
clearing rates because this policy was introduced alongside exemp-
tions that enabled a substantial amount (approximately 87%) of rural
clearing to occur without offsetting the loss (data from Office of
Environment and Heritage of the New South Wales Government,
2014). That is it appears that any additional constraints imposed on
development through the introduction of offsetting were defrayed
by exemptions that were introduced progressively over the 10 years
that the policy persisted. Thus, our results neither support the
hypothesis that offsetting leads to reduced land clearing nor the
hypothesis that offsets leads to increased rates of land clearing.
Instead, we hypothesize that the amount of land clearing is princi-
pally driven by underlying economic stimuli, which in our study area
is increasingly linked to demand for agricultural products in the
emerging economies of China and India. This hypothesis has partial
support from studies in Australia by Macintosh (2014), Evans (2016)
and (Department of Climate Change and Energy Efficiency, undated)
who have found correlations between macroeconomic variables
associated with agriculture (the key driver of land clearing in Aus-
tralia) and deforestation rates.
4.4
|
Policy recommendations
Our study provides a number of general lessons for biodiversity off-
setting policy.
First, our study highlights the value of integrating sound data col-
lection and reporting systems with the implementation of policy. The
policy examined here was introduced alongside the establishment of
a public register of approval decisions, the metric created to imple-
ment the policy required the collection of consistent, quantitative
data at each site and investment was made by the government regu-
lator to regularly map changes to the area of native vegetation (using
satellite imagery). Most of these data were made publicly available.
However, lacking from these data is the extent to which conditions
imposed upon developers were undertaken on the ground. Further-
more, a subset of offset sites should be monitored so the predicted
change in condition could be compared with actual data. However,
that the New South Wales Government reviewed the policy exam-
ined here without a thorough review of these data (Byron, Craik,
Keniry, & Possingham, 2014) highlights a point made by Mcdonald-
Madden, Baxter, and Possingham (2008) that collecting appropriate
data to evaluate policy is only effective if the organization is also
willing to act upon this evidence.
Second, our study confirms warnings by Maron et al. (2015),
Gordon et al. (2015) and Gibbons et al. (2016) that biodiversity gains
procured from averted loss offsetting can be easily overstated. Our
study highlights that this leads to considerable delays before no net
loss occurs, which is not consistent with the principle of intergenera-
tional equity—a key principle underpinning sustainability. Biodiver-
sity offsetting policies should define appropriate sources of averted
loss, justify how averted losses can be calculated on land subject to
a no net loss policy and make explicit the rates of loss that are used
when calculating averted losses.
Third, our study indicates that biodiversity offsetting that is
underpinned by metrics that allow variables to be substituted will
lead to the replacement of biodiversity attributes that are difficult to
restore with those that are easier to restore. Affording attributes
that are difficult to restore a greater weight within these metrics
does not address this issue. Instead, biodiversity attributes should be
enumerated and offset separately (Gibbons et al., 2016; Maron et al.,
2016).
Fourth, our results suggest that biodiversity offsetting may
encourage clearing in more fragmented landscapes and offsetting in
more intact landscapes, which has potential to exacerbate the loss
GIBBONS ET AL.
|
9
of already overcleared ecosystems if the policy does not restrict off-
sets to the same ecosystem types as those impacted.
Finally, we detected no measurable change in land clearing
rates despite the aim of the policy “to prevent broad-scale clearing
unless it improves or maintains environmental outcomes.”In a
wide-ranging review, the Millennium Ecosystem Assessment (2005)
concluded that “ecosystem degradation can rarely be reversed
without actions that address one or more indirect drivers of
change.”That is, using regulation such as biodiversity offsetting
alone to stop the loss of biodiversity is unlikely to be effective. A
policy that seeks to achieve no net loss of biodiversity must iden-
tify ways that ongoing demands from population growth and eco-
nomic growth can be met without impacting on biodiversity. This
could be in the form of a no net loss policy introduced alongside
policies promoting more intensive use of existing land (e.g. urban
infill) (Villase~
nor, Tulloch, Driscoll, Gibbons, & Lindenmayer, 2017),
increasing income from primary production or resource extraction
by means other than increasing production (e.g. value-adding) and/
or macroeconomic reform that transitions an economy away from
primary production or resource extraction (Angelsen & Kaimowitz,
1999; Kaimowitz, Byron, & Sunderlin, 1998; Rudel et al., 2005). For
changes such as these to occur, biodiversity offsetting policy must
be introduced as part of holistic reforms across government rather
than in isolation of those parts of society and the economy that
represent the ultimate drivers of biodiversity loss.
ACKNOWLEDGEMENTS
We thank several officers in the New South Wales Office of Envi-
ronment and Heritage for providing data and fielding questions
about these data. We thank Peter Burnett at The Australian National
University for commenting on a draft of the manuscript. PG
acknowledges support from the Nagoya University, Japan for sup-
porting him as a visiting fellow when much of this research was
completed. The graphical summary was created by Clive Hilliker. The
authors have no conflicts of interest to declare.
REFERENCES
Angelsen, A., & Kaimowitz, D. (1999). Rethinking the causes of deforesta-
tion: Lessons from economic models. The World Bank Research Obser-
ver,14,73–98. https://doi.org/10.1093/wbro/14.1.73
Apostolopoulou, E., & Adams, W. M. (2015). Biodiversity offsetting and
conservation: Reframing nature to save it. Oryx,51,1–9.
BBOP (2012a) Resource paper: No net loss and loss-gain calculations in bio-
diversity offsets. Washington, D.C.: Business and Biodiversity Offsets
Programme.
BBOP (2012b). Standard on biodiversity offsets. Washington, D.C.: Busi-
ness and Biodiversity Offsets Program.
Brown, P. H., & Lant, C. L. (1999). The effect of wetland mitigation bank-
ing on the achievement of no-net-loss. Environmental Management,
23, 333–345. https://doi.org/10.1007/s002679900190
Bull, J., Gordon, A., Law, E., Suttle, K., & Milner-Gulland, E. (2014). Impor-
tance of baseline specification in evaluating conservation interven-
tions and achieving no net loss of biodiversity. Conservation Biology,
28, 799–809. https://doi.org/10.1111/cobi.2014.28.issue-3
Bureau of Rural Sciences (1998). Australia’s the state of the forests report.
Canberra: Bureau of Rural Sciences.
Byron, N., Craik, W., Keniry, J., & Possingham, H. (2014). A review of bio-
diversity legislation in NSW. Final report. Sydney: State of New South
Wales and Office of Environment and Heritage.
Cunningham, R. B., Lindenmayer, D. B., Crane, M., Michael, D. R., Barton,
P. S., Gibbons, P.,... Stein, J. A. (2014). The law of diminishing
returns: Woodland birds respond to native vegetation cover at multi-
ple spatial scales and over time. Diversity and Distributions,20,59–
71. https://doi.org/10.1111/ddi.2013.20.issue-1
Curran, M., Hellweg, S., & Beck, J. (2014). Is there any empirical support
for biodiversity offset policy? Ecological Applications,24, 617–632.
https://doi.org/10.1890/13-0243.1
Curran, M., Hellweg, S., & Beck, J. (2015). The jury is still out on biodi-
versity offsets: Reply to Qu
etier et al.. Ecological Applications,25,
1741–1746. https://doi.org/10.1890/14-2363.1
Department of Climate Change and Energy Efficiency (undated). Drivers
of land clearing in Australia. Canberra: Australian Government Depart-
ment of Climate Change and Energy Efficiency.
Department of Conservation (2014). Guidance of good practice biodiversity
offsetting in New Zealand. New Zealand Department of Conservation:
Unpublished.
Department of Environment Climate Change and Water (2011). Opera-
tional manual for BioMetric 3.1. Sydney: New South Wales Depart-
ment of Environment, Climate Change and Water.
Department of Environment Climate Change and Water Nsw (2011).
Operational manual for BioMetric 3.1. Sydney: Department of Environ-
ment, Climate Change and Water.
Evans, M. C. (2016). Deforestation in Australia: Drivers, trends and policy
responses. Pacific Conservation Biology,22, 130–150. https://doi.org/
10.1071/PC15052
Ferraro, P. J., & Pattanayak, S. K. (2006). Money for nothing? A call for
empirical evaluation of biodiversity conservation investments. PLoS
Biology,4, e105. https://doi.org/10.1371/journal.pbio.0040105
Fromond, L., Simil€
a, J., & Suvantola, L. (2009). Regulatory innovations for
biodiversity protection in private forests—towards flexibility. Journal
of Environmental Law,21,1–31.
Gardner, T. A., Hase, A., Brownlie, S., Ekstrom, J. M., Pilgrim, J. D., Savy,
C. E.,... Ten Kate, K. (2013). Biodiversity offsets and the challenge of
achieving no net loss. Conservation Biology,27, 1254–1264. https://d
oi.org/10.1111/cobi.12118
Gibbons, P., Briggs, S., Ayers, D., Seddon, J., Doyle, S., Cosier, P.,...
Roberts, K. (2009). An operational method to assess impacts of land
clearing on terrestrial biodiversity. Ecological Indicators,9,26–40.
https://doi.org/10.1016/j.ecolind.2008.01.006
Gibbons, P., Briggs, S. V., Zerger, A., Ayers, D., Seddon, J., & Doyle, S.
(2007). Assessing the biodiversity value of stands and patches in a
landscape context. In D. B. Lindenmayer & R. J. Hobbs (Eds.), Manag-
ing and designing landscapes for conservation (pp. 215–228). Oxford:
Blackwell Publishing and Zoological Society of London.
Gibbons, P., Evans, M. C., Maron, M., Gordon, A., Roux, D., Hase, A.,...
Possingham, H. P. (2016). A loss-gain calculator for biodiversity off-
sets and the circumstances in which no net loss is feasible. Conserva-
tion Letters,9, 252–259. https://doi.org/10.1111/conl.2016.9.issue-4
Gordon, A., Bull, J. W., Wilcox, C., & Maron, M. (2015). FORUM: Perverse
incentives risk undermining biodiversity offset policies. Journal of Applied
Ecology,52,532–537. https://doi.org/10.1111/1365-2664.12398
Habib, T. J., Farr, D. R., Schneider, R. R., & Boutin, S. (2013). Economic
and ecological outcomes of flexible biodiversity offset systems. Con-
servation Biology,27, 1313–1323. https://doi.org/10.1111/cobi.
12098
Harper, D. J., & Quigley, J. T. (2005). A comparison of the areal extent of
fish habitat gains and losses associated with selected compensation
projects in Canada. Fisheries,30,18–25. https://doi.org/10.1577/
1548-8446(2005)30[18:ACOTAE]2.0.CO;2
10
|
GIBBONS ET AL.
Hilderbrand, R. H., Watts, A. C., & Randle, A. M. (2005). The myths of
restoration ecology. Ecology and Society,10(1), 19 (online). Retrieved
from http://www.ecologyandsociety.org/vol10/iss1/art19/
Jenner, N., & Balmforth, Z. (2015). Biodiversity offsets: Lessons learnt from
policy and practice. Country summary report: South Africa. Cambridge,
UK: Business and Biodiversity Offset Programme and Flora and
Fauna International.
Kaimowitz, D., Byron, N., & Sunderlin, W. (1998) Public policies to
reduce inappropriate tropical deforestation. Agriculture and the envi-
ronment: Perspectives on sustainable rural development.
Laitila, J., Moilanen, A., & Pouzols, F. M. (2014). A method for calculating
minimum biodiversity offset multipliers accounting for time discount-
ing, additionality and permanence. Methods in Ecology and Evolution,
5, 1247–1254. https://doi.org/10.1111/2041-210X.12287
Landsberg, J. J. (1993). Rural dieback and insect damage in remnants of
native woodlands. Victorian Naturalist,110(1), 37.
Lehmann, E. A., Wallace, J. F., Caccetta, P. A., Furby, S. L., & Zdunic, K.
(2013). Forest cover trends from time series Landsat data for the
Australian continent. International Journal of Applied Earth Observation
and Geoinformation,21, 453–462. https://doi.org/10.1016/j.jag.2012.
06.005
Macintosh, A. (2014). Mitigation targets, burden sharing and the role of
economic modeling in climate policy. Australian Journal of Public
Administration,73, 164–180. https://doi.org/10.1111/1467-8500.
12074
Maron, M. (2015). Stop misuse of biodiversity offsets. Nature,523, 401.
https://doi.org/10.1038/523401a
Maron, M., Bull, J. W., Evans, M. C., & Gordon, A. (2015). Locking in loss:
Baselines of decline in Australian biodiversity offset policies. Biologi-
cal Conservation,192, 504–512. https://doi.org/10.1016/j.biocon.
2015.05.017
Maron, M., Hobbs, R. J., Moilanen, A., Matthews, J. W., Christie, K.,
Gardner, T. A.,... McAlpine, C. A. (2012). Faustian bargains? Restora-
tion realities in the context of biodiversity offset policies. Biological
Conservation,155, 141–148. https://doi.org/10.1016/j.biocon.2012.
06.003
Maron, M., Ives, C. D., Kujala, H., Bull, J. W., Maseyk, F. J., Bekessy, S.,...
Possingham, H.P. (2016). Taming a wicked problem: Resolving con-
troversies in biodiversity offsetting. BioScience,66, 489–498.
https://doi.org/10.1093/biosci/biw038
Maseyk, F., Barea, L. P., Stephens, R. T. T., Possingham, H. P., Dutson, G.,
Maron, M. (2016). A disaggregated biodiversity offset accounting
model to improve estimation of ecological equivalency and no net
loss. Biological Conservation,204, 322–332.
May, J., Hobbs, R. J., & Valentine, L. E. (2017). Are offsets effective? An
evaluation of recent environmental offsets in Western Australia. Bio-
logical Conservation,206, 249–257. https://doi.org/10.1016/j.biocon.
2016.11.038
Mcdonald-Madden, E., Baxter, P. W. J., & Possingham, H. P. (2008). Mak-
ing robust decisions for conservation with restricted money and
knowledge. Journal of Applied Ecology,45, 1630–1638. https://doi.
org/10.1111/jpe.2008.45.issue-6
Mckenney, B. A., & Kiesecker, J. M. (2010). Policy development for
biodiversity offsets: A review of offset frameworks. Environmental
Management,45, 165–176. https://doi.org/10.1007/s00267-009-
9396-3
Millennium Ecosystem Assessment (2005). Ecosystems and human well-
being: Synthesis. Washington, DC: Island Press.
Miller, K. L., Trezise, J. A., Kraus, S., Dripps, K., Evans, M. C., Gibbons,
P.,... Maron, M. (2015). The development of the Australian environ-
mental offsets policy: From theory to practice. Environmental Conser-
vation,42, 306–314. https://doi.org/10.1017/S037689291400040X
Moilanen, A., Teeffelen, A. J. A. V., Ben-Haim, Y., & Ferrier, S. (2009).
How much compensation is enough? A framework for incorporating
uncertainty and time discounting when calculating offset ratios for
impacted habitat Restoration Ecology,17, 470–478. https://doi.org/
10.1111/rec.2009.17.issue-4
Moreno-Mateos, D., Maris, V., B
echet, A., & Curran, M. (2015). The true
loss caused by biodiversity offsets. Biological Conservation,192, 552–
559. https://doi.org/10.1016/j.biocon.2015.08.016
New South Wales Government Office of Environment and Heritage
(2013). Native vegetation regulation 2013: Environmental outcomes
assessment methodology. Sydney: Office of Environment and Her-
itage.
NOAA (2002). Habitat equivalency analysis: An overview. Washington,
D.C.: National Oceanic and Atmospheric Administration.
Norton, D. A. (2009). Biodiversity offsets: Two New Zealand case studies
and an assessment framework. Environmental Management,43, 698–
706. https://doi.org/10.1007/s00267-008-9192-5
Office of Environment and Heritage for New South Wales(2014). Frame-
work for Biodiversity Assessment. Sydney: State of NSW and Office of
Environment and Heritage.
Office of Environment and Heritage of the New South Wales Govern-
ment (2014). New South Wales Annual Report on Native Vegetation
2011–13. Sydney: Office of Environment and Heritage.
Ozolins, A., Brack, C., & Freudenberger, D. (2001). Abundance and
decline of isolated trees in the agricultural landscapes of Central New
South Wales. Pacific Conservation Biology,7, 195–203. https://doi.
org/10.1071/PC010195
Parkes, D., Newell, G., & Cheal, D. (2003). Assessing the quality of native
vegetation: The ‘habitat hectares’approach. Ecological Management and
Restoration,4, S29–S38. https://doi.org/10.1046/j.1442-8903.4.s.4.x
Pickett, E. J., Stockwell, M. P., Bower, D. S., Garnham, J. I., Pollard, C. J.,
Clulow, J., & Mahony, M. J. (2013). Achieving no net loss in habitat
offset of a threatened frog required high offset ratio and intensive
monitoring. Biological Conservation,157, 156–162. https://doi.org/10.
1016/j.biocon.2012.09.014
Queensland Department of Science, Information, Technology and Innova-
tion (2015). Land cover change in Queensland 2012–13 and 2013–14:
A Statewide Landcover and Trees Study (SLATS) report. Brisbane:
Queensland Department of Science, Information Technology and
Innovation.
Qu
etier, F., Regnery, B., & Levrel, H. (2014). No net loss of biodiversity
or paper offsets? A critical review of the French no net loss policy.
Environmental Science & Policy,38, 120–131. https://doi.org/10.
1016/j.envsci.2013.11.009
Qu
etier, F., Van Teeffelen, A., Pilgrim, J., Von Hase, A., & Ten Kate, K.
(2015). Biodiversity offsets are one solution to unmitigated biodiver-
sity loss–a response to Curran et al.. Ecological Applications,25,
1739–1741. https://doi.org/10.1890/14-1217.1
R Development Core Team (2010). R: A language and environment for sta-
tistical computing. Vienna, Austria: R Foundation for Statistical Com-
puting.
Radford, J. Q., Bennett, A. F., & Cheers, G. J. (2005). Landscape-level
thresholds of habitat cover for woodland-dependent birds. Biological
Conservation,124, 317–337. https://doi.org/10.1016/j.biocon.2005.
01.039
Rubec, C. D., & Hanson, A. R. (2009). Wetland mitigation and compensa-
tion: Canadian experience. Wetlands Ecology and Management,17,3–
14. https://doi.org/10.1007/s11273-008-9078-6
Rudel, T. K., Coomes, O. T., Moran, E., Achard, F., Angelsen, A., Xu, J., &
Lambin, E. (2005). Forest transitions: Towards a global understanding
of land use change. Global Environmental Change,15,23–31.
https://doi.org/10.1016/j.gloenvcha.2004.11.001
Rundcrantz, K., & Sk€
arb€
ack, E. (2003). Environmental compensation in
planning: A review of five different countries with major emphasis on
the German system. European Environment,13, 204–226. https://doi.
org/10.1002/eet.v13:4
Sudol, M. F., & Ambrose, R. F. (2002). The US Clean Water Act and habi-
tat replacement: Evaluation of mitigation sites in Orange County,
GIBBONS ET AL.
|
11
California, USA. Environmental Management,30, 0727–0734.
https://doi.org/10.1007/s00267-002-2787-3
Ten Kate, K., Pilgrim, J., Brooks, T., Gibbons, P., Hughes, J., Mackey,
B.,... Watson, J. (2014). Biodiversity Offsets Technical Study Paper.
Gland: IUCN.
Tischew, S., Baasch, A., Conrad, M. K., & Kirmer, A. (2010). Evaluating
restoration success of frequently implemented compensation mea-
sures: Results and demands for control procedures. Restoration Ecol-
ogy,18, 467–480.
Villase~
nor, N. R., Tulloch, A. I., Driscoll, D. A., Gibbons, P., & Lindenmayer,
D. B. (2017). Compact development minimizes the impacts of urban
growth on native mammals. Journal of Applied Ecology,54, 794–804.
https://doi.org/10.1111/1365-2664.12800
Walker, S., Brower, A. L., Stephens, R., & Lee, W. G. (2009). Why barter-
ing biodiversity fails. Conservation Letters,2, 149–157. https://doi.
org/10.1111/conl.2009.2.issue-4
Weinberg, A., Gibbons, P., Briggs, S. V., & Bonser, S. P. (2011). The
extent and pattern of Eucalyptus regeneration in an agricultural land-
scape. Biological Conservation,144, 227–233. https://doi.org/10.
1016/j.biocon.2010.08.020
How to cite this article: Gibbons P, Macintosh A, Constable
AL, Hayashi K. Outcomes from 10 years of biodiversity
offsetting. Glob Change Biol. 2017;00:1–12. https://doi.org/
10.1111/gcb.13977
12
|
GIBBONS ET AL.
