Valuing the Benefits of Forests

Final Report

MPI Technical Paper No: 2017/68

Prepared for Elizabeth Heeg

Ministry for Primary Industries
Pastoral House 25 The Terrace
PO Box 2526
Wellington 6140

by Patrick Walsh, Landcare Research; Tarek Soliman, Landcare

Research; Suzie Greenhalgh, Landcare Research; Norman Mason,
Landcare Research; and David Palmer, Palmer Consulting

ISBN No: 978-1-77665-738-4 (online)

ISSN No: 2253-3923 (online)

October 2017
Disclaimer
While every effort has been made to ensure the information in this publication is accurate, the
Ministry for Primary Industries does not accept any responsibility or liability for error of fact,
omission, interpretation or opinion that may be present, nor for the consequences of any
decisions based on this information.

Requests for further copies should be directed to:

Publications Logistics Officer

Ministry for Primary Industries
PO Box 2526
WELLINGTON 6140

Email: [email protected]
Telephone: 0800 00 83 33
Facsimile: 04-894 0300

This publication is also available on the Ministry for Primary Industries website at
http://www.mpi.govt.nz/news-and-resources/publications/

© Crown Copyright - Ministry for Primary Industries

Reviewed by: Approved for release by:

Florian Eppink Pike Brown

Senior Economist Research Priority Area Leader – Supporting
Landcare Research Business & Policy
Landcare Research

Landcare Research Contract Report: LC2788

Contents Page

Executive Summary 1

1 Introduction 2

2 National analysis 3

3 Manawatū catchment analysis 21

4 Summary of benefits and costs in the Manawatū Catchment 41

5 Works cited 48

i
Executive Summary
This report is the product of a Ministry for Primary Industries (MPI) commissioned project to
explore the benefits of permanent forests compared with plantation forests and other land
uses. The project is related to the Permanent Forest Sink Initiative (PFSI), which was
established in 2008 and acts as a mechanism for landowners to access the value of carbon
sequestration in exchange for establishing a forest sink covenant on their land. New Zealand
has a variety of national and regional priorities associated with afforestation, which can assist
the country in achieving its domestic and international climate goals. There are also a range
of other related benefits, including water quality improvement, native species habitat, and
other ecosystem services.

MPI commissioned Landcare Research NZ to perform several analyses related to

afforestation in erosion-prone pastoral hill country. A range of quantitative geographic,
biological, and economic models were integrated, with several related qualitative
assessments. Following the specifications set out by MPI, the report first conducts a broad
analysis at the national level, and then performs a more detailed analysis in the Manawatū
catchment in the lower North Island. There are three afforestation scenarios evaluated in both
the national and Manawatū analyses. The first scenario assumes that land identified for
afforestation is converted to exotic pine plantations. The other two scenarios assume that the
land is converted to indigenous forest, with one scenario maintaining permanent indigenous
forest with no production activities and the other allowing some productive uses of the land,
e.g. honey production.

The national-level analysis estimates the opportunity cost of the land use conversion to
permanent forest, in terms of lost agricultural revenue and land value. In addition, the
increased revenue from the afforestation scenarios (if applicable) was estimated alongside the
change in greenhouse gas emissions.

In the Manawatū analysis, opportunity costs and increased revenue from the afforestation
scenarios were estimated using a similar approach to the national level analysis. Carbon
values, however, were estimated using a time path of forest growth. Water quality benefits
were also monetized. A 50-year time horizon was used to estimate the net present value for
these scenarios for the Manawatū catchment. In addition to these monetised values, changes
in biodiversity values were quantified and an ecosystem services framework was used to
quality describe the impacts of the afforestation scenarios on a range of ecosystem services.

Even under the most conservative policy assumptions, whereby the full value of afforested
land is included as an opportunity cost, two out of three afforestation scenarios had a positive
net present value across 50 years in the Manawatū catchment. When an alternate policy is
used, and the full value of land is not included, all three afforestation scenarios have positive
net benefits, with the highest benefit cost ratio at 9.2.

The monetised benefits and costs analysis would suggest that exotic afforestation would be
the best approach to pursue given it has the highest NPV and benefit-cost ratio. However,
there were many benefits that could not be monetized, including biodiversity, recreation,
cultural values, and broader impacts on ecosystem services.

Several important limitations with current data and models were encountered in the course of
this analysis and these limitations are outlined to illustrate where additional analysis could be
undertaken in the future.

Ministry for Primary Industries Valuing the Benefits of Forests  1

1 Introduction
This report is the product of a Ministry for Primary Industries (MPI) commissioned project to
explore the benefits of permanent forests compared with plantation forests and other land
uses. The project is related to the Permanent Forest Sink Initiative (PFSI), which was
established in 2008 and acts as a mechanism for landowners to access the value of carbon
sequestration in exchange for establishing a forest sink covenant on their land. However, to
be more widely applicable, the analysis does not fully adhere to the conditions required under
the PFSI. New Zealand has a variety of national and regional priorities associated with
afforestation, which can assist the country in achieving its domestic and international climate
goals. There are also a range of other related benefits, including water quality improvement,
native species habitat, and other ecosystem services.

With carbon prices increasing, PFSI forest or alternative afforestation programs could
represent the most economically and environmentally viable land use for some of New
Zealand’s most erosion-prone pastoral hill country. To help inform discussions surrounding
the PFSI and these erosion-prone areas, the current report explores several scenarios related
to those lands. The main goals for this report, as laid out by MPI, are to determine the:

– additional forest land cover across NZ that could be realised and the proportion that
would most suit long-term forest cover, along with an estimate of the opportunity
costs;
– and the economic, environmental, and sustainable land management co-benefits of
permanent forest plantings over a period of 50 years compared with benefits of
plantation forestry and other land uses using an ecosystem services approach. Where
possible, these benefits are monetised to enable a cost-benefit ratio to be determined
for the different types of land uses.

MPI commissioned Landcare Research to perform these analyses, which are summarized in
this report. To achieve these goals, we integrate quantitative geographic, biological, and
economic models, with several related qualitative assessments. Following the specifications
set out by MPI, the report is structured around the following central components:

1. Identification of potential new areas suitable for afforestation across New Zealand.
2. Catchment analysis of ecosystem services of different land uses.
3. Non-market valuation of ecosystem services and cost-benefit analysis.

The main analysis in this paper takes place in Section 2 in the Manawatū Catchment, where
two policy scenarios are analysed across 50 years. In that analysis a variety of ecosystem
services are explored in a spatially explicit manner, with several monetized or quantified.
However, to give a rough idea of potential impacts at the national level, we also present a
national analysis in Section 1. The national analysis is presented for a single year instead of
50, where we assume that the policy is fully implemented (so no transition path is presented).
There are therefore some differences in the methods between the 50 year analysis and the one
year analysis.

In the national analysis (Section 2) we identify marginal lands across New Zealand that could
be converted to long-term forest cover. These afforestation areas were identified using
several geospatial maps and models from previous New Zealand literature. Throughout this
report these additional marginal land areas identified for afforestation are referred to as ‘new
afforestation areas.’ However, in some cases the identified areas will not be new. For

2  Valuing the Benefits of Forests Ministry for Primary Industries

example, some identified marginal lands already have exotic or indigenous forestry on them.
In that case, those lands will be assumed to remain in their current (forest) state. 1

As directed by MPI, two main policy scenarios were developed, which are applied
throughout this report:

 Scenario E assumes that exotic plantation forestry is planted on the new afforestation
areas While this rotational plantation model would not be eligible for PFSI registration,
the inclusion of this scenario provides an alternative ‘forestry’ land use, which will
deliver some of the benefits of forest with different rates of return, including additional
forestry-related revenue.
 Scenario I assumes that the new afforestation areas are afforested with permanent, non-
rotation indigenous forest.
– In order to capture alternate production methods, Scenario I is further subdivided:
o Scenario Ia assumes the new afforestation areas are removed from production and
converted to indigenous forest.
o Scenario Ib assumes the land suitable for mānuka/kānuka in the new afforestation
areas remain in mānuka/kānuka and are used for enterprises such as medical or
commercial honey production.

After identifying the new afforestation areas, Section 2 identifies the areas of indigenous
forest that would require active planting versus passive afforestation. In order to estimate
productive potential of the indigenous forest areas, the areas suitable for mānuka/kānuka are
identified. The opportunity costs of the afforestation scenarios can then be estimated, as well
as the annual carbon-related benefits.

Section 3 analyses the afforestation scenarios in more detail for the Manawatū catchment in
the Manawatū-Wanganui Region of the North Island. The majority of the land area in the
Manawatū catchment is covered by pasture, as much of the previous indigenous forest was
cleared for farming over the last 150 years or so. This has caused erosion and other problems
on steep slopes in the area. There is also significant variation in current land use, serving as a
good setting for this case study. In order to identify, at a finer resolution, the additional land
suitable for afforestation, areas for passive regeneration, and areas suitable for
mānuka/kānuka, similar analyses as Section 2 are undertaken in the Manawatū Catchment.
The analysis in Section 3 goes into greater detail for the economic analysis and uses an
ecosystem services approach to classify the estimated costs and benefits. For the Manawatū
analysis, benefits and costs are projected over a 50 year period.

2 National analysis
This section focuses on national economic and environmental impacts of the afforestation
scenarios. GIS maps and empirical models are first used to identify the marginal pasture land
across New Zealand that is suitable for afforestation. Some of the new afforestation areas
may not naturally revert to indigenous forest because of distance to other native forest which

1
This initial identification of marginal land assessment identifies areas likely to be suitable for afforestation and
earning carbon credits. However, there are particular conditions for eligibility under the PFSI, the Emissions
Trading Scheme or other carbon schemes which may not necessarily be met. This limitation applies to all
analysis within this report.

Ministry for Primary Industries Valuing the Benefits of Forests  3

would act as a seedbank. Therefore, we identify areas that would require active planting for
Scenario I and those areas that would passively revert. For Scenario Ib, we also identify areas
suitable for mānuka/kānuka, as these areas could have a productive use, e.g. for honey
production.

The focus of the national analysis is on the annual costs and benefits from a representative
year where the policy scenario is fully implemented. The two policy scenarios are compared
to the baseline land uses, assuming that the current land use is roughly representative of what
may have happened in the absence of the policy. A more detailed analysis at the national
level would require significantly more modelling, data, and analysis beyond the scope of this
project.

Using the New Zealand Forest and Agriculture Regional Model (NZFARM; Daigneault et al.
2016) and several other inputs, we are able to estimate the opportunity costs related to these
afforestation scenarios. As a reminder, in Scenario E all new afforestation areas are converted
to exotic plantation forestry. In scenario I, all new afforestation areas are converted into
indigenous forest. The opportunity cost contains two main inputs – (1) the value of the
earnings in the previous land use, and (2) the value of the converted land. We also calculate
the carbon-related benefits of these scenarios using several different valuation methods.
Although these carbon-related benefits are likely some of the larger impacts, it is important to
recognize that they are only one of a range of potential ecosystem service benefits. This
broader range of ecosystem service benefits can be difficult to quantify and monetize,
particularly at the national level. The reasons for this are explained in more depth in the
Manawatū case study.

2.1 Generation of national maps of areas most suitable for plantation forestry
and most likely indigenous species
To identify areas of marginal land most suitable for afforestation, we drew heavily from maps
produced by Watt et al. (2011) (a previous project funded by the SLMACC programme2).
This earlier work identified three potential options for afforestation, depending on
assumptions about land suitability which was based on Land Use Capability (LUC) classes. 3
LUC classes 5-8 are designated as unsuitable for arable land use, and have slight (LUC 5),
moderate (6), severe (7), and extreme (8) erosion. The afforestation scenarios developed in
Watt et al (2011) focus mostly on differences in land in LUC 6, where forestry is a suitable
land use to control erosion. Each scenario also includes lands from LUC 7 and 8 that would
be suitable for forest.

The most conservative scenario option in Watt et al. (2011) assumed that only the least
versatile land classes with severe to extreme erosion would be afforested, the second option
included lands with moderate to extreme erosion, and the third included slight to extreme
erosion area. Using the most recent land cover map (2012–2013), we have reproduced these
three options. For all our analysis below, we use the first scenario of Watt et al. (2011), which
represents the most conservative assumptions on areas suitable for afforestation – and has the

2
Sustainable Land Management and Climate Change (SLMACC) Research Programme:
https://www.mpi.govt.nz/funding-and-programmes/farming/sustainable-land-management-and-climate-change-
research-programme/
3
For more information on LUC’s, see http://www.landcareresearch.co.nz/publications/books/luc. A GIS map of
LUC’s can be found here: https://lris.scinfo.org.nz/layer/76-nzlri-land-use-capability/.

4  Valuing the Benefits of Forests Ministry for Primary Industries

smallest afforestation area of the three options. That land is less likely to face competition for
other higher value agricultural uses, and it is therefore the least disruptive policy option.

To identify the new afforestation areas, high resolution national-level GIS datasets are used
that cover land use, terrain attributes, and a range of environmental variables. The primary
data source was the New Zealand Land Cover Database (LCDB), which is a digital map of
land cover created by grouping together similar classes that were identified from satellite
images.4

A map of additional areas suitable for afforestation, based on the conservative option
described in Watt et al. (2011) is found in Figure 1. Potential areas for afforestation appear in
black. A total of 695,566 ha of afforestation area were identified for our preferred option. Of
that area, approximately 26,000 ha are classified as “reserves.” Excluding the reserve land,
there are 669,966 ha available for afforestation. Of that area, 531,051 ha, or 79%, is on the
North Island, while 138,914 ha (22%) is on the South Island.5 To compare across the updates
of the different scenarios from Watt et al. (2011),
Table 1 illustrates the total area in each scenario, as well as the distribution across the North
and South islands.

4
The most recent version of the NZ LCDB, which was released across 2012/2013, can be found at:
https://lris.scinfo.org.nz/layer/423-lcdb-v41-land-cover-database-version-41-mainland-new-zealand/.
5
In comparison, the two less conservative options had 1,094,361 ha and 2,834,962 ha of land being afforested.

Ministry for Primary Industries Valuing the Benefits of Forests  5

Figure 1: New afforestation areas

Table 1: Updated afforestation scenarios from Watt et al. (2011)

Option 1 Option 2 Option 3

LUC 6 erosion Slight to
Severe to extreme Moderate to extreme
assumptions extreme
North Island Area (‘000
531 756 1,640
ha)
South Island Area (‘000
138 375 1,283
ha)
Total Area (‘000 ha) 670 1,131 2,923

6  Valuing the Benefits of Forests Ministry for Primary Industries

2.2 Identification of areas of passive regeneration and active afforestation for
indigenous areas
The previous section identified potential areas for new afforestation. In scenario E, these
areas will be actively planted and managed for Pinus radiata timber production. In Scenario
I, however, some areas are likely to revert naturally to indigenous species, while other areas
will require active management to convert to indigenous forest. The probability of natural
reversion depends on landscape and location-based factors. In this section we estimate which
of the new afforestation areas will passively revert as opposed to those areas which will
require active planting and management for indigenous afforestation. Mason et al. (2013)
provides the foundation for our analysis of passive versus active afforestation for indigenous
forest.

Mason et al. (2013) selected 10,061 plots from New Zealand’s National Vegetation Survey
Databank, and analyzed them to identify central environmental and land cover influences that
affect the occurrence probability of indigenous vegetation. Plot surveys were conducted
between 1982 and 2008. LCDB also played a central role in this effort, in addition to a range
of environmental and land cover GIS datasets. Mason et al. (2013) found that the most
important environmental variable was mean annual temperature, while the most predictive
land cover variables were local woody cover and distance to forest. We use GIS maps from
Mason et al. (2013), and link them with our other analyses and maps to identify the areas that
are more suitable for active or passive indigenous afforestation.6

Figure 2 shows the potential areas of indigenous afforestation, split into areas of active and
passive afforestation. The pink areas represent active afforestation, while the blue areas
indicate passive afforestation. The importance of temperature found by Mason et al. (2013) is
noticeable in the map, as colder areas at higher elevation fall more generally into the areas
that require active afforestation.

To summarize the content of Figure 2 in an alternative format, Table 2 illustrates breakdown

between active and passive areas in hectares (for the identified afforestation areas). Across
New Zealand, approximately 581,070 ha are in areas of passive afforestation, while 88,895
ha are in areas that require active afforestation. Although the actual success of natural
reversion depends on a variety of factors, as emphasized by Mason et al. (2013), areas of
passive afforestation would be expected to revert to natural forest with a high probability
after the land was no longer actively managed for its current use.

6
Note that the Mason et al. (2013) work produced a probability map describing the probability of passive
afforestation, based on a range of inputs. Based on recommendations from the authors of that work, we used a
probability cutoff of 0.6 to differentiate between the two areas. Greater than 0.6 was delineated as passive
afforestion areas while less than 0.6 was active afforestation areas.

Ministry for Primary Industries Valuing the Benefits of Forests  7

Table 2: Areas requiring active verus passive afforestation to indigenous species

Indigenous afforestation Total Passive Active

(ha) (ha) (ha)
North Island 531,051 480,429 50,622
South Island 138,914 100,641 38,273
Total 669,966 581,070 88,895

Figure 2: Potential areas of active and passive indigenous afforestation

8  Valuing the Benefits of Forests Ministry for Primary Industries

2.3 Mānuka and Kānuka suitability
In the two indigenous scenarios analyzed in this report, Scenario Ia and Ib differ by their
assumptions around the use of indigenous forest. Scenario Ia assumes all new afforestation
areas are not used for any production uses, whereas Scenario Ib assumes that some of that
land can be used for (non-harvested) production. For our analysis, we assume this productive
use relates to mānuka/kānuka-related products such as honey. For Scenario Ib, we also
estimate where mānuka/kānuka is more likely to occur, and changes the opportunity cost
estimations between Scenarios Ia and Ib.

We adapted the work of Watt et al. (2012) as they included a predictive model of those two
species. The growth of mānuka/kānuka stands were modelled using a physiologically based
growth model (CenW 4.0) that explored the relationship of mānuka/kānuka stands to
environmental conditions and other related factors. Mean annual temperature and
precipitation were identified as the best predictors, with mānuka/kānuka completely absent
from sites below 5⁰C. Therefore, we used the temperature and precipitation prediction
equations from Watt et al. (2012) to define the probability of occurrence of mānuka/kānuka
as a function of temperature and precipitation. 7

For temperature:

(1) ProbTemp(Occur) = 7.159 × (TempC) – 30.734

Where TempC is the mean annual temperature in degrees Celsius, and

ProbTemp(Occur) is the probability of occurrence, based on temperature.

For precipitation, where mean annual precipitation is below 1,550 mm/yr:

(2a) ProbPrecip(Occur) = 0.060 × Precip – 16.5

For annual precipitation greater than 1,550 mm/yr:

(2b) ProbPrecip(Occur) = –2(E–10) × Precip3 + 5(E–6) × Precip2 – 0.0418 × Precip +

119.57

In areas where the probability is based on precipitation and temperature, both of which are
greater than 50%, the land is designated as having the potential for mānuka/kānuka stands.
Table 3 outlines where mānuka/kānuka afforestation is likely to occur, where green areas are
suitable for mānuka/kānuka, while red areas are not. Approximately 52% of the new
afforestation area is suitable for mānuka/kānuka (Table 3), with a larger percent of the North
Island being suitable for mānuka/kānuka than the South Island.

7
Given the difficulty in predicting mānuka and kānuka, it is important to emphasize that this model may miss
other characteristics related to their growth, such as soil conditions.

Ministry for Primary Industries Valuing the Benefits of Forests  9

Table 3: Summary of afforestation areas and mānuka/kānuka suitability

Total new afforestation Area suitable for

Percent suitable for for
area mānuka/kānuka
mānuka/kānuka
(ha) (ha)
New Zealand 669,966 348,055 52
North Island 531,051 337,172 63
South Island 138,914 10,883 8

Figure 3: New afforestation areas suitable for mānuka/kānuka species.

10  Valuing the Benefits of Forests Ministry for Primary Industries

2.4 Opportunity cost of converting from existing land use to permanent forest
The opportunity cost for each scenario is estimated as:
 Scenario E: the new afforestation areas convert from pasture production to exotic
forestry, so the land remains in productive use. The opportunity cost is represented by the
profit from the previous land use.
 Scenario Ia: the new afforestation areas convert to indigenous forest. The land is
removed from production. The opportunity cost is composed of the profit from the
previous land use plus the value of the converted land.
 Scenario Ib: the new afforestation areas convert to indigenous forest with those areas
suitable for mānuka/kānuka being used for used for productive purposes. In this instance
for medical or edible honey production. The opportunity cost is composed of the profit
from the previous land use plus the value of the converted land. Due to the honey
production, there is less land opportunity cost than Scenario Ia.

To estimate the opportunity cost associated with changes in profitability, we employed the
NZFARM, along with related data on land value. First, the change in agricultural profits
resulting from Scenario E and Scenario I, which is a main component of opportunity cost, is
estimated. Agricultural profits are represented as average annual lost EBIT (earnings before
interest and taxes) for each scenario.

The agricultural EBIT opportunity cost estimates were derived using NZFARM. NZFARM
incorporates data and estimates from economic and land use databases and biophysical
models. The model is a non-linear, partial equilibrium, mathematical programming model of
New Zealand land use. Current land uses were derived from a national land use map (based
on AgriBase and the NZ Land Cover Database version 4 (LCDBv4)). Economic returns were
obtained by integrating several sources on farm budgets for relevant enterprises (Newsome et
al. 2008; MPI 2013a, b; Lincoln University Budget Manual 2013). Net farm revenues vary by
the farm/enterprise type, size, and location and can be aggregated by region and land use. The
model enables the comparison of EBIT between the current land use baseline and each
scenario.

The foregone EBIT from previous land uses is a part of the opportunity cost for all scenarios,
as they all assume that the same marginal land is converted from its existing use. 8 Where land
moves from a productive use to a non-productive use we assume the land is purchased from
the current land owner and retired from its productive use.9 This is the value of the land
portion of the opportunity cost estimation.

To estimate the value of the land opportunity cost, we use the median sales price per hectare
for farm land from the Real Estate Institute of New Zealand (REINZ). REINZ differentiates

8
There is, however, one important caveat. We assume that existing forest remains as it is. So in Scenario E,
which replaces marginal land with exotic plantation forestry, we assume the existing indigenous forest is not
converted. Likewise, for the indigenous Scenarios Ia and Ib, we assume that existing exotic forestry remains
plantation forestry.
9
Note that there are several alternative policy instruments that could effectively sideline land for afforestation.
For instance, placing a covenant on the land to dedicate it to forest. Our chosen policy instrument likely
represents a conservative upper bound on the potential opportunity cost of land, as the others will keep some
land under citizen ownership and potentially allow for other productive and related activities.

Ministry for Primary Industries Valuing the Benefits of Forests  11

land value regionally and by farm type. These estimates represent an upper bound, since the
profitability of the land is normally factored in its value. Therefore, profit (as represented by
lost EBIT) is not likely independent of the value of the land. In treating land value and
profitability as additive we assume that they are independent. To calculate the annual
opportunity cost, the value of land is annualized, so can be interpreted as foregone rent on the
land. 10

2.4.1 EBIT opportunity costs

Table 4 contains information about the current land uses in areas identified as the new
afforestation areas. The current land uses are dominated by sheep & beef (72%), native forest
(12%), and scrub (4.5%), while dairy, arable, and grapes represent less than 1% of the total
area. The majority of the sheep & beef farms are generating minor or no profits, while dairy,
fruits, vegetables, and grapes are generating the highest profits. Comparing the amount of
land to the net revenue figures, the more profitable enterprises are fruit, viticulture (grapes),
vegetables and dairy. Note that the minor difference between the total ha at that bottom of
Table 4 and the total afforestation area of 695,566 ha identified section 2.1 are due to slight
differences in the underlying GIS datasets.11

Table 4: Distribution of area and EBIT of current land use, by farm type
Current land use in new Corresponding Corresponding EBIT
Corresponding area
afforestation area agricultural EBIT per ha
(ha)
($NZ) ($NZ/ha)
Arable 37 61,750 1,669
Dairy 5,325 13,247,235 2,488
Deer 5,203 5,177,032 995
Exotic Forestry 20,357* 11,421,411* 561
Fruit 71 522,605 7,361
Native Forest 87,240* 0 0
Other 18,563 0 0
Other pasture 20,204 388,594 19
Pig 7 9,612 1,373
Scrub 31,175 0 0
Sheep & beef 501,992 31,088,779 62
Tussock 2,637 0 0
Vegetables 111 361,388 3,256
Grapes 1,920 9,224,887 4,805
Total 694,844 71,503,293
Total Without Exotic and Native 587,245 60,081,882*
* Note that we assume that in Scenario I, exotic forestry is not converted into indigenous forest. Similarly, in Scenario E,
native forest is not converted to exotic forestry.

10
So in the case of the producer owning the land, they are essentially “renting” to themselves.
11
Also, the approximately 26,000 ha of reserves discussed in Section 2.1 is classified as either “Native forest”
or “Other” here.

12  Valuing the Benefits of Forests Ministry for Primary Industries

2.4.2 Total opportunity costs by scenario
Table 5 presents the opportunity costs of converting existing land in the new afforestation
areas. The table lists these opportunity costs by region. As explained below, the shaded
columns represent the total opportunity costs for the three scenarios – E, Ia, and Ib.

For Scenario E, the total opportunity cost of converting is the lost EBIT, approximately $60
million. Since the land stays in production – as it is converted to exotic forestry – there is no
opportunity cost of the value of the land. This Scenario E total opportunity cost is displayed
in the second column, and is approximately $60 million. Note that this value differs from the
total of Table 4, as that table includes existing exotic forestry (which is not converted).

The third and fourth data columns outline the value of converted land for Scenario Ia and Ib,
respectively. These are estimated from the REINZ land value data mentioned above. The
total value of a property is the net present value of future returns, so to get annual values the
REINZ data are annualized. To annualize the data we use an 8% rate. There are several
potential interest rates that could be used. New Zealand Treasury recommends a 4-6% rate
for social benefit-cost analysis, although they previously recommended an 8% rate for the
social cost of capital. 12 The opportunity cost of Scenario Ib is less than Scenario Ia because
the proportion of the land assumed to stay in mānuka-related production (except for Otago
and Southland, which had no afforestation areas suitable for mānuka/kānuka) don’t have the
same opportunity cost of the value of converted land, since it is still in a productive use. The
final two columns contain the opportunity costs for Scenarios Ia and Ib, which are the sum of
lost EBIT and value of converted land.

It is important to note that the opportunity costs for Scenarios Ia and Ib, in reality, will
depend on the policy used to implement those scenarios. If the policy implemented involves
the government purchasing the land from the current owner, then the opportunity cost is as
illustrated in Table 5. However, if an alternative policy is deployed, such as placing
permanent forest covenants on the land, then the opportunity cost is less straightforward. In
this policy context, the original land owner may still be able to generate carbon credits from
the land, or other ecosystem services, and this portion of the opportunity cost is no longer the
full value of the land. As stressed earlier, the estimates in this report likely represent an upper
bound, since the profitability of the land is normally factored in its value, and hence the profit
(as represented by lost EBIT) is probably not independent from the value of the land. In
treating them as additive here, we assume they are independent.

12
In this report, we use a wider range of rates to illustrate the full range of potential estimates, so consistently
report values using 3% and 8%. The 4-6% rates currently recommended by Treasury should fit within this
range, while allowing for a wider range of estimates. Other countries frequently use 7%, such as the US and
Britain. http://www.treasury.govt.nz/publications/guidance/planning/costbenefitanalysis/currentdiscountrates.
The 8% used here for the social cost of capital should yield more conservative estimate of net benefits, since
many of the costs occur closer to the present than the benefits.

Ministry for Primary Industries Valuing the Benefits of Forests  13

Table 5: Annual opportunity costs by region ($NZ) of converting the new afforestation area

Region Lost EBIT + Lost EBIT +

Total afforested area Lost agricultural Value of converted Value of converted
converted land value converted land value
(ha) EBIT ($NZ) land Scenario Ia land Scenario Ib
Scenario Ia Scenario Ib
Scenario E Scenario Ia Scenario Ib
Opportunity Cost Opportunity Cost Opportunity Cost
Auckland Region 1,364 148,726 3,568,367 152,689 3,717,093 301,415
Bay of Plenty Region 5,385 2,103,301 37,541,507 5,722,285 39,644,808 7,825,586
Canterbury Region 71,755 4,509,952 73,185,872 72,081,232 77,695,824 76,591,184
Gisborne Region 104,254 1,044,412 148,737,670 33,377,089 149,782,082 34,421,501
Hawke's Bay Region 66,805 15,999,711 67,652,921 26,048,800 83,652,632 42,048,511
Manawatū-Wanganui Region 169,005 11,029,688 100,661,686 59,031,742 111,691,374 70,061,430
Marlborough Region 15,422 2,143,932 16,500,500 11,683,279 18,644,432 13,827,211
Nelson Region 1,122 77,753 1,451,508 545,750 1,529,261 623,503
Northland Region 4,449 261,838 4,338,391 108,156 4,600,229 369,994
Otago Region 29,616 2,244,047 38,717,878 38,717,878 40,961,925 40,961,925
Southland Region 3,693 784,665 7,686,860 7,686,860 8,471,525 8,471,525
Taranaki Region 21,620 2,714,981 24,996,834 1,035,058 27,711,815 3,750,039
Tasman Region 10,486 2,078,737 12,683,781 10,011,865 14,762,518 12,090,602
Waikato Region 29,807 5,392,166 63,053,003 10,380,983 68,445,169 15,773,149
Wellington Region 44,466 7,110,840 30,164,709 14,226,323 37,275,549 21,337,163
West Coast Region 7,997 2,437,132 5,209,318 4,047,616 7,646,450 6,484,748
Total 587,247 60,081,881 636,150,805 294,857,604 696,232,686 354,939,485

Ministry for Primary Industries Valuing the Benefits of Forests  14

2.5 Increases in exotic forestry earnings
In Scenario E the afforested land will be converted from its previous land use into exotic
forest production. Using NZFARM, we calculate the average annual EBIT expected from the
new forestry operations. In this part of the analysis, we are focusing on annual values. These
values are the estimated EBIT for forestry operations, assuming that the local area has the
proper infrastructure to support the increase in forestry, and assuming that local farmers are
willing to change their land use. The results of these estimates are outlined in Table 6. The
largest increase in EBIT is the Manawatū-Wanganui Region, while the smallest is the Nelson
area. An important caveat to these estimates is that the modelling uses current profitability;
they do not reflect any changes in market conditions that may be affected by the increase in
forested areas.

Table 6: Estimated national increase in forestry earnings, annual average

Region Gains from timber

($NZ)
Auckland Region 1,037,278
Bay of Plenty Region 3,333,665
Canterbury Region 19,178,675
Gisborne Region 61,909,180
Hawke's Bay Region 33,980,225
Manawatū-Wanganui Region 100,972,455
Marlborough Region 5,546,250
Nelson Region 364,031
Northland Region 3,448,134
Otago Region 7,638,856
Southland Region 1,411,443
Taranaki Region 14,220,458
Tasman Region 3,394,194
Waikato Region 18,434,312
Wellington Region 24,800,276
West Coast Region 2,892,839
Total 302,562,270

Table 7 presents the change in EBIT from converting to exotic forestry by previous land use
instead of region. Note we assume that the areas of indigenous forest are not converted to
exotic forestry operations, and the existing exotic plantations remain as they were.

Ministry for Primary Industries Valuing the Benefits of Forests  15

Table 7: Estimated increase in forestry earnings, annual average by previous land use

Enterprise Total afforested area Gains from timber

(Ha) ($NZ EBIT)
Arable 37 20,196
Dairy 5,325 3,012,997
Deer 5,203 2,252,974
Exotic forestry 0 0
Fruit 71 40,217
Native forestry 0 0
Other 18,563 0
Other pasture 20,204 11,005,215
Pig 7 4,210
Scrub 31,175 9,865,897
Sheep & beef 501,992 274,590,924
Tussock 2,637 800,010
Vegetables 111 38,412
Grapes 1,920 931,218
Total 587,245 302,562,270

2.6 Changes in carbon

Reductions in greenhouse gasses can yield a wealth of benefits, most importantly human
health benefits, but also including aesthetic and habit impacts (US EPA, 2014). NZFARM
provides outputs related to greenhouse gas emissions and carbon sequestration, which can be
used to identify some of those benefit categories. For both scenarios E and I, we calculate the
additional carbon sequestration in the new afforestation areas, as well as the greenhouse gas
(GHG) emissions avoided from moving from the previous land use to a forest use. These
represent the average carbon stock change as the forest reaches a steady state of carbon under
the new afforestation scenarios, where the new scenario is assumed to be fully implemented,
and the annual values depict a representative year.13

13
At the national level, we do not go into detail about the transition path to these new steady state values. In the
later case study on the Manawatu, we more explicitly estimate the transition from the existing land use to the
new afforestation areas.

16  Valuing the Benefits of Forests Ministry for Primary Industries

Table 8 outlines the avoided GHG emissions and carbon sequestered in the new afforestation
areas for Scenario E. The avoided GHG emissions are the emissions that would have
occurred if the land remained in its previous use, while the carbon sequestration is from the
new plantings of exotic forestry.

Ministry for Primary Industries Valuing the Benefits of Forests  17

Table 8: Annual GHG emissions and carbon sequestration for Scenario E (exotic forest)

Total GHG avoided and

GHG avoided Carbon sequestration
Land use carbon sequestered
(tons CO2e) (tons CO2e)
(tons CO2e)
Arable 62 429 491
Dairy 10,969 61,004 71,973
Deer 879 59,604 60,483
Forestry 0 0 0
Fruit 41 808 849
Native 0 0 0
Other 0 212,653 212,653
Other pasture 17,550 231,451 249,001
Pig 18 86 104
Sheep & beef 120,006 5,750,587 5,870,593
Vegetables 31 1,266 1,297
Grapes 1,313 21,993 23,306
Total 150,869 6,339,881 6,490,750

18  Valuing the Benefits of Forests Ministry for Primary Industries

Table 9 outlines the results of Scenario I where the new afforestation areas are converted to
indigenous forest. Based on the forestry literature, indigenous forest sequesters less carbon
than exotic plantation forestry. This is mainly due to the higher density and volume of timber
in exotic plantations as compared to indigenous forest, among other things.14

14
See, for example, the NZ ETS lookup table estimates of carbon storage for several species across time:
http://www.mfe.govt.nz/climate-change/reducing-greenhouse-gas-emissions/new-zealand-emissions-trading-
scheme

Ministry for Primary Industries Valuing the Benefits of Forests  19

Table 9: GHG emissions and carbon sequestration for Scenario I (indigenous forest)

Total GHG avoided and

GHG avoided Carbon sequestration
Land use carbon sequestered
(tons CO2e) (tons CO2e)
(tons CO2e)
Arable 62 22 84
Dairy 10,969 3,195 14,164
Deer 879 3,122 4,001
Forestry 0 0 0
Fruit 41 42 83
Native 0 0 0
Other 0 11,138 11,138
Other pasture 17,550 12,123 29,673
Pig 18 4 23
Sheep & beef 120,006 301,195 421,201
Vegetables 31 66 97
Grapes 1,313 1,152 2,465
Total 150,869 332,061 482,929

There are several assumptions that underpin this analysis. To reflect current production, the
NZFARM model assumes that the forest area was planted pre-1990. This may induce some
uncertainty with the new afforestation areas, since they will be new plantings. To annualize
profitability estimates, the model also assumes a small harvest each year instead of one large
one at the end of a rotation (see Daigneault et al. 2016 for more details). These assumptions
may produce some differences with the estimates for the Manawatū catchment presented
below, which better account for the time path of forest growth.

The reduction in net GHG emissions can be valued using several different methods. GHG
emissions have diverse impacts. These impacts have been explored in a range of international
research projects, and many OECD member countries estimate the monetary impact of
carbon emissions (OECD 2013). Focusing specifically on carbon dioxide emissions, some
countries use an estimate of the “Social Cost of Carbon” (SCC) in regulatory impact analysis
to monetize the benefits or costs of regulation (Marten et al. 2015). The SCC represents the
present value of the future stream of damages associated with marginal carbon emissions. 15

Under a set of strict equilibrium assumptions, a properly functioning carbon market can
provide insights about marginal abatement costs of carbon, and potentially the marginal
benefits of emissions to firms. Active trading in a competitive carbon permit market can
theoretically price carbon at the value at which the marginal abatement cost of a unit of
emissions are equal to the marginal benefit. Several countries, and even some US states, have
implemented carbon permit markets to mixed results. Other countries, such as Ireland, have
an explicit tax on carbon (with a current tax rate of €20/ton CO2e). For more background on

15
Note that the SCC is a global figure, so contains benefits from outside of New Zealand. Additionally, the

20  Valuing the Benefits of Forests Ministry for Primary Industries

these different approaches, the New Zealand Treasury provides an overview:
http://www.treasury.govt.nz/publications/research-policy/wp/2005/05-02/06.htm.

As a result of the Climate Change Response Act of 2002 and subsequent amendments , New
Zealand has an active carbon trading market: the New Zealand Emissions Trading Scheme
(NZETS). The NZETS also puts a price on carbon. 16 Forestry is already covered in the
NZETS along with several other economic sectors. However, agriculture is not currently
covered (Daigneault et al. 2017).

Due to the uncertain nature of the future impacts of climate change, as well as uncertainty
about future international agreements related to carbon pricing and trading, there are a wide
range of estimates for the social cost of carbon in the literature. To simplify our analysis, we
use two data sources to value the GHG emission reductions associated with each scenario.
First, we use the current carbon price in the New Zealand Emissions Trading Scheme (ETS)
approximately $17. This price, however, is likely an underestimate, as it only captures current
trading and does not directly incorporate the potential for future agreements, regulations, and
impacts that might increase (decrease) the demand for (supply of) carbon credits and
therefore raise their price. In addition, we are not modelling the potential revenue that could
be gained from the afforestation scenarios at the national level, which would require more
data and modelling. We are simply using the ETS price as a proxy for the value of emissions
reductions. In modelling the impact of its target under the Paris Agreement (on Climate
Change), New Zealand assumed a global carbon price of $50 by 2030.17

For a wider sensitivity analysis, we also employ the United States estimates for the SCC,
which vary by several factors and provide a fairly wide range of valuations. 18 An important
feature of the US estimates is that they capture the range of many other current estimates,
such as the UK traded carbon values, 19 the Irish Carbon Tax, and various EU estimates.
Furthermore, a 2010 report commissioned for the New Zealand Parliamentary Commission
for the Environment estimated future New Zealand ETS prices, and those future prices fit
within the US SCC estimates.20

Table 10 shows the range of value estimates for the reduction in net GHG emissions for
Scenarios E and I. The 2020 US SCC estimates are used for these annual values, alongside
the current New Zealand ETS price. The SCC estimates vary by the discount rate applied
(such as 5%, 3%, or 2.5%). For Scenario E, with approximately an 8 million tonne reduction
in net GHG emissions, the value of carbon benefits ranges from $137 million to $833 million
(NZ) dollars. It is again important to note that this represents an average year, which masks

16
http://www.mfe.govt.nz/climate-change/reducing-greenhouse-gas-emissions/new-zealand-emissions-trading-
scheme
17
See
http://www.mfe.govt.nz/sites/default/files/media/Climate%20Change/INDC_cabinet_paper_for_public_release.
pdf, as well as http://www.mfe.govt.nz/climate-change/international-forums-and-agreements/united-nations-
framework-convention-climate-0
18
See https://www.epa.gov/climatechange/social-cost-carbon
19
https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/600710/Updated_short-
term_traded_carbon_values_for_appraisal_purposes_2016.pdf
20
http://www.pce.parliament.nz/media/1292/covec-final-report-19-07-10.pdf

Ministry for Primary Industries Valuing the Benefits of Forests  21

important temporal fluctuations in carbon sequestration. The Scenario I benefits are estimated
to range between $10 million and $58 million.

Table 10: Value of the reduction in net GHG emissions

2020 SCC
NZ$/tonne Scenario E Scenario I
US$/tonne
(2017) (2017 $NZ) (2017 $NZ)
(2007 $US)
Reduction in GHG emissions (tonnes CO2e)
- - 6,490,750 482,929
Value of reductions in net GHG emissions
Current NZ emissions price - $17 $110,342,748 $8,209,797
US SCC 5% average $12 $19.94 $129,404,783 $9,628,063
US SCC 3% average $42 $69.78 $452,916,739 $33,698,221
US SCC 2.5% average $62 $103.01 $668,591,376 $49,744,994
Notes: Estimates for SCC are dollar-year and emissions year specific. We first convert from 2007 to 2017 dollars (assuming
1.17% increase), and then from US to NZ dollars (assuming 1.42 NZD: 1 USD exchange rate). For additional information on
the estimates, see https://www.epa.gov/climatechange/social-cost-carbon. Note that we do not use the high impact, 95th
percentile estimates, which are $123 (US) per ton.

2.7 Summary of the national analysis

Three afforestation scenarios were analysed in this section, which vary by their assumptions
about the post conversion land use. We identified new afforestation areas based on previous
work by Watt et al. (2011). Within the new afforestation areas we also identified land suitable
for mānuka and kānuka species, as well as identified those areas which would passively
revert to indigenous forest and those areas that would need active management for indigenous
afforestation.

We used NZFARM and related resources to calculate several notable national afforestation
impacts. These include the opportunity costs for each of the three scenarios, increases in
exotic forestry earnings, and carbon-related changes. The agricultural opportunity cost of lost
production was estimated to be approximately $60 million for each scenario. The opportunity
cost of land was highest for Scenario Ia, at approximately $600 million. Scenario Ib had an
opportunity cost of land of approximately $300 million. The land opportunity costs are
directly dependent on the policy used to implement afforestation, so are more uncertain than
the opportunity cost of lost EBIT.

The afforestation scenarios are expected to produce a range of benefits. We have only
monetized two in this national analysis: increases in timber earnings and the value of carbon.
We monetize and quantify several additional impacts in the Manawatū catchment case study
below, including the value of honey production and water quality benefits.

Scenario E is estimated to produce $300 million in increased Pinus radiata profits, as well as
$140 million in carbon benefits. Scenarios Ia and Ib are estimated to result in approximately
$10 million in carbon-related benefits.

22  Valuing the Benefits of Forests Ministry for Primary Industries

2.8 National-level analysis: limitations, uncertainties, and assumptions
Some of the important uncertainties and assumptions in the national-level analysis are
outlined in Table 11. As noted in the text, we have attempted to be conservative in our
assumptions where possible.

Table 11: Uncertainties and assumptions

Issue Impact on estimate Notes

Only annual values are Uncertain In some analyses, such as opportunity costs, the average
presented annual value is likely a good proxy. However, for other
impacts, such as carbon or other ecosystem services that
have a distinct time path, annual values may be a poor
representation.
Use most conservative Underestimate the total The area of land afforested could be higher, depending on the
afforestation scenario from potential land for Watt et al (2011) scenario chosen or incentivised.
Watt et al. (2011) afforestation
Assume value of land and Overestimate This addition may be double counting some opportunity costs,
lost EBIT are additive opportunity costs so represents an upper bound.
Missing a range of Underestimate benefits For this analysis, it was only possible to quantify or monetise
ecosystem services and a portion of the changes in ecosystem services, although
other costs several others are described.
Carbon valued using NZ Underestimate of full We use a fairly conservative value for a tonne of carbon,
ETS welfare impacts of which based on current rates. Several sources in the literature
climate change indicate that the actual welfare-based value is larger, and
other sources indicate that the ETS price may increase
significantly in the coming years.
Carbon valued using steady Effect depends on The estimates of carbon sequestered are based on
state average scenario assumptions from NZFARM that were made to annualize
emissions/sequestration estimates, such as a small annual harvest. More details can
from NZFARM be found in Daigneault et al. (2016).
In Scenario Ia, we assume Overestimate Other policy mechanisms might be used to incentivise
that all land is purchased opportunity cost of land permanent afforestation, such as covenants on the land. In
from landowners. that case, the opportunity cost of the land will be significantly
less.

3 Manawatū catchment analysis

To explore the potential impacts of the three afforestation scenarios in more detail, we
undertake a related analysis in the Manawatū catchment. However, in this analysis we
explore impacts over time and consider the transition path to the new land use. In
comparison, the previous national results portrayed a representative year once the policies
were fully implemented.

The Manawatū catchment is located in the Manawatū-Wanganui Region in the North Island
(Figure 4), which is the region with the largest area of new afforestation (Table 4). Most of
the Manawatū catchment is covered by pasture, as much of the original indigenous forest has
been cleared for farming over the last 150 years or so. The clearing of the indigenous forest
has led to erosion and other problems on steep slopes in the area (Ausseil et al. 2013). The
Manawatū catchment overlaps some of the area of six Territorial Authorities (TA): Central
Hawke’s Bay, Horowhenua, Manawatū, Masterton, Palmerston North City, and Tararua.

Ministry for Primary Industries Valuing the Benefits of Forests  23

Figure 4: The Manawatu catchment.

3.1 Identification of suitable areas for plantation and indigenous forest in the
Manawatū catchment
Using the same process described in the national analysis in Section 2, we identified the areas
in the Manawatū catchment for new afforestation using the most conservative afforestation
scenario in Watt et al. (2011) (Figure 5, where dark green represents the new afforestation
areas). Similar to the national analysis, we omit all conservation and reserve land from the
analysis. Approximately 40,000 ha of land were identified for new afforestation in the
Manawatū catchment. The distribution of current land uses in the new afforestation areas is
found in

24  Valuing the Benefits of Forests Ministry for Primary Industries

Table 12. The most prevalent existing land use is sheep & beef, at approximately 32,000 ha.

Ministry for Primary Industries Valuing the Benefits of Forests  25

Table 12: Current land use area for future afforestation within the Manawatū catchment

Enterprise Class Area (ha)

Dairy 800
Deer 362
Forest 642
Native 637
Other 416
Other pasture 1,326
Sheep & beef 31,606
Scrub 3,608
Arable <1

Figure 5: Afforestation areas in the Manawatu catchment.

26  Valuing the Benefits of Forests Ministry for Primary Industries

3.2 Active and passive indigenous afforestation in the Manawatū catchment
For the new afforestation areas (Figure 5), we estimate which areas would require active
afforestation in Scenarios Ia and Ib. The approach for identifying these areas is described for
the national analysis in Section 2. There were only four small land areas (total 3.3 ha)
identified as requiring active afforestation.21 These are all located in the southwest of the
Manawatū catchment (red circle in

Figure 6). The average cost of indigenous afforestation would be lower in the Manawatū
compared to the national average cost.

21
These four areas may be an artefact of overlaying multiple GIS layers, and under certain policies might not
qualify for carbon credits.

Ministry for Primary Industries Valuing the Benefits of Forests  27

 Figure 6: Active indigenous forest areas for the Manawatu catchment.

3.3 Areas suitable for Mānuka/kānuka in the Manawatū catchment

As with the national analysis, we use the models from Watt et al. (2012) to identify the new
afforestation areas suitable for mānuka/kānuka (Figure 7). Of the almost 40,000 ha identified
as new afforestation areas, approximately 24,000 ha are suitable for mānuka/kānuka. The
identified areas are coloured purple in Figure 7, while the new afforestation areas not suitable
for mānuka/kānuka appear in dark green.

Figure 7: Areas suitable for mānuka/kanuka in the Manawatu catchment.

28  Valuing the Benefits of Forests Ministry for Primary Industries

3.4 Ecosystem services classification
To better evaluate the full range of costs and benefits for the afforestation scenarios, we use
the ecosystem services classification (MA 2005). The Millennium Ecosystem Assessment
(MA)22 defined ecosystem services as “the benefits people obtain from ecosystems,” which
the MA classify as provisioning, regulating, supporting, and cultural services (Table 13).23

Table 13: Ecosystem services categories

Provisioning Services Regulating Services Cultural Services

Crops Air quality regulation Recreation & eco-tourism
Livestock: milk Climate regulation Ethical & spiritual values
Livestock: meat Water regulation: flow Educational & inspirational values
Capture fisheries Water regulation: quality
Wildfoods: honey Erosion control
Timber & wood Water purification & waste treatment Supporting Services
Fibres & resins Biological control Habitat Provision
Ornamental resources Disease regulation
Biomass fuel Pollination
Freshwater Natural hazard regulation
Genetic resources
Biochemicals, natural medicines &
pharmaceuticals
a: adapted from MA (2005).

Provisioning services include the direct products from ecosystems that people use. Many of
the agricultural products, such as meat, milk, and honey fit into this category (MA 2005).
According to Statistics NZ, approximately 8 percent of New Zealand’s GDP was derived
from primary industries in 2009, which includes agriculture, fishing, forestry, and mining,
illustrating the large amount of resources dependent on these ecosystem services.

Regulating services include the impacts from the ecosystem people obtain that help regulate
ecosystem processes, such as the regulation of air pollution by trees, the control of erosion by
tree roots, and the purification of water by plants (MA 2005). The NZETS creates a market
for the climate regulating services provided by nature.

Cultural services are the non-material benefits that people receive from ecosystems (MA
2005). New Zealand has a variety of sites that have specific cultural significance to many
people, and rapid landscape change is likely to affect these values. There are many areas with
a history of spiritual practices, experiences, and values that depend on the composition of the
landscape. For instance, historic vistas might be significantly altered by large exotic forestry
plantations. In elevated, erosion prone areas these areas’ visibility may be quite expansive. A
transition from pasture farming to forestry may also affect farming lifestyle and the

22
The MA was created by an active group of scientists, along with representatives from governments, private
sector, and nongovernmental organizations.
23
For more information on the MA, see http://millenniumassessment.org/en/index.html

Ministry for Primary Industries Valuing the Benefits of Forests  29

associated cultural experience. Similarly, the degradation/improvement in water quality from
changing land uses will affect cultural values. Recreation-related benefits were recently found
to have a very high value in Turner et al. (2011), although those benefits were related to
mountain biking and walking trails, which were quite unique to the setting of their study.

Finally, supporting services are seen as inputs into the other ecosystem services categories,
which can be necessary for their production (MA 2005). For instance, the provision of habitat
for pollinators is a necessary input to a range of agricultural products.

Although the term “ecosystem services” was formalised and popularised by the 2005 MA
report, economists have valued many of these services for decades (Freeman 2003; Atkinson
et al. 2012). Estimating these values serves an important role in policy analysis, and is
enshrined in the official requirements for regulatory analysis in several countries. 24 Placing
values on these services helps to convey their importance, and the integral role they can play
in various sectors of the economy. For instance, Gallai et al. (2009) suggest that insect
pollinators contribute approximately ($US) 190 billion to the pollination of crops used for
human consumption. However, significant challenges continue in this research, especially in
the areas of ecological production functions and related complexities in quantifying changes
in environmental outputs (Boyd & Banzhaf 2007; Ferraro & Hanauer 2011; Atkinson et al.
2012).

The analysis and valuation of ecosystem services has progressed significantly since the initial
2005 MA report. When monetising ecosystem service values, some economists such as Boyd
and Banzhaf (2007), recommend classifying ecosystem services as either intermediate or
final services to avoid double counting. Water quality, for example, is an intermediate service
for the production of fish. However, the issue is nonetheless quite complex, as water quality
is also a final service for recreators such as swimmers.

The concept of ecosystem services is being used in New Zealand to provide a consideration
of the wider impacts of land management decisions.25 Greenhalgh and Hart (2015), for
example, detail several important lessons from recent New Zealand applications, and find that
it holds considerable promise for future policy analysis and planning.

24
See, for example, the US EPA’s Guidelines for Preparing Economic Analysis:
https://www.epa.gov/environmental-economics/guidelines-preparing-economic-analyses
25
http://www.landcareresearch.co.nz/science/portfolios/enhancing-policy-effectiveness/best/integrating contains
several recent examples of ecosystem service approaches in applied policy.

30  Valuing the Benefits of Forests Ministry for Primary Industries

3.5 NZFARM Opportunity cost estimates for the Manawatū catchment
We use NZFARM26 to estimate a range of the provisioning services (Table 13), in particular
those related to agricultural or forestry land uses, for each of the three afforestation scenarios.
The opportunity cost, by TA, for each of the scenarios is outlined in Table 14 (represented by
the grey shading). As with the national analysis, the opportunity cost for each scenario is
estimated as:

 Scenario E: the new afforestation areas convert from pasture production to exotic
forestry, so the land remains in productive use. The opportunity cost is represented by the
profit from the previous land use.
 Scenario Ia: the new afforestation areas convert to indigenous forest. The land is
removed from production. The opportunity cost is composed of the profit from the
previous land use plus the value of the converted land.
 Scenario Ib: the new afforestation areas convert to indigenous forest with those areas
suitable for mānuka/kānuka being used for used for productive purposes. In this instance
for medical or edible honey production. The opportunity cost is composed of the profit
from the previous land use plus the value of the converted land. Due to the honey
production, there is less land opportunity cost than Scenario Ia.

Table 14: Opportunity costs in the Manawatū catchment (2017 NZ dollars)

Lost EBIT + Lost EBIT +

Value of Value of
converted land converted land
Lost EBIT ($NZ) converted land converted land
value value
Scenario Ia Scenario Ib
Scenario Ia Scenario Ib
Scenario E Scenario Ia Scenario Ib
opportunity cost opportunity cost opportunity cost
Central Hawke's
59,213 223,062 164,913 282,274 224,126
Bay
Horowhenua 28,949 299,046 159,939 327,995 188,888
Manawatū 1,268,287 3,380,893 2,232,817 4,649,180 3,501,105
Masterton 0 5,638 2,891 5,638 2,891
Palmerston North
43,166 367,483 195,845 410,649 239,011
City
Tararua 1,843,688 16,670,522 5,560,100 18,514,210 7,403,788
Total annual
3,243,303 20,946,644 8,316,505 24,189,946 11,559,808
values
Total value over 50
years (using 8% 42,851,048 276,750,447 109,879,029 319,601,526 152,730,078
discount rate)

Table 14 indicates that for Scenario E the opportunity cost for the catchment is approximately
$3.2 million/year in lost EBIT. For Scenario Ia, the opportunity cost is approximately $24

26
The version of NZFARM we use has the most up-to-date input data available. For instance, we are now using
the most recent version (4.1) of the New Zealand Land Cover Database (LCDB).

Ministry for Primary Industries Valuing the Benefits of Forests  31

million/year as the value of the converted land is included in the opportunity cost estimate.
This assumes the new afforestation areas are purchased by the government and set aside for
indigenous afforestation. Scenario Ib, has an opportunity cost of about $11.6 million as some
of the new indigenous afforestation is being used for productive purposes, e.g. honey
production.27

The last row of Table 14 contains estimates of opportunity cost projected over 50 years. This
estimate assumes the same annual EBIT figures are earned for 50 years. Since we do not have
any a priori assumptions or forecasts about trends in profitability, a constant flow of EBIT
likely represents the best estimate of those values. The 50-year estimate uses an 8% discount
rate, as suggested by the New Zealand Treasury for cost benefit analysis of this type. All
figures are presented in 2017 NZ dollars.

Table 15 shows the change in EBIT by land use. The highest land uses occupying the new
afforestation areas are dairy and sheep & beef.

Table 15: EBIT in afforestation areas, by previous land use

Land use EBIT ($/yr) Area (ha)

Arable 115 <1
Dairy 1,550,303 804
Deer 365,178 367
Exotic forestrya 403,314 658
Fruit 110 0
Native forestry 0 4,303
Other 0 419
Other pasture 26,973 1,353
Sheep & beef 1,299,791 31,548
Vegetables 833 0
a: The exotic forestry is presented for illustrative purposes. The areas in already in exotic forest are assumed to remain in
exotic forestry for all scenarios.

3.6 New revenue sources

The afforestation scenarios will result in several new streams of revenue, some of which we
can quantify or monetise. In Scenario E, the new afforestation areas will be planted with
exotic forestry, in this case Pinus radiata. Under similar assumptions about production
discussed for the national assessment, we use NZFARM to model the increase in EBIT
derived from to the additional exotic forestry areas. This increase in EBIT represents the
monetised change in timber and wood ecosystem services. There are also several other
ecosystem services that can either be quantified or described qualitatively. The new exotic
forest area may directly or indirectly effect a number of cultural services. For example,

27
Note again that these estimates represent an upper bound, since the profitability of the land is normally
factored in its value, and hence the profit (as represented by lost EBIT) is probably not independent from the
value of the land. In treating them as additive here, we assume that they are independent.

32  Valuing the Benefits of Forests Ministry for Primary Industries

nearby recreational opportunities may increase such as birdwatching and hiking.
Alternatively, aesthetic values (related to scenic views) may increase or decrease as some
people prefer indigenous vegetation over exotic vegetation (Brown & Mortimer 2012).
Moving from a pastoral land use to forested land may increase aesthetic values.28

3.6.1 Exotic forestry

The increase in exotic forestry in the Manawatū catchment in Scenario E is estimated to
increase EBIT by more than $17 million/year (Table 16). The Tararua district is expected to
experience the largest gain in EBIT from the new afforestation areas while Masterton gains
the least.

The third column in the table shows the difference between a full conversion to exotic
forestry and the previous earnings. The bottom row of Table 16 shows the total net present
value across 50 years, using an 8% discount rate. All of the other figures in the table
represent annual values.

There are some important caveats to these results. First, a large expansion in forestry would
require a parallel expansion in the underlying local infrastructure, such as nearby mills,
durable roads, and skilled workers.29 Second, the likelihood of farmers converting pastoral
land to exotic forestry is probably mixed. The large upfront costs and long lag time before the
trees are harvested means that exotic forestry may not be considered a viable option for some
farmers, particularly those more risk averse farmers. Risk-averse farmers tend to be the older
and more experienced farmers – who are becoming the majority. 30 On the other hand, current
and future carbon prices may send strong incentives for the conversion of marginal lands to
exotic forestry. All our scenarios, however, assume that all land identified for new
afforestation will be afforested.

28
See http://www.doc.govt.nz/Documents/conservation/human-values/evaluating-non-market-impacts-of-
wilding-conifers-on-cultural-values.pdf for further discussion of cultural values and views.
29
A recent New Zealand Herald article discussed future infrastructure problems with the current forestry
rotation: http://www.nzherald.co.nz/business/news/article.cfm?c_id=3&objectid=11692463.
30
See the results from the Survey of Rural Decision Makers
http://www.landcareresearch.co.nz/science/portfolios/enhancing-policy-effectiveness/srdm2015/15-
demographics-education-and-community/15-1-demographics

Ministry for Primary Industries Valuing the Benefits of Forests  33

Table 16: Increases in exotic forestry EBIT for Scenario E

EBIT ($) from exotic EBIT ($) from

Territorial Authority Difference
afforestation existing land usea
Central Hawke's Bay District 220,618 61,920 158,698
Horowhenua District 349,148 51,851 297,298
Manawatū District 3,142,778 1,324,465 1,818,313
Masterton District 6,043 0 6,043
Palmerston North City 363,589 69,693 293,896
Tararua District 17,072,045 2,138,689 14,933,356
Total annual EBIT 21,154,221 3,646,617 17,507,604
Total EBIT over 50 years
279,493,023 46,858,483 231,313,325
(8% discount rate)
a: This EBIT differs from Table 14 as it also includes the EBIT from existing exotic forestry areas. For instance, the previous
land use EBIT for Central Hawke’s Bay is $61,920 here, as compared with $59,213 in Table 14, because of the $2,708 in
existing forestry EBIT.

3.6.2 Honey production

In Scenario Ib, we assume that some of the new afforestation area will revert to
mānuka/kānuka and be used for mānuka honey production. Mānuka honey falls under the
wildfoods provisioning service as well as the biochemicals, natural medicines and
pharmaceuticals provisioning service. Our analysis for the Manawatū catchment uses the
mānuka prediction equations outlined in Section 2. Estimating honey revenue is difficult, as
there is limited information on the profitability of honey production in New Zealand. There
also appears to be significant variability in profits and honey quality. 31

For our analysis, we use information gathered from trade associations and beekeepers by
Daigneault et al. (2015) to estimate the EBIT from honey. 32 Based on that report, we
developed three profitability types for honey production. The least profitable honey operation
is a self-managed operation. The next most profitable operation involves hiring a beekeeper,
which is marginally more profitable than a self-managed operation. The most profitable
honey operation involves the use of a beekeeper and the production of high UMF honey,
which is quite rare. There is unfortunately a dearth of information in the literature and from
the trade associations about predicting UMF level on a particular landscape. Discussions with
Apiculture New Zealand indicated that some of the central influences on UMF include: past
history of production from a landscape, soil quality, rainfall, climate and altitude, and
genotype. 33

31
For instance, the April 2017 issue of New Zealand Beekeeper highlights significant variation in honey yields
over the past year, even on the same plot of land.
32
The final report can be found at: http://www.maniapoto.iwi.nz/wp-content/uploads/2016/04/1.-Economic-
Analysis.pdf

33
There is a new MBIE programme “Building resilience and provenance into an authentic Māori honey
industry” led by Landcare Research that is starting to address some of these challenges.

34  Valuing the Benefits of Forests Ministry for Primary Industries

Since several of those factors are included in the Watt et al. (2012) prediction equations we
use to identify land suitable for mānuka/kānuka, we have adapted them to predicting the
potential of honey operations. Those equations produce probability scores for each area on
the ability to support mānuka. As a fairly conservation assumption, we assume that only the
99th percentile of mānuka afforestation areas, in terms of rainfall and temperature probability,
are suitable for high UMF production. The remaining area with mānuka is assumed to be split
between the other two profitability types.

The estimated annual honey-related profit, by TA, is listed in Table 17, along with the
discounted net present value of those profits over a 50-year period. This calculation assumes
that the returns will remain constant over the 50 years.

Table 17: Increases in honey EBIT for Scenario Ib

Territorial Authority EBIT ($)

Central Hawke's Bay District 23,580
Horowhenua District 186,853
Manawatū District 484,053
Masterton District 1,730
Palmerston North City 79,116
Tararua District 4,920,116
Total annual EBIT 5,695,448
Total over 50 years (8% discount rate) 75,249,196

3.7 Water quality valuation

Increased afforestation will affect several important regulating ecosystem services, such as
water quality and water quantity. Water quality should improve as land is converted from
agricultural uses to forested land. Afforesting land can reduce nutrient runoff, mitigate
erosion, and prevent excess stormwater runoff. On the other hand, afforestation reduces water
yield, meaning there may be less water available for irrigation and other activities (Ausseil et
al. 2013). Due to data and time constraints for this analysis, we focus primarily on the
changes in water quality. Additionally, the changes in water quantity produce uncertain
changes in values. On one hand, it may be more expensive for farmers to irrigate their crops.
However, that may only have an impact at certain levels of existing water which are hard to
predict. On the other hand, recent literature indicates that citizens may have a positive
willingness to pay for water going to forests instead of to agriculture (Baskaran et al. 2009).

To value the benefits of water quality improvements, we employ a benefit transfer approach.
Benefit transfers use estimated non-market values from a study (or studies) to evaluate
another area or policy (Freeman 2003). Benefit transfer is commonly employed when time or
budget constraints prevent original analysis. Although there are a variety of water quality
valuation in the US and Europe, the New Zealand literature is much thinner. There are
several different types of benefit transfers, including unit transfers, function transfers, and
meta-analysis function transfers (US EPA 2014).

We use a function transfer, which allows us to correct for the characteristics of the local
population. Function transfers are generally recommended over unit transfers, as they allow
for some correction between the population of the original study and that of the site the

Ministry for Primary Industries Valuing the Benefits of Forests  35

values are being transferred to (US EPA 2014). Unfortunately, there are not enough water
quality valuation studies in New Zealand that use comparable measures of water quality, to
construct a meta-analysis function transfer solely based on New Zealand studies.

In choosing water quality valuation studies for benefit transfer purposes, there are several
central criteria. Most important, the studies must use water quality parameters that match the
outputs of our policy simulations. From NZFARM, we have data on the projected reduction
in nutrient loadings, so studies that value changes in nutrients are ideal. We are also looking
for stated preference studies in order to capture more aspects of people’s willingness to pay
(WTP) for water quality improvements. It is also important that the study is done in New
Zealand, ideally in an area similar to the Manawatū.

After reviewing a range of potential studies, we selected Baskaran et al. (2009) for our
benefit transfer.34 Baskaran et al. (2009) estimates the value of percentage reductions in
nitrate leaching from agricultural activities, which is a good match to the outputs of
NZFARM. Baskaran et al. (2009) ask respondents about their WTP for either a 10% or 30%
reduction in nitrate leaching. Their econometric model estimates different WTP values at
different income levels, which then allows us to tailor their results to the Manawatū area. We
use TA median household income data from the 2013 New Zealand Census to derive incomes
for the Manawatū catchment.

In their paper, Baskaran et al. (2009) present WTP values for 10% and 30% nitrate reductions
at several different income levels. We harness the variation in their estimates to create linear
and non-linear approximations of their WTP functions at different income levels to estimate
the WTP for the NZFARM estimated changes in nutrient runoff. Our preferred results are the
non-linear approximations, since they allow for diminishing WTP as the percent change in
water quality decreases, and because they allow WTP to approach zero as the percent change
approaches zero. Both the linear and non-linear approximations are shown in Figure 8, where
the lighter lines represent the linear approximations and the darker lines are the
corresponding non-linear approximations. They show the relationship between the percent
change in nutrients and the estimated WTP.

34
In addition to general internet searches, we used the New Zealand Non-Market Valuation Database
(http://selfservice.lincoln.ac.nz/nonmarketvaluation/default.asp), the Environmental Valuation Reference
Inventory (https://www.evri.ca/Global/Splash.aspx), and the University of Waikato working paper “Review of
freshwater Non-Market Value Studies”
(https://www.waikatoregion.govt.nz/assets/PageFiles/30275/2997672Review_of_Freshwater_Non-
Market_Value_Studies.pdf).

36  Valuing the Benefits of Forests Ministry for Primary Industries

Figure 8: Baskaran WTP approximation functions.

We calculate the total WTP for water quality changes at the TA level, based on the water
quality outputs of NZFARM. To estimate the water quality benefits across a 50 year
timeframe, we also need an estimate of population growth for each TA. For this, we base our
population growth estimates on the most recent Census. Table 18 contains the NZFARM
results for predicted reduction in nitrogen (N) leaching. The first three columns show the total
N leaching based on current land use and the two afforestation scenarios. The next two
columns convert the change from baseline to each scenario into a percent. However, this
percent assumes that all new afforestation areas have been afforested. To estimate the annual
change, the total change is split into time increments, which depend upon the time a tree
species takes to reach maturity. For this analysis, we assume it takes 30 years for exotic forest
to reach full maturity, and 50 years for an indigenous forest to reach full maturity (Carver and
Kerr, 2017). Based on those assumptions, the exotic areas see reductions in nitrogen leaching
for the first 30 years, and none thereafter, which is why the Scenario E annual reductions in
the last two columns are larger than the Scenario I reductions – they are spread over a shorter
time period.35

35
It is also likely that there is a pulse of nutrients during and shortly after harvesting, followed by some
additional reductions. Modelling that change is outside of the scope of this work, so for simplicity we assume
that it is zero leaching after the first harvesting

Ministry for Primary Industries Valuing the Benefits of Forests  37

Table 18: Predicted changes in nutrients

N Leaching (kg) Estimated % Change Annual % Change

Territorial Authority Current Scen. E Scen. I Scen. E Scen. I Scen. E Scen. I
Central Hawke's Bay 5,251 789 473 0.850 0.910 0.028 0.018
Horowhenua 5,929 1,390 834 0.766 0.859 0.026 0.017
Manawatū 69,776 12,479 7,488 0.821 0.893 0.027 0.018
Masterton 128 40 24 0.686 0.812 0.023 0.016
Palmerston North
6,516 1,507 904 0.769 0.861 0.026 0.017
City
Tararua 370,242 62,701 37,621 0.831 0.898 0.028 0.018

We apply the benefits transfer function to the water quality change in each TA36 in each year
to calculate annual WTP. The WTP figures were then applied at the household level, which is
the unit of analysis in the Baskaran et al. (2009) study. The 2013 Census population growth
figures are used to extrapolate population out 50 years to calculate the full path of benefits
(Table 19). Finally, the net present value of the benefits stream is calculated using two
alternate discount rates. The first discount rate, 8%, is the recommended rate by the NZ
Treasury. The second discount rate, 3%, is a common rate recommended in the general
valuation literature to discount social welfare benefits (US EPA 2014). The alternate discount
rate is used because the 8% figure is more representative of capital expenditures, and likely
does not represent social discounting of WTP (US EPA 2014). At both discount rates,
Scenario E has a higher WTP than Scenario I (Table 19).

Table 19: Total WTP for water quality benefits in the Manawatū catchment over 50 years (NZ$a)

Scenario E 3% Scenario E 8% Scenario I 3% Scenario I 8%

Central Hawke's Bay District 1,413,481 799,167 1,335,079 611,987
Horowhenua District 1,916,060 1,076,445 1,864,398 841,829
Manawatū District 3,061,206 1,697,848 3,032,572 1,327,608
Masterton District 2,364,891 1,320,119 2,391,795 1,063,530
Palmerston North City 7,716,217 4,348,821 7,454,413 3,391,467
Tararua District 1,749,103 1,001,482 1,614,293 762,041
Total WTP over 50 years 18,220,958 10,243,883 17,692,549 7,998,462
a: values are in 2017 New Zealand dollars

In terms of water quantity, the expected change in value is uncertain. On one hand, it may be
more expensive for farmers to irrigate their crops, given there is likely less water available for
irrigation. However, this may only be an issue at river low flow times. Without detailed
hydrological modelling, the availability of water and when water restrictions may occur are
difficult to predict. On the other hand, recent literature indicates that citizens may have a
positive WTP for water going to forests instead of agriculture (Baskaran et al. 2009). The
value of the water quantity reductions is therefore uncertain.

36
This estimate uses 2013 Census data for income (inflated to 2017 dollars).

38  Valuing the Benefits of Forests Ministry for Primary Industries

3.8 Carbon benefits
Using NZFARM, we estimate changes in net GHGs (avoided GHG emissions plus carbon
sequestered) with each afforestation scenario. Scenario E has lower net GHG emissions than
Scenario I (Table 20). For instance, in Central Hawke’s Bay District there are approximately
960 tonnes CO2e emitted under the existing land use. However, when Scenario E is fully
implemented and the trees are mature, that district is a carbon sink of almost 5,000 tonnes.
Under Scenario I, the district is only a 1,000 tonnes carbon sink

Table 20: Change in net GHGs in the Manawatū catchment

TA Existing land Scenario E Scenario I Scenario E Scenario I Total

use emissions Sequestered Sequestered Total Change Change
(tonnes CO2e) (tonnes CO2e) (tonnes CO2e) (tonnes CO2e)a (tonnes CO2e)a
Central Hawke's Bay District 960 –4,857 –237 5,817 1,197
Horowhenua District 586 –-8,226 –417 8,811 1,002
Manawatū District 11,614 –69,122 –3,744 80,735 15,357
Masterton District 16 –244 –12 260 28
Palmerston North City 597 –9,836 –452 10,433 1,049
Tararua District 62,945 –392,951 –18,810 455,895 81,755
TOTAL 76,718 –485,236 -23,672 561,951 100,388
a: a positive number means carbon is being sequestered.

The GHG figures presented in Table 20represent the steady state levels of GHG emissions.
To estimate the benefits over 50 years, we need to know the emissions transition path rather
than the steady state levels. In the absence of forest modelling to forecast the time paths for
the afforestation scenarios (which are beyond the scope of this analysis) we use the 2015
New Zealand Ministry for Primary Industries ETS lookup tables37 to estimate the growth of
various tree species, including Pinus radiata and indigenous species. The lookup tables
provide carbon sequestration rates for different regions of New Zealand.

For our calculations we assume that the exotic forest (Scenario E) is a harvest in year 30,
while the indigenous forest (Scenario I) keeps growing over the 50-year time period. The
lookup tables allow us to control for the size of the harvest, as well as carbon remaining after
the harvest in stumps and the soil, which diminishes over time. We incorporate both effects in
our estimates. The valuation assumes that credits generated are sold in the same year, and that
credits have been purchased to cover any harvest-related GHG emissions.38 Note that an
estimate based solely on the ETS lookup tables is likely an underestimate, since it only values

37
https://www.mpi.govt.nz/document-vault/4762
38
Although common in the literature, these assumptions can affect the economic viability of these of these
options, depending on whether the carbon price is expected to increase or decrease.

Ministry for Primary Industries Valuing the Benefits of Forests  39

the carbon sequestered as a result of the new land use. The reductions in emissions from the
previous agricultural land use are not included in these calculations. 39

Similar to the national case, we use the NZETS price at discount rates of 3% and 8% and
SCC at 3%. A constant NZETS price is a strong assumption, particularly if new sectors are
covered by the NZETS in the future which changes market conditions. Additionally, future
international agreements, and New Zealand’s integration into them, could significantly affect
the market price. A recent report by the Parliamentary Commission for the Environment
projects that 2030 carbon prices could be as high as $150 per tonne CO2e, with a low estimate
of $20 per tonne CO2e.40 The US EPA SCC 3% is used to account for potentially higher
future carbon prices. Exploring SCC values across time is somewhat more complicated than
the annual values presented in the national analysis, since the social cost of CO2e is both
dollar year and emissions year dependent. We only use the 3% estimate here for comparison
to the NZETS 3% estimate. There is a wide international literature recommending lower
values for social welfare, particularly those related to environmental benefits (Pearce 2003;
Guo et al. 2006; David et al. 2009). Given the PCE forecasts and the higher US EPA SCC
values, our estimates are likely conservative and may therefore be underestimates of the true
value of carbon for each of the afforestation scenarios.

The monetised benefits of carbon sequestration, by TA, for Scenario E are outlined in Table
21. All estimates in the table represent the net present value across 50 years. Depending on
the price assumption, benefits range from approximately $105 million to almost $700 million
for the Manawatū catchment.

Table 21: Carbon benefits for Scenario E over 50 years

NZETS 3% NZETS 8% SCC 3%

Central Hawke's Bay District 1,732,879 1,046,857 7,071,811
Horowhenua District 3,052,358 1,843,973 12,291,778
Manawatū District 27,410,838 16,559,281 110,382,821
Masterton District 88,109 53,228 354,811
Palmerston North City 3,310,675 2,000,027 13,332,014
Tararua District 137,724,026 83,201,060 554,611,514
Total 173,318,885 104,704,425 698,044,750
Estimates are presented in 2017 dollars (NZ)

The benefits for Scenario Ia and Ib are outlined in (Table 22) and produce slightly lower
carbon monetised benefits than Scenario E. For example, the NZETS 3% price is estimated to
generate $118 million in carbon benefits in Scenario I as opposed to $173 million in Scenario
E. Although the average tree density – and hence sequestered carbon – of exotic forests is
more than indigenous forests, they are periodically harvested.

39
To more accurately model the impact of a particular policy, such as the PFSI, would require a more detailed
analysis. For instance, areas of forest over 100 ha would require a field measurement approach, which might
differ from the lookup tables.
40
http://www.pce.parliament.nz/media/1292/covec-final-report-19-07-10.pdf

40  Valuing the Benefits of Forests Ministry for Primary Industries

Table 22: Carbon benefits for Scenarios Ia and Ib over 50 years

NZETS 3% NZETS 8% SCC 3%

Central Hawke's Bay District 1,182,273 521,915 5,367,236
Horowhenua District 2,082,500 919,321 9,393,804
Manawatū District 18,701,300 8,255,702 84,358,394
Masterton District 60,113 26,537 271,159
Palmerston North City 2,258,739 997,122 10,188,790
Tararua District 93,963,502 41,480,252 423,853,425
Total 118,248,428 52,200,848 533,432,808
Estimates are presented in 2017 dollars (NZ)

3.9 Biodiversity-related benefits

To estimate changes in biodiversity, we employ a measure of ‘ecological integrity.’ This
measure was originally defines by Lee et al. (2005) as ‘‘the full potential of indigenous biotic
and abiotic factors, and natural processes, functioning in sustainable communities, habitats,
and landscapes” Carswell et al. (2015). Indicators of ecological integrity are now widely
employed, and the New Zealand Department of Conservation uses ecological integrity as
their primary biodiversity goal. Our measure of ecological integrity is based on catchment-
scale natural regeneration of indigenous forests on agricultural lands, and has been used in
several recent papers (Mason et al. 2012; Carswell et al. 2015; Mason et al. 2016). The
measure is called the “restored significance,” and it is a measure of the potential gain in
environmental representation through natural regeneration. Larger restored significance
values indicate that there is a larger potential increase in biodiversity from converting a
particular plot of land to indigenous forest. The units of this indicator are parts per billion
(ppb), where one billion represents the ideal ecological utopia of natural (prehumen)
conditions (Carswell et al. 2015).

The distribution of restored significance (hereafter referred to as “SRS”) throughout the

Manawatū catchment is shown in Figure 9. The darker the blue indicates a higher SRS score,
indicating that more biodiversity could be gained from allowing those areas to revert to
indigenous forest. Similarly, the lighter blue areas indicate there is less to gain from allowing
an area to revert to indigenous forest. The two large lighter areas in the map are areas that are
already heavily forested and surrounded by other forested land, and hence have little to
“gain” by being “converted.”

Ministry for Primary Industries Valuing the Benefits of Forests  41

Figure 9: SRS in the Manawatū.

To gauge the potential biodiversity benefits from indigenous afforestation, we overlay the
new afforestation areas with the SRS ecological integrity data. Table 23 contains several
descriptive statistics on the new afforestation area. Table 23 applies specifically to indigenous
afforestation, and hence represents the potential ecological benefits for Scenario I. Note that
the SRS score is based on several detailed local criteria and assessed at the hectare scale. That
detailed focus helps explain the average SRS scores for the areas previously classified as
native and forest, which at first glance seem counter-intuitive. Average SRS for indigenous
forest areas, for example, is higher than sheep & beef areas. As suggested by Figure 9, most
of the new afforestation areas are on marginal lands that could see significant benefit from
conversion to indigenous forest. These estimates also represent the maximum potential
biodiversity once the area has been fully restored. These SRS calculations require two

42  Valuing the Benefits of Forests Ministry for Primary Industries

important caveats. First, although our SRS estimates are based on very spatially explicit
underlying data, the actual realized biodiversity may differ from the predicted estimate. We
therefore present multiple descriptive statistics to better gauge the distribution of potential
benefits. In addition, the SRS estimates represent the full benefits once the plot of land is
fully regenerated. According to the underlying studies the derivation of SRS is based on,
regeneration is likely to take 40–50 years. Therefore, the SRS is approximating the
biodiversity potential at the end of that time period.

Table 23: SRS in afforestation areas by existing land use, for Scenario I

25th Percentile
Minimum SRS Mean SRS Max SRS Std Dev. SRS
Enterprise Class SRS
Scrub 0.0 426.901 455.3 647.9 99.8
Deer 114.9 398.513 455.7 633.7 113.3
Native 0.0 436.056 465.0 715.0 122.3
SNB 0.0 451.316 468.8 740.6 104.2
Other 0.0 473.2595 485.1 750.8 106.1
Forest 0.0 479.3905 497.3 655.0 88.2
Dairy 0.0 483.875 497.6 655.0 99.3
Other pasture 0.0 485.231 506.7 655.0 69.8

The estimates for Table 23 use an SRS value that was developed for indigenous afforestation.
Biodiversity for Scenario E will differ from Scenario I due to differences in forest type.
Scenario E will have much less plant diversity, which will support less diversity of other
species. Another fundamental difference in exotic forestry is the periodic harvesting where
land will be harvested shortly after reaching peak potential biodiversity potential. Harvesting
is likely to damage biodiversity significantly. We, therefore, update the SRS estimate for
exotic forest based on several studies that compare biodiversity under native and exotic
forests, including Pawson et al. (2008) and Deonchat et al. (2009). The updated SRS values
for Scenario E are outlined in Table 24.

Table 24: SRS in Afforestation areas by existing land use, for Scenario E

25th Percentile
Enterprise Class Minimum SRS Mean SRS Max SRS Std Dev. SRS
SRS
Scrub 0.0 192.1055 204.9 291.6 44.9
Deer 51.7 179.3309 205.1 285.1 51.0
Native 0.0 196.2252 209.3 321.8 55.0
SNB 0.0 203.0922 211.0 333.3 46.9
Other 0.0 214.2 218.3 337.9 47.7
Forest 0.0 212.9668 223.8 294.8 39.7
Dairy 0.0 215.7257 223.9 294.8 44.7
Other pasture 0.0 217.7437 228.0 294.8 31.4

Assuming that the exotic forests will be harvested roughly twice in a 50 year period, the level
of biodiversity presented in Table 24 are only likely to be achieved twice in this period.
Therefore, the average annual biodiversity benefits will therefore be much lower.

Ministry for Primary Industries Valuing the Benefits of Forests  43

4 Summary of benefits and costs in the Manawatū Catchment
When assessing the benefits and costs of the three afforestation scenarios for the Manawatū
catchment it is important to consider those that can be monetised, those that can be quantified
and those that can only be qualitatively described.

The previous sections outline a number of impacts related to the afforestation scenarios that
can be quantified and/or monetised impacts. While it is possible to quantify and/or monetise
other impacts, the budgetary and time constraints for this project mean that no additional
primary data or ecosystem modelling could be undertaken. In terms of benefits and costs, the
benefits are often more challenging to quantify. For instance, aesthetic values are difficult to
quantify, and aesthetic preferences can vary significantly across the population and across
time. To estimate those benefits properly, a stated preference survey would be ideal (Freeman
2003). Similarly, more advanced ecosystem modelling would be needed to quantify the
indirect impacts of changes in biodiversity.

In the absence of additional data collection and ecosystem modelling we use an ecosystem
services framework to describe the broader range of impacts and the subsequent benefits and
costs of the afforestation scenarios. Table 25 presents a summary of ecosystem service
impacts, including effects that can be quantified or monetised as well as a short narrative on
the potential impacts. The narratives in the table, in particular for those ecosystem service
impacts that are qualitatively described, are not comprehensive. The table does, however,
draw on expected ecosystem service relationships and insights gained from other situations
which are likely transferable to this context.

44  Valuing the Benefits of Forests Ministry for Primary Industries

Table 25: Summary of ecosystem service impacts

Category Ecosystem Service Effect of Afforestation Quantified Monetized Methods/ Notes

Scenario
Provisioning Crops Reduced production X X NZFARM was used to examine agricultural impacts
Livestock: milk Reduced production X X NZFARM was used to examine agricultural impacts
Livestock: meat Reduced production X X NZFARM was used to examine agricultural impacts
Capture fisheries Likely improvement Fish habitat is expected to improve as water quality improves and with additional stream
shading expected with the afforestation scenarios. Decreased stream flows associated
with afforestation, however, may have some negative impacts on fish habitat. Improved
fish habitat is likely to enhance commercial fishery harvest for freshwater species such as
eel or recreational trout catch. To estimate the full effects would require hydrological and
fish modelling which is beyond the scope of this project. Any impacts on the ocean fishery
are unknown.
Freshwater Improvement in Water quality is expected to improve due to decreases in nutrient inputs and other forms
quality/decrease in of farm runoff associated with pasture land, and thereby improving drinking and stock
quantity water quality. In addition, freshwater contact recreation should be improved, yielding
human health impacts. Water yield, however, is expected to decrease with greater areas
of forested land. This may affect irrigation in the area. Hydrological modelling is required
to determine the spatial and temporal impacts on water flows.
Wildfoods Likely increase Wildfood harvests should increase, particularly in indigenous afforestation scenarios
(Scenario I). Trout and eel habitat should improve with better water quality leading to
greater fish abundance and catch. Honey will increase particularly in Scenario Ib).
Timber & wood Increase in Scenario E X X NZFARM was used to examine forestry impacts.
Fibres & resins Potential Increase Afforestation may yield products in addition to timber.
Biomass fuel Potential increase Forestry by-products could be used for biomass fuel.
Ornamental resources Potential increase With indigenous forest (Scenario I) we expect greater availability of ornamental resources
such as flax.
Biochemical, natural Potential increase High-grade mānuka honey, among other products, has several medical applications.
medicines and Mānuka is one of the first successional species that is anticipated after reversion from
pharmaceuticals pastoral farming to indigenous vegetation. Rongoā is also likely to increase in Scenario I.
Regulating Air quality and climate Improvement X X NZFARM outputs and ETS materials are used to quantify and value changes in carbon, in
regulation particular the carbon sequestration potential of forests. Forests also improve air quality in
terms of reduced particulates. Pine pollen, however, could be an issue in some areas.

Ministry for Primary Industries Valuing the Benefits of Forests  45

 Category Ecosystem Service Effect of Afforestation Quantified Monetized Methods/ Notes
Scenario
Water regulation Mixed The afforestation scenarios will likely decrease water yield in the area as runoff from
erosion-prone and pastoral areas is reduced. Alternatively, improvements in water quality
will reduce water treatment costs for drinking and agriculture water.
Erosion control Improvement Afforestation will improve erosion control.
Water quality or Improvement X X NZFARM nutrient outputs are used for a benefit transfer of stated preference WTP values
purification to monetise the value of improved water quality.
Pollination Potential improvement We expect an increase in native pollinators with indigenous forest (Scenario I); the extent,
however, will depend on the availability of floral resources. There is also an increase in
honey production (from honey bees) under Scenario I that will likely have additional
indirect pollination benefits.
Natural hazard regulation Improvement A reduction in water yield should reduce stormwater impacts, such as stream scouring,
and potentially reduce peak flooding flows
Cultural Recreation and May increase Increased afforestation may induce greater local recreation, particularly in areas with
Ecotourism greater public accessibility. This could be hiking, biking or similar recreation.
Improvements in water quality should improve the swimability of streams and also improve
the recreational experience and the health of the recreational fishery (e.g. trout). There is
some evidence of aesthetic preferences for indigenous species over exotic species
(Brown et al. 2012), which may mean greater recreation and ecotourism services are
provided by indigenous forest (Scenario I).
Ethical and spiritual Expected improvement With indigenous forest (Scenario I) there is an expected increase in the spiritual values
associated with the landscape, especially when native species increase (e.g. taonga
species).
Aesthetic values Expected improvement Views will be changed, particularly when afforested areas are elevated. The local value of
changing viewscapes depends on the local population and the particular scenario. In a
farmer workshop on ecosystem services in the Manawatū in 2015, the farming community
noted the importance of the aesthetic value of their catchment and how these attracted
international visitors.
Cultural heritage values Expected improvement Indigenous afforestation scenarios (Scenario I) may promote the return of indigenous
species with particular cultural values. Water quality improvements in culturally important
waterbodies should provide additional benefits.

46  Valuing the Benefits of Forests Ministry for Primary Industries

 Category Ecosystem Service Effect of Afforestation Quantified Monetized Methods/ Notes
Scenario
Social relations Mixed There is likely to be a change in the rural population with afforestation. With less farm
labour required there is likely to be an initial reduction of people in the catchment.
However, over time different people are expected to move into the area, but with different
employment preferences. Anecdotally, this is what happened in the Taupō catchment
when a portion of the land was afforested leading to an initial decrease in social
relations/cohesion followed by an increase when new people moved into the catchment
(Mike Barton, Farmer Lake Taupō, March 2016).
Sense of place Mixed The ‘look’ of the catchment will change with a move from pastoral land to forested land in
the marginal areas. Therefore, the sense of place may be altered (and potentially
reduced), especially for those who grew up surrounded by pastoral land. However, older
generations may feel a greater sense of place with a reversion to forest.
Cultural diversity Unclear The expected initial reduction in the rural population is likely to decrease cultural diversity.
However, as noted above this will likely change over time as new people are expected to
move into the catchment.
Supporting Habitat Provision Increase X The habitat for native species is expected to increase, particularly in the indigenous
scenario (Scenario Ia and Ib).

Ministry for Primary Industries Valuing the Benefits of Forests  47

A summary of the monetized benefits and costs for the Manawatū catchment for the three
afforestation scenarios are provided in Table 26. All figures are in 2017 dollars, and use a
discount rate of 8%.41 There are a variety of issues (not presented here) involving the
magnitude of the discount rate.42

Scenario I has the highest discounted net present value of the opportunity cost at
approximately $317 million, whereas Scenario E has the lowest opportunity cost at $43
million. Each of the scenarios faces same loss of EBIT ($43 million) associated with the
existing land use before afforestation. The opportunity cost for Scenarios Ia and Ib, however,
also includes the additional converted value of land as well. This value reflects the policy
context we used for this analysis which involved the government purchasing the land from
the current owner. Other policy context, e.g. using covenant, may not include the converted
value of land as part of the opportunity cost.

In terms of discounted benefits, Scenario E yields the largest monetized benefits, at

approximately $400 million, which includes increased profit from exotic forestry, water
quality benefits, and carbon-related benefits. Scenarios Ia and Ib have lower carbon and water
quality benefits, as well as production-related revenue (which is zero for Scenario Ia). These
differences in benefits between scenarios are important, as they each come with their own
caveats. Future policy, climate change, and farmer preferences can significantly affect the
benefits realised by each scenario in different ways.

The overall NPV and the benefit-cost ratio show all scenarios as having a positive benefit-
cost ratio. However, Scenario Ia has a negative NPV of $190 million, while Scenarios E and
Ib have a positive NPV. The negative NPV for Scenario Ia is largely driven by the
opportunity cost of the converted value of land. As noted above, this portion of the
opportunity cost is related to the policy context we used in this analysis. Therefore, these
results should be viewed as the upper-bound of estimates.

In a different policy context that does not involve land sales, the opportunity costs are lower
and only reflect on the loss of EBIT from the existing land uses. Where only lost EBIT is
included in the opportunity cost, both the NPV and benefit-cost ratio increase for Scenarios Ia
and Ib, but are unchanged for Scenario E. Overall, Scenario E has the highest NPV and
benefit-cost ratio.

Although Scenario E has the highest monetised net benefits, there are many other benefits
that were not monetised. For instance, biodiversity benefits were found to be considerably
higher for indigenous forest, although they were not assigned dollar values. Cultural,
recreation, aesthetic, and human health impacts were also not monetised or quantified. The
preferable afforestation scenario therefore depends on the preferences and constraints of the
policy makers. Overall, we find that both exotic and indigenous permanent forest have the

41
The traditional default discount rate recommended recommended by Treasury was 8%:
http://www.treasury.govt.nz/publications/guidance/planning/costbenefitanalysis/discountrates/discount-rates-
jul08.pdf. Note, however, that recent (2016) guidance has suggested alternate discount rates
http://www.treasury.govt.nz/publications/guidance/planning/costbenefitanalysis/currentdiscountrates. A full
comparison of these rates is outside the scope of this analysis. By using 3 and 8 percent in most sections, we
should capture a wide range of sensitivity.
42
For instance, higher discount rates may penalize “lumpy” effects that occur in the future, as opposed to up-
front costs.

Ministry for Primary Industries Valuing the Benefits of Forests  48

potential to produce significant benefits. With flexible policy that provides balanced
incentives to producers, both types of forest can achieve multiple regional and national goals.

Table 26: Monetized benefits and costs across 50 years (8% discount rate)

Scenario E Scenario Ia Scenario Ib

Opportunity Costs
Lost EBIT 42,851,048 42,851,048 42,851,048
Converted value of land 276,750,447 109,879,029
Total opportunity cost 42,851,048 316,660,879 152,730,078

Increases in EBIT
Forestry 279,493,023
Honey 75,249,196

Ecological Benefits
Water quality 10,243,883 7,998,462 7,998,462

Carbon Benefits
Carbon valuation
104,704,425 118,248,428 118,248,428
(Current NZ ETS price)

Total monetized benefits 394,441,331 126,246,890 201,496,086

Overall NPV 351,590,283 -190,413,989 48,766,008

Benefit-cost ratio 9.2 0.40 1.3

Sensitivity Analysis
Overall NPV – Lost EBIT only 351,590,283 83,395,842 158,645,038
Benefit-cost ratio – Lost EBIT only 9.2 3.0 3.7

Ministry for Primary Industries Valuing the Benefits of Forests  49

4.1 Assumptions and uncertainties
Although we have attempted to be comprehensive within the time and budget constraints of
the project, this analysis does not include all benefits and costs associated with the
afforestation scenarios. It is important to acknowledge these limitations, as well as some
uncertainties with the analysis.

Issue Impact on estimate Notes

Discount rates used are Underestimate benefits The choice of discount rate can affect constant values across
relatively high and costs, at different time (i.e., a constant annual cost) differently than “lumpy”
rates values. Future analysis might explore these differences in
additional sensitivity analysis.
Use most conservative Underestimate the total The area of land afforested could be higher, depending on the
afforestation scenario from potential land for Watt et al. (2011) scenario chosen or incentivised.
Watt et al. (2011) afforestation
Assume value of land and Overestimate This addition may be double counting some opportunity costs,
lost EBIT are additive opportunity costs so represents an upper bound.
Not all impacts on Underestimate benefits It was only possible to quantify or monetise a portion of the
ecosystem services (or changes in ecosystem services in this analysis. A qualitative
costs) are quantified or description was provided of likely relationships and impacts.
monetised
Carbon valued using NZ Underestimate of full We use a fairly conservative value for a tonne of carbon,
ETS welfare impacts of based on current rates. Several sources in the literature
carbon indicate that the actual welfare-based value is larger, and
other sources indicate that the ETS price may increase
significantly in the coming years.
Climate change Uncertain Several of the models are based on current and historical
assumptions information, which may not reflect future climate change. For
instance, predictions of tree growth from the ETS lookup
tables might be affected by different climate change scenarios.
Population and income Uncertain For our analysis of water quality, we assume a constant
growth assumptions population growth rate based on current projections, as well
as current income levels. Since the WTP depends on
population and income, and those variables may change
across time, it can affect the estimated value of water quality.

Limitations
Several important limitations with current data and models were encountered in the course of
this analysis. These limitations illustrate where additional analysis could be undertaken in the
future. These include:

 Better forecasts of carbon prices and values across time.

 The incorporation of the benefits and costs of erosion were not feasible within time and
budget constraints. Some issues encountered were due to differences in land use and land
cover databases within New Zealand. Many erosion databases are based on current land
cover, whereas this report is focused on land use. These issues were further complicated
by erosion projections for hill country areas. Spatially explicit predictions and modeling
in those areas were limited.
 The literature on the value of water quality improvements in New Zealand is somewhat
thin. Our benefit function transfer was done on only one study. Although international
sources might be used in the future, there are concerns with the transferability of benefits

50  Valuing the Benefits of Forests Ministry for Primary Industries

 estimates to New Zealand. There are also only limited sources for the cultural values
around water quality.
 Only a small subset of ecosystem services was able to be monetised with the data, time,
and resources available for this report. Additional analysis related to those ecosystem
services would be beneficial. This may involve the collection of primary data as well as
ecosystem modelling.

5 Works cited
Atkinson G, Bateman I, Mourato S 2012. Recent advances in the valuation of ecosystem
services and biodiversity. Oxford Review of Economic Policy 28(1): 22–47.

Ausseil AGE, Dymond JR, Kirschbaum MUF, Andrew RM, Parfitt RL 2013. Assessment of
multiple ecosystem services in New Zealand at the catchment scale. Environmental
Modelling & Software 43: 37–48.

Baskaran R, Cullen R, Colombo S 2009. Estimating values of environmental impacts of dairy

farming in New Zealand. New Zealand Journal of Agricultural Research 52(4): 377–
389.

Boyd J, Banzhaf S 2007. What are ecosystem services? The need for standardized
environmental accounting units. Ecological Economics 63(2–3): 616–626.

Brown P, Mortimer C, Meurk C 2012. Landscape preferences in the Canterbury Region.

Landcare Report LC964.

Brown P. 2015. Survey of rural decision makers. Landcare Research NZ Ltd.

Carswell FE, Mason NWH, Overton JM, Price R, Burrows LE, Allen RB 2015. Restricting
new forests to conservation lands severely constrains carbon and biodiversity gains in
New Zealand. Biological Conservation 181: 206–218.

Carver, Thomas and Suzi Kerr 2017. Facilitating Carbon Offsets from Native Forests. Motu
Working Paper 17-01, Motu Economic and Public Policy Research.

Daigneault A, Wright W, Samarasinghe O 2015. Economic analysis of land use opportunities

in the Maniapoto rohe. Landcare Research Report LC2415.
http://www.maniapoto.iwi.nz/wp-content/uploads/2016/04/1.-Economic-Analysis.pdf

Daigneault A, Greenhalgh S, Samarasinghe O 2017. Economic impacts of multiple agro-

environmental policies on New Zealand land use. Environmental and Resource
Economics: 1–23.

David A, Richard SJT, Gary WY 2009. Risk aversion, time preference, and the social cost of
carbon. Environmental Research Letters 4(2): 024002.

Ferraro PJ, Hanauer MM 2011. Protecting ecosystems and alleviating poverty with parks and
reserves: ‘win-win’ or tradeoffs? Environmental and Resource Economics 48(2): 269–
286.

Ministry for Primary Industries Valuing the Benefits of Forests  51

Freeman M. 2003. The measurement of environmental and resource values. Washington, DC,
Resources for the Future Press.

Gallai N, Salles J-M, Settele J, Vaissière BE 2009. Economic valuation of the vulnerability of
world agriculture confronted with pollinator decline. Ecological Economics 68(3): 810–
821.

Greenhalgh S, Hart G 2015. Mainstreaming ecosystem services into policy and decision-
making: lessons from New Zealand’s journey. International Journal of Biodiversity
Science, Ecosystem Services & Management 11(3): 205–215.

Guo J, Hepburn CJ, Tol RSJ, Anthoff D 2006. Discounting and the social cost of carbon: a
closer look at uncertainty. Environmental Science & Policy 9(3): 205–216.

La Notte A, D’Amato D, Mäkinen H, Paracchini ML, Liquete C, Egoh B, Geneletti D,

Crossman ND 2017. Ecosystem services classification: a systems ecology perspective
of the cascade framework. Ecological Indicators 74: 392–402.

Marten A, Kopits E, Griffiths CW, Newbold S, Wolverton A 2015. Incremental CH4 and
N2O mitigation benefits consistent with the US Government's SC-CO2 estimates.
Climate Policy 15(2): 272–298.

MA 2005. Ecosystems and human well-being: a framework for assessment.

http://www.millenniumassessment.org/en/Framework.html

MPI 2013a. Situation and outlook for primary industries. MPI Policy Publication.
Wellington, New Zealand, Ministry for Primary Industries.

MPI 2013b. Farm monitoring report. MPI Policy Publication. Wellington, New Zealand,
Ministry for Primary Industries,.

Mason NWH, Ausseil A-GE, Dymond JR, Overton JM, Price R, Carswell FE 2012. Will use
of non-biodiversity objectives to select areas for ecological restoration always
compromise biodiversity gains? Biological Conservation 155: 157–168.

Mason NWH, Palmer DJ, Vetrova V, Brabyn L, Paul T, Willemse P, Peltzer DA 2016.
Accentuating the positive while eliminating the negative of alien tree invasions: a
multiple ecosystem services approach to prioritising control efforts. Biological
Invasions: 1–15.

Newsome PFJ, Wilde RH, Willoughby EJ 2008. Land resource information system spatial
data layers: data dictionary. Palmerston North, New Zealand, Landcare Research New
Zealand Ltd. https://lris.scinfo.org.nz/document/162-lris-datadictionary-v3/download/
(accessed 14 May 2015).

New Zealand Treasury 2008. Public sector discount rates for cost benefit analysis.
Wellington, New Zealand Treasury.

OECD 2013. Climate and carbon, aligning prices and policies, OECD Environmental Policy
Paper No. 2013-1.

52  Valuing the Benefits of Forests Ministry for Primary Industries

Pearce D. 2003. The social cost of carbon and its policy implications. Oxford Review of
Economic Policy 19(3): 362–384.

Turner, James A., Bhubaneswor Dhakal, Richard Yao, Tim Barnard, and Colin Maunder
2011. Non-timber Values from Planted Forests: Recreation in Whakarewarewa Forest.
New Zealand Journal of Forestry 55(4): 24-31.

US EPA 2014. Guidelines for preparing economic analysis. U. EPA. Washington DC,
https://yosemite.epa.gov/ee/epa/eerm.nsf/vwAN/EE-0568-50.pdf/$file/EE-0568-50.pdf.

Watt MS, Palmer DJ, Hock BK 2011. Spatial description of potential areas suitable for
afforestation within New Zealand and quantification of their productivity under Pinus
radiata. New Zealand Journal of Forestry Science 41: 115–129.

Watt MS 2012. Future forest systems. MPI Technical Paper No 2012/40. Prepared for the
Ministry for Primary Industries by Scion.

Ministry for Primary Industries Valuing the Benefits of Forests  53