Environmental Impact Assessment

Project Number: 52111-001 February 2020

Samoa: Alaoa Multi-purpose Dam Project

Volume 2: Assessment of Ecological Effects: and Bats (Part 1 of 9)

Prepared by Rebecca Stirnemann and Briar Taylor-Smith for the Asian Development Bank.

This environmental impact assessment is a document of the borrower. The views expressed herein do not necessarily represent those of ADB's Board of Directors, Management, or staff, and may be preliminary in nature. Your attention is directed to the “terms of use” section on ADB’s website.

In preparing any country program or strategy, financing any project, or by making any designation of or reference to a particular territory or geographic area in this document, the Asian Development Bank does not intend to make any judgments as to the legal or other status of any territory or area.

0

Proposed Alaoa Dam

Vaisigano River, Upolu,

Assessment of Ecological Effects:

Birds and Bats

1

Prepared by Dr Rebecca Stirnemann and Dr Briar Taylor-Smith for the Asian Development Bank Pictures: 1) Manumea 2) MNRE staff Fialelei Enoka and a Mao. Contents

1 Executive Summary 4 2 Introduction 4 2.1 Background 4 2.2 Policy, Legal and Administrative Framework 9 2.3 Critical habitat 10 2.4 Legal protection and international recognition of the site 10 2.5 Relative Importance of the Project Area 11 2.5.1 Ecology and Distribution 11 2.5.2 Manumea 12 2.5.2.1 Ecology and Distribution 12 3 Methods 14 3.1 Assessment of Ecological Effects methodology 14 3.1.1 Evaluating ecological features 14 3.1.2 Assessing potential and actual effects 15 3.1.3 Establishing impact management options 15 3.2 Vegetation mapping 16 3.3 Avifauna surveys 16 3.3.1 Playback surveys 16 3.3.2 Automated recorders 17 3.3.3 Territory determination 17 3.3.4 Targeted habitat identification 17 3.4 Bat surveys 18 4 Description of Existing Ecological Values 19 4.1 Habitat 19 4.3 Habitat value & threat status 22 4.4 Mao 23 4.4 Manumea 27 4.5 Other species 29 4.6 Bats 29 5 Project effects 29 5.1 Summary of adverse ecological effects 29 5.2 Effects on vegetation 32 5.3 Effects on avifauna 33 5.4 Effects on bats 38

2

6 Avoidance, remediation, mitigation and offsetting/compensation requirements 38 6.1 Mitigation Strategy 38 6.2 Vegetation mitigation 41 6.3 Avifauna mitigation & monitoring 42 6.3 Bat mitigation & monitoring 43 References 43 Appendix I: Vegetation types - species lists 45 Appendix II 47

3

1 Executive Summary

The proposed Alaoa dam site contains habitat for the endangered mao ( samoensis) and the critically endangered manumea (Didunculus strigirostris) as well as habitat for additional range restricted species such as the Samoan fruit bat and other threatened Samoan avifauna. Territories of mao occur within the footprint. The site is defined as critical habitat for both mao and manumea. The Project is categorized as environment Category A, because potential adverse environmental impacts are not only site-specific due to the potential population level on mao and Manumea. Samoan fruit bats and some threatened range restricted bird species will also be impacted and are detailed in this document. Impacts on key species could have significant adverse environmental impacts if mitigation is not carefully designed and implemented. Stringent environmental management will be required for the Alaoa dam project to proceed. Mitigation and environmental management would be needed to be carefully designed to ensure these species are maintained and enhanced by the development. It is critical that there are no significant adverse environmental impacts that are irreversible, diverse, or unprecedented for these populations. An Environmental Management Plan (EMP) would be required should the project proceed.

2 Introduction 1.1.1 2.1 Background The Terrestrial Ecology Study is part of the wider Technical Feasibility and other Scientific Assessments that will provide the information for the resource consenting process for the Alaola Dam Project. The purpose of this Assessment of Ecological Effects (AEE) report is to present the environmental impacts and risks associated with the proposed project and make recommendations regarding measures to avoid, mitigate or offset potential adverse effects. Specifically, the objectives are to: To conduct some additional biodiversity surveys focusing on endangered/critically endangered birds and bats, to map the vegetation of the area and provide qualitative and quantitative data on the presence and abundance of sensitive species.

The biodiversity expert will:

● Conduct biodiversity surveys of Samoan endangered birds and bats in the area of influence for the project. ● Conduct botanical surveys of the vegetation present in the area of influence of the project to identify the habitat and the presence of /habitat potentially able to sustain the . ● Delivery of a biodiversity report indicating the presence and abundance of the targeted species, basic habitat mapping of vegetation to determine the habitat value for key species and suggestions on how to mitigate the environmental impacts of the proposed project.

The scope of the project is limited to the methodology milestones set out in the Request for Proposal documents from ADB and the extent of physical works and flooding associated with the proposed Alola Dam as shown in the proposed site layout in Figures 1 and 2 below - provided by ADB and Entura. This report does not cover quantification of habitat loss outside of the flood area (within dotted pink Fig. 1) and the dam wall as seen in Figure 1 and the white line seen in Figure 2. Due to the lack of available information at the time the report was drafted the effect of roads and other construction cannot be quantified.

4

5

Figure 1. Proposed dam site layout. The footprint as defined in this document are the areas within the pink dotted line and inclusive of the grey dam wall as seen in this figure and the white line on Figure 2.

6

1.1.2 2.2 Policy, Legal and Administrative Framework The Project shall comply with the requirements of the Planning and Urban Management Act 2004 and the Environmental Impact Assessment (EIA) Regulations 2007. For development of hydropower projects, development consent must be obtained from the Planning and Urban Management Agency (PUMA) within the Ministry of Works, Transport & Infrastructure and from the Ministry of Natural Resources and Environment (MNRE). The development consent application must include an environmental assessment which complies with the Act and Regulation requirements. The Project should also comply with the requirements of the Asian Development Bank (ADB) Safeguard Policy Statement (ADB, 2009) to: (i) avoid adverse impacts of projects on the environment and affected people;

(ii) where possible; minimize, mitigate, and/or compensate for adverse project impacts on the environment and affected people when avoidance is not possible; and

(iii) help borrowers/clients to strengthen their safeguard systems and develop the capacity to manage environmental and social risks.

The environment safeguard requires due diligence which entails addressing environmental concerns, if any, of a proposed activity in the initial stages of project preparation. Proposed projects are screened according to type, location, scale, and sensitivity and the magnitude of their potential environmental impacts, including direct, indirect, induced, and cumulative impacts. The SPS categorizes potential projects or activities into categories of impact (A, B or C) to determine the level of environmental assessment required to address the potential impacts. Projects are assigned to one of the following four categories: (i) Category A: A proposed project is classified as Category A if it is likely to have significant adverse environmental impacts that are irreversible, diverse, or unprecedented. These impacts may affect an area larger than the sites or facilities subject to physical works. An Environmental Impact Assessment is required.

(ii) Category B: A proposed project is classified as Category B if its potential adverse environmental impacts are less adverse than those of Category A projects. These impacts are site-specific, few if any of them are irreversible, and in most cases mitigation measures can be designed more readily than for category A projects. An Initial Environmental Examination is required.

(iii) Category C: A proposed project is classified as category C if it is likely to have minimal or no adverse environmental impacts. No environmental assessment is required although environmental implications need to be reviewed.

(iv) Category FI: A proposed project is classified as category FI if it involves investment of ADB funds to or through a financial intermediary.

The Project is categorized as environment Category A, because potential adverse environmental impacts are not only site-specific due to the potential population level on mao (Gymnomyza samoensis) and manumea (Didunculus strigirostris). These effects will affect an area larger than the footprint and will also include sites or facilities subject to physical works. Impacts are likely to have significant adverse environmental impacts if mitigation is not carefully designed and implemented. This EIA has been prepared as the requisite level of assessment to address the potential impacts in line with the SPS.

1.1.3 2.3 Critical habitat Critical habitat is defined by ADB for purposes of informing investment decisions as a subset of both natural and modified habitat that deserves particular attention. Critical habitat includes areas with high biodiversity value, including habitat required for the survival of critically endangered or endangered species; areas having special significance for endemic or restricted-range species; sites that are critical for the survival of migratory species; areas supporting globally significant concentrations or numbers of individuals of congregatory species; areas with unique assemblages of species or that are associated with key evolutionary processes or provide key ecosystem services; and areas having biodiversity of significant social, economic, or cultural importance to local communities (ADB, 2009). Critical habitats include those areas either legally protected or officially proposed for protection, such as areas that meet the criteria of the World Conservation Union classification, the Ramsar List of Wetlands of International Importance, and the United Nations Educational, Scientific, and Cultural Organization’s world natural heritage sites (ADB, 2009). ADB Safeguard Policy Statement (SPS) requirements, state that project activities should not be implemented in areas of critical habitats, unless (i) there are no measurable adverse impacts on the critical habitat that could impair its ability to function, (ii) there is no reduction in the population of any recognized endangered or critically endangered species, and (iii) any lesser impacts are mitigated. If a project is located within a legally protected area, implement additional programs to promote and enhance the conservation aims of the protected area. In an area of natural habitats, there must be no significant conversion or degradation, unless (i) alternatives are not available, (ii) the overall benefits from the project substantially outweigh the environmental costs, and (iii) any conversion or degradation is appropriately mitigated. Use a precautionary approach to the use, development, and management of renewable natural resources (ADB 2009).

1.1.4 2.4 Legal protection and international recognition of the site The existing Alaoa area is contained within the Catchments KBA. A KBA is a site of critical importance for the conservation of globally important biodiversity (Conservation International et al., 2010). KBAs support the regular occurrence of one or more globally threatened species. The Apia KBA totals 8,335 hectares and includes the mid to high elevation portions of catchments that drain to Apia, including the . There are eight globally or nationally threatened species found within this KBA: Samoan bush palm, manumea or toothbilled pigeon, ground dove, mao or Mao mao, Samoan broadbill (Myiagra albiventris), Samoan triller (Lalage sharpie), Samoan flying fox (Pteropus samoensis) and the land snail (Thaumatodon hystrucelloides).

The closest legally protected area is Mount Vaea Scenic Reserve, located on the summit of Mount Vaea approximately three kilometres to the north. This area includes the Reserve which is the burial ground of author . The O le Pupu Pue National Park is located approximately five kilometres to the south-east and is immediately adjacent to the Vaisigano catchment. The footprint is located on the edge of the Apia Catchment KBA (Birdlife 2019). However the KBA site has no legal protection.

1.1.5 2.5 Relative Importance of the Project Area

2.5.1 Ecology and Distribution The ma’oma’o or mao is endemic to Samoa (Photos 1. a&b). The species is listed as endangered by the IUCN. Mao were previously found on both Upolu and Savaii in the upland and foothill forests (Watling, 2004). The mao was also historically found on Tutuila, , but it is now locally extinct on all American . The species was last collected from Tutuila in 1924 and the last unconfirmed report of the bird was in 1977 (Watling, 2004). The reason for the local is not known. Following cyclones Ofa and Val a decline in the number and distribution of Ma were noted. This decline is likely to be linked to the severe storm damage to critical habitat. During a rapid biodiversity survey in 2014 biorap, out of 135 random point counts over four areas in both Savaii and Upolu mao were recorded twice (ICCRIFS 2014). Both points were within the Lake Lanoto’o reserve, an extremely small area which can only maintain a small number of territories (R. Stirnemann, pers. obs). These results suggest that mao distribution is more limited than previously thought. Mao populations occur predominantly on village owned land outside of protected areas. The current population size of the mao is unknown. The IUCN lists an estimated population of 375-1,499 individuals. The recovery plan (MNRE, 2006) suggests that mao may be present in smaller numbers than the manumea, a species which has recently been classified as critically endangered. The exact cause of the decline of the mao in Samoa has been linked loss of habitat and the presence of invasive species, such as rats, which predate the chick and may result in mortality of the breeding female (Stirnemann et al. 2016). Slow life history traits, such as a small clutch size and low maximum annual reproductive success, also contribute to the vulnerability of this species (Stirnemann et al. 2016). At maximum, only a single chick is produced each year (Stirnemann et al. 2016a). The chick is dependent on the mother bird for 2-2.5 months post fledging (Stirnemann et al. 2016a). Nest predation by rats is a problem, particularly in modified sites near plantations where rat numbers are higher (Butler & Stirnemann, 2013). Mao only lay a single egg and are not known to have more than one brood in a season, as a result the annual fecundity is low (Stirnemann, 2016). In the Vaisigano and Lake Lanoto’o catchments a reproductive success rate of 0.33 chicks per adult female per year was observed. This was reduced to 0.125 chicks per female near plantations where predation of rats is high. If rat numbers are reduced and nests are actively protected (through trapping and rat rings around trees), reproductive success and adult and fledgling survival increases (Stirnemann pers com). Cats may also predate juveniles which spend time on the ground and females are vulnerable to predation while on the nest (Stirnemann et al. 2016).

Mao are absent from logged sites but can occupy modified habitat such as plantations where large trees still remain. The Samoan government identified six key sites for mao conservation. The Vaisigano catchment is identified as one of five sites for mao conservation on Upolu, and one of five “first priority” sites in Samoa. In Malololelei intensive pest control activity has now been occurring for over 2 years. Mao have been responding well to this conservation effort. Early results suggest a possible spill over of mao outside the target pest control area into the surrounding habitat.

Photos. 1. a&b. A male mao ringed and carrying a transmitter.

2.5.2 Manumea

2.5.2.1 Ecology and Distribution The manumea is endemic to Samoa and is the country’s national bird (photos 2. a&b). Manumea is listed as critically endangered by the IUCN. This is the highest threat ranking for extant species in the wild. The threat ranking for manumea was raised in 2014 from a former classification of endangered because the population was estimated to be smaller than previously thought. Low numbers of recent records and a decreasing number of sightings suggest that the population is now extremely small. In the 1980s, Dr Ulf Beichle, who studied the manumea in the 1980s and 1990s, estimated the population of manumea was to be approximately 4800-7200 birds (MNRE, 2006). During this period flocks of Manumea could still be seen. Over the last few decades large areas of forest in Samoa were deforested by logging activity and also heavily impacted by cyclones (Cyclones Ofa, 1990 and Val 1991). In 2000, a population of fewer than 2,500 mature Manumea were estimated. The current population is considered to be approximately 300-100 birds. Between 1978 to 2000, birds were recorded in many locations from near sea level to the uplands of Upolu, Savai’i and on Nuutele Island (MNRE, 2006). Recent surveys indicate that manumea are now rare (Butler & Stirnemann, 2013). A survey in the Savai’i uplands took place with the hope that the area was a possible stronghold for the species (Butler & Stirnemann, 2013). These surveys suggest manumea may not occur in high numbers in the Savai’i uplands. In 2005-2006, manumea were recorded at 10 out of 24 sites surveyed (MNRE, 2006). Between 2011 and 2019, 48 possible sightings of manumea were recorded from scientists and local people. Several sightings of manumea occurred within or near the water catchments above Apia (Butler & Stirnemann,

2013), including flying across Cross Island Road and in Malololelei. Confirmed sightings in the Vaisangango catchment have been made near Magiagi, in Malololelei, flying over cross island road. Furthermore, Moeumu Uili and Fialelei Enoka from MNRE confirmed and photographed a juvenile Manumea at Salelologa (MNRE, 2014). The juvenile had just fledged the nest confirming breeding in this area. The Manumea was once considered to be a bird of upland montane forests (MNRE, 2006). However, recent observations suggests that lowland areas maybe critical for this species. The manumea has been observed in mature and secondary forest including in forest edge habitat, along forest roads and sometimes visiting clearings where large native fruiting trees remain. The species has been recorded from sea- level to 947m (MNRE, 2006). However, it is likely the species elevational distribution is seasonal reflecting fruiting of key species. Manumea distribution has been closely linked to native fruit-bearing trees such as Dysoxylum (Dysoxylum maoto, D. samoense and D. huntii) and aoa or banyan trees (Ficus obliqua). Manumea play a vital ecological role by distributing the seeds of native Samoan trees (Butler & Stirnemann, 2013).Very little is known about the breeding biology of the Manumea. Old records and Samoan sayings suggest it may nest close to the ground. Threats to manumea include hunting, predation by feral cats, rats, pigs and loss of habitat, and the effects of climate change, including increased risk of cyclones and droughts (MNRE, 2006, Stirnemann pers com). Loss of habitat can also occur through forest loss caused by the effects of invasive [smothering forest trees] such as rubber trees replacing native trees. Manumea was traditionally a highly esteemed source of food, especially for the high Chiefs in Samoa (Stirnemann, 2017). Recent surveys, however, have shown that in recent times hunting by local people focuses on the Lupe or Pacific Pigeon which is preferred for its meat. When manumea are currently shot it is predominantly as bycatch to the Lupe or Pacific Pigeon (Stirnemann, 2017). In the 1800s cats were first identified as a threat to the species when they were observed killing adult birds. Rats may also prey on eggs and chicks. Foraging and nesting low to the ground would put eggs, chicks and adult birds at increased risk of predation. The Manumea Recovery Plan (MNRE, 2006) identified eight key sites for conservation of manumea, comprising of five sites on Upolu and three sites on Savai’i. The locations included upland and lowland sites, although the Vaisigano catchment was not identified as a key site. A more recent report (MNRE, 2014) has identified only three key sites for conservation, consisting of two lowland sites on Savai’i and one upland site on Upolu. These are Aopo forest; Faala to Salelologa; and the Vaisigano catchment and Lake Lanoto’o. The map also highlights the upper, central and western Vaisigano catchment (MNRE, 2014).

Photo 2. a). Adult and juvenile manumea specimens, New York. b) Manumea in native forest.

2. 3 Methods 2.1.1 3.1 Assessment of Ecological Effects methodology The potential ecological effects on birds and bats by the proposed Alaoa Dam project were assessed using the Ecological Institute of Australia and New Zealand (EIANZ) Ecological Impact Assessment Guidelines (EIANZ 2018). These guidelines provide a framework for identifying, quantifying and evaluating the potential impacts and effects of defined actions on ecosystems. Although these guidelines were designed for use in Australia and New Zealand, they provide a scientifically defensible approach to ecosystem management in the context of development and can easily be adapted to other countries as best practice guidelines. The key steps to the application of these guidelines are as follows:

1. Scoping 2. Describing ecological features through detailed investigations 3. Evaluating ecological features 4. Assessing potential and actual effects 5. Establishing impact management options 6. Developing monitoring requirements

3.1.1 Evaluating ecological features Work carried out during the detailed planning and design stages, to identify and describe ecological features within the zone of influence. This will give scientific basis for evaluation, impact assessment and impact management and can feed into project design and AEE. To achieve this both qualitative and quantitative information are collected.

3.1.2 Assessing potential and actual effects To determine the nature and magnitude of project effects an assessment of impact/effects was undertaken with identification and prediction of potential positive and adverse effects of the activity. Where there is high confidence about the actual effects, it is relatively easy to describe these and the extent to which an ecological feature would be impacted. But where there is more uncertainty, a more cautious approach is taken, and a worst case scenario assumed. If this uncertainty involves an ecological component of high value, then a precautionary approach is taken. The degree of effect and the need for impact and effect avoidance, remedy and mitigation, as well as other management opportunities such as enhancement were also part of this assessment.

3.1.3 Establishing impact management options The report was prepared by assessing baseline conditions, identifying risks during site visit and desktop evaluation, evaluating potential Project impacts and benefits, and assessing mitigation and management mechanisms relative to ADB safeguards, and national and international statutory requirements (refer to section 2.2 above). As part of the overall report broad recommendations have been made as to possible monitoring, avoidance, remediation and mitigation strategies to minimise any potential adverse ecological effects associated with the proposed dam scheme on each element of the affected terrestrial ecosystems and habitats for the key avifauna and bat species. The magnitude of the effect is described in accordance with Table 1 and the level of effect as described in Table 2. We also assess the timescales for duration of effects as outlined in Table 3 (EIANZ 2018).

Table 1: Criteria for describing magnitude of effect (from EIANZ 2018)

Table 2: Criteria for describing level of effect (from EIANZ 2018)

Table 3: Possible timescales for duration of effects (from EIANZ 2018)

2.1.2 3.2 Vegetation mapping Vegetation on the site was mapped using Georeferenced high quality LIDAR imagery obtained from MNRE PUMA (2017-2018). The proposed footprint obtained from ADB and Entura was overlaid over the site to enable accurate mapping of habitat and vegetation. Key species lists were developed for each vegetation type. This was then used for targeted habitat identification.

2.1.3 3.3 Avifauna surveys Site visits to the Alaoa site were undertaken by the project team between 10-21 September 2019. Members from EPC and MNRE were present during visits. Targeted surveys occurred for mao, manumea and Samoan triller. The time of year was appropriate for Mao. The timing was suboptimal for Manumea in the target catchment as none of the fruiting trees which attract them were fruiting. The survey timing was a little late for some of the other which tend to breed earlier and therefore are not as responsive at this time. The timing was optimal for assessing nesting use of seabirds in the catchment. Additional birds of interest were recorded at an adhoc basis. Four methods were used to survey key species, as described below.

3.3.1 Playback surveys Playback surveys of the three bird species occurred both within the site and in the surrounding area. A Foxpro playback recorder was used which had a mixture of calls and songs from the target species. Surveys occurred in the early morning (6-10am) and in the late afternoon (4-8pm) targeting the preferred calling time for each species of interest. Playback was used 5 times at 100m intervals within the footprint.

3.3.2 Automated recorders The use of automated acoustic recording devices (ARDs) for detecting birds and other has increased rapidly in recent years (Photo 3a). Advantages include the ability to estimate the number of species present at many sites simultaneously, the generation of a permanent and reviewable record over prolonged time periods, minimal disturbance to wildlife and the ability to sample the audible soundscape 24 hours per day. ARDs were developed and designed by DOC, each incorporating 4 x wm61a electret microphones in parallel with a foam filter and custom-made low noise pre-amplifier with a DSP anti-aliasing filter (Photo 3b). Recordings were saved to Secure Digital (SD) memory card as a series of uncompressed 32 kHz, 16bit audio files in waveform audio file format (‘.WAV’ file extension) with a bit-rate of 512 kbps, each approximately 15 minutes in length. In total, a maximum of 80 hours of recordings can be generated for each point. Points for automatic recorders were established at random locations >200 m apart along randomly-selected areas within the footprint. This distance allowed independence between the recorders. All recorders were set at high sensitivity and covered a period from in the morning from 6-10am and in the evening from 5-8pm. Recordings were analysed using ‘AvianNZ’ software to determine presence or absence per point. This software generated sonograms from the ARD recordings and allowed audio playback for identification. This allows a sound file to be reviewed in 6–10 minutes (rather than 15–20 minutes to listen to the whole sound file manually in the absence of software).

Photo. 3a) An automatic sound recorder being placed in a tree, b) Automatic sound recorders.

3.3.3 Territory determination Identified individual birds were followed on foot where possible to determine spatial usage. Behavioural cues were recorded opportunistic in regards to breeding and feeding to determine if the habitat core was within the footprint for species such as mao.

3.3.4 Targeted habitat identification During the site visits a qualitative assessment of vegetation and habitat was undertaken, with an emphasis on the quality of habitat for threatened avifauna, specifically mao , manumea and Samoan triller. Vegetation maps were developed based on aerial imagery and ground truthing. Phenological status of the vegetation was recorded as it predicts seasonal use and presence of the target species. For

instance, epiphytic mistletoe (Decaisnea sp) is a strong predictor of mao presence and was used as an indicator of the value of the area to this species. We also assessed the vegetation for the key trees species that manumea feed upon.

2.1.4 3.4 Bat surveys Surveys for Tongan fruit bats occurred in the evening when they are active. Samoan fruit bats, a day bat, were surveyed during the day when they are active. We recorded bat presence within the footprint. We also assessed the vegetation for the key trees species that fruit bats feed upon.

3. 4 Description of Existing Ecological Values 3.1.1 4.1 Habitat Approximately 24.4 ha of vegetation will be directly impacted by dam construction and operation through direct removal1. This vegetation can be categorised into two broad vegetation types: secondary forest; and mixed plantation. A map of vegetation types is shown in Figure 3.Detailed lists, their importance for threatened bird species and phenological observations are shown in Appendix I. Approximately 6.6 ha of the project footprint comprises mixed plantation, comprising forest clearings with wild and cultivated crops such as taro, banana and cocoa (Photo 4 a&b, Appendix I Table 1). Trees include Panama rubber trees, albizia containing native mistletoe, wild bush banana, large tava and large Dysoxylum spp. Critical food- source species for mao and/or manumea observed within this vegetation type include:

● Mistletoe (Decaisnea sp); ● Dysoxylum sp.; ● Mountain ginger (Etlingera cevuga); and ● Gatae (Erythrina variegata). ● Maota (Dysoxylum maota);

Approximately 17.8 ha of the project footprint comprises secondary forest containing a mixture of indigenous and exotic vegetation. This vegetation type is located on steeper valley slopes and ridges. This vegetation is dense, has higher species diversity and contains some large native trees. A large number of invasive species are also present in this vegetation type (Appendix I Table 2, Photo 4c&d). Critical food- source species for mao and/or manumea observed within this vegetation type include those present in the plantation vegetation type but also:

● Tufaso (Dysoxylum samoense); ● Pua lulu (Fagraea berteroana); ● Mati (Ficus sp.); ● Laufao (Heliconia laufao); and ● Mamalava (Planchonella samoensis).

These vegetation types also provide foraging and roosting habitat for other bird species and bats (see Appendix I). The two vegetation types mapped in Figure 3 will be directly impacted by dam construction and operation. An additional area of vegetation adjacent to the dam footprint will not be removed but is likely to be modified through edge effects. Edge effects will be directly modifying the biota and are likely to result in long-term changes in their composition and structure. Unfortunately, there is virtually no data on the changing composition of forest remnants in Samoa. However, in New Zealand Microclimate effects do not extend beyond c. 20 m into the forest (Norton 2002). The

1 Insufficient detail is available on the roads and additional infrastructure for vegetation loss calculations to be made. However, we do note that valuable habitat was along the proposed roads and territorial mao pairs also occur in these areas.

microclimatic edge effect of 20 m in width was estimated for the footprint adding an additional 10 ha of impacted forest (see Figure 3). However, it is important to note edge effects are observed past this point and vegetation responses can be apparent up to 70 m into the forest for some variables (e.g. tree ferns and bryophytes) (Norton 2002). The responses of canopy invertebrates appear similar to those of vegetation, although the effect of edge on predator abundance appeared to extend into the forest for at least 100 m (Norton 2002). Studies, both in New Zealand and elsewhere suggest that edge effects of 50-100 m are a significant issue for small forest remnants in particular (Norton 2002).

a b

c d

Photo 4. The two vegetation types found within the footprint, (a&b) Mixed plantation and (c&d) mixed forest.

3.1.2 4.3 Habitat value & threat status The footprint of the site was assessed for the presence of critical habitat, as defined by the ADB SPS (ADB, 2009). The Alaola site is modified exotic and native forest that provides habitat to three bird species of particular concern: mao ,manumea or toothbilled pigeon (Didunculus strigirostris) and Samoan triller or miti tae (Lalage sharpei). The Alaola site is also habitat to the following range restricted bird species; Flat-billed Kingfisher (Todiramphus recurvirostris), Samoan Whistler ( flavifrons), Samoan (Rhipidura nebulosa), Samoan Flycatcher (Myiagra albiventris), and Samoan ( atrifusca) as well as the Red headed parrot finch (Erythrura cyaneovirens). Of these there is also some concern for the Samoan Broadbill and the Samoan red headed parrot finch. The threat status of these species is described in Table 4. The Samoan broadbill has a very small range and a small population and is suspected to have declined owing to small-scale habitat loss for shifting agriculture (BirdLife International 2019). The Samoan populations of the newly defined red headed parrot finch E. cyaneovirens are widespread, though uncommon. Based on known records, the Samoan sub- population sizes are likely more than 1,000 mature individuals (BirdLife International 2019). The flat billed king fisher has an unknown unquantified population trend. However, the population is not believed to be decreasing rapidly and it is not believed to approach the thresholds for Vulnerable. The Samoan whistler also appears to have a stable population trend though the population size has not been quantified (BirdLife International 2019). The Samoan fantail and the Samoan starling both have a restricted range, however neither are believed to approach the thresholds for Vulnerable under the range size criterion and both population appear stable (BirdLife International 2019). The two flying fox species (Table 4) both occurred in the footprint of the site. Because the Samoan flying fox is near threatened because it is range restricted it needs special consideration, in contrast the Tongan fruit bat’s status is of less concern. Table 4. Threat classification listing for all species of concern.

Group Common name Scientific Threat Reference Ecological name classification value

Birds Manumea, Didunculus Critically BirdLife Very high Toothbilled strigirostris Endangered International pigeon (2019)

Mao, Mao mao Gymnomyza Endangered BirdLife Very high samoensis International

(2019)

Samoan triller, Lalage sharpei Near BirdLife High Miti tae threatened International (but likely to (2019) be upgraded) Samoan Myiagra Near BirdLife High broadbill/Flycat albiventris threatened International cher (2019)

Flat-billed Todiramphs Least concern BirdLife Low Kingfisher recurvirostrs International (2019)

Samoan Pachycephala Least concern BirdLife Lowmedium Whistler flavifrons International (2019)

Samoan parrot E. c. Near BirdLife Medium finch gaughrani threatened International (2019)

Samoan Fantail Rhipidura Least concern BirdLife Low nebulosa International (2019)

Samoan Aplonis Least concern BirdLife Low Starling atrifusca International (2019)

Mega Tongan fruit bat Pteropus Least concern IUCN (2012) Low tonganus bats Samoan fruit Pteropus Near Brooke, A. & High bat samoensis threatened Wiles, G. IUCN (2008)

3.1.3 4.4 Mao The 5 acoustic sound recorders placed within the footprint all contained mao calls (n=5) (Fig. 4). In addition territorial pairs were found throughout the footprint of the site and the surrounding area over the survey period (Fig.5). One pair has the core of its territory where the dam wall would be built. Mao has extremely high territorial site fidelity and therefore I would expect this pattern to be consistent. A large number of the larger trees, invasive and native, contained mistletoe a plant which is strongly correlative with mao occurrence (Photo 5a). No breeding behaviour was identified

during the field survey. However breeding occurs from April-Nov and only a single chick is produced in a season. Therefore it is not unlikely breeding may have already have occurred before the survey for this study was undertaken. The Vaisigano catchment is one of five “first priority” sites for conservation of mao in Samoa. Only three of the five key sites have had recent mao sightings, in the Savaii uplands, at Lake Lanoto’o/Malololelei and in the Vaisigano catchment. Only lake Lanoto’o/Malololelei and the Vaisigano catchment are in Upolu. This makes these sites of particular importance for maintaining this species on Upolu. Mao are known to occupy and breed within the Vaisigano catchment, and both adults and juvenile birds have been observed within or near the site at O le Fale o le Fe'e. Successful breeding is also known to occur in the eastern Vaisigano catchment where a long term study was made of the species. The Lake Lanoto’o and nearby Malololelei site has only a small area of suitable habitat for mao and can only maintain a few territories. The proposed project area is an important site where mao are potentially successfully breeding. Given the limited distribution of this species, and how few large native forest canopy trees remain on Upolu, any modification of habitat must be carefully managed to avoid adversely impacting the population.

Photo 5. Critical habitat a) for Mao Mistletoe in invasive albizia tree are and b) ficus fruit from the Aoa a favourite food source of the Manumea.

3.1.4 4.4 Manumea Field work conducted during as part of this survey did not detect manumea within the project site. This is a cryptic species with extremely low detectability and survey did cover the fruiting period within the catchment. Manumea move around the landscape tracking the availability of fruit which can differ markedly in phenology across the landscape. Critical manumea habitat was identified in the Alaoa footprint with multiple large mature moata, a favoured feeding tree identified. During the survey period Aoa (Ficus prolixa) (Photo 5b), Moata (Dysoxylum moata), (Syzygium dealateum) and (Myristica inutilis) were found to be fruiting at low levels within 2 km of the footprint. However, in the Alaola footprint no trees with ripe fruit were found during the survey. Because the trees in the footprint were not fruiting it is unlikely that manumea will be using the site for feeding during the survey period. The absence of manumea sightings during the survey therefore cannot exclude the presence or absence of manumea from within the project site. Manumea have been observed outside this survey period within 1-2km of the footprint (See Fig. 6). It is likely that manumea would be feeding on species, such as moata, within the footprint during the fruiting season in this location. Manumea have been detected within the Vaisigano catchment. Adult birds have also been observed flying in and of the Vaisigano catchment across Cross Island Road probably so they can access various mature fruits on which they rely. Manumeau have also been observed east and west of the Vaisigano catchment; at Tiapapata, Lake Lanoto’o and O le Fale o le Fe'e and by hunters directly above the proposed site. The project site was found to contain mature Dysoxylum trees, an important food source of the species. Dysoxylum were locally common. It is therefore likely that the birds feed within the project footprint when these trees are fruiting.

This valley area is also important for manumea because of its landscape value. Cyclone Evan in 2012 completely stripped trees of leaves and fruit strongly impacted food supply both pigeons and fruit bats across most of the landscape however valleys acted as refugia sites (Stirnemann pers com). The footprint is expected to be especially important for food supply for pigeons, such as manumea, after severe cyclones.

3.1.5 4.5 Other bird species A single Samoan triller was heard in the footprint of the dam. The habitat within the footprint is likely to be appropriate habitat for this species to breed. This endemic species occurs in forested areas, forest edges and clearings (Watling, 2004). Playback where the Samoan triller was previously found this species in only two sites (n=10). This species has become increasingly rare in Samoa and particularly in Upolu. However it appears to respond positively to pest control (Stirnemann pers com). The Samoan broadbill was also found within the footprint and is likely to be breeding within the footprint. In addition to the target birds of interest the footprint also supports breeding congregatory brown noddies (Anous stolidus) as well as white terns along the end of the Eastern arm of the Vaisigano river within the footprint (Fig. 5). A cave of breeding White-rumped Swiftlet (Aerodramus spodiopygius) is also located along this arm of the river (Fig. 5). A few Samoan broadbill were seen and heard during the survey period both inside and nearby the footprint. Calls indicated territorial pairs. The red headed parrot finch was also seen in the footprint and has been previously been observed nesting within 1km from the site. Flat billed king fisher, Samoan fantail and Samoan were also seen in and around the site and footprint. No freiendly ground doves were observed in the site or footprint area.

3.1.6 4.6 Bats Both Samoan and Tongan fruit bats were observed in the footprint and surrounding area. Twelve observations of Samoan fruit bats were made in the footprint over the survey period. This is not surprising given that they are often seen breeding in the fale O’le fe’e site less than 2km from the site. No bats were seen feeding in the footprint but given that none of the fruiting trees were ripe this was also not surprising.

4. 5 Project effects 4.1.1 5.1 Summary of adverse ecological effects Table 5 outlines the primary effects of the Project on key ecological values in relation to the relevant SPS criteria. The ADB SPS (ADB, 2009) also requires more stringent criteria to be met in order for a project to proceed within an area of critical habitat. In the absence of effective management, the clearance of vegetation and loss of habitat will have both direct and indirect effects on endangered avifauna. Given the footprint under consideration is critical habitat the ADB SPS (ADB, 2009) requires stringent criteria to be met in order for a project to precede within an area of critical habitat.

The application of the EIANZ guidelines (EcIA) provides guidance as to the level of adverse effect on target avifauna as a consequence of the development. The EcIA has been used to ascertain the following:

● The level of ecological value of the affected ecological values; ● The magnitude of ecological effect from the proposed development on the key fauna species and their habitats; ● The importance of the ecological effect; and ● The overall level of effect to quantify the extent of mitigation and biodiversity offsetting or compensation required.

When mitigation is applied, based on the EIANZ guidelines, the overall level of effects of the project on mao and manumea is likely to be ‘Very High’ (‘Very High’ ecological value x ‘High’ magnitude of effects – referring to Table 2). ‘Very high’ represent a high level of effect on ecological or conservation values and warrant avoidance and/or extremely high intensity mitigation and biodiversity offset or compensation remediation actions. In addition, mitigation measures should be designed to achieve at least no net loss of biodiversity. It is recommended that the ‘residual effects’ associated with the loss of habitat for key avifauna and bat species associated with the construction and operation of the dam are addressed using the 10 biodiversity offset principles outlined by the Business and Biodiversity Programme Offsets Programme (BBOP 2009) as a core objective. Under the principles of the BBOP programme, mitigation for the indigenous vegetation and fauna habitat loss should result in a minimum standard of no net loss of ecological value. Further, any mitigation should preferably result in a net gain for all biodiversity components affected by the Project in terms of potential and actual adverse effects on habitats and threatened species. Accordingly, a comprehensive mitigation and offsetting/compensation package will be needed to compensate for the unavoidable loss of habitat associated with the Dam’s construction and operation. Table 5. Critical habitat assessment for the proposed development

SPS Criteria Existing information Is Criterion met? (ADB, 2009) Areas of high - Botanical values are moderate to low. Yes it has high biodiversity Vegetation is mixed exotic with some large biodiversity value emergent native trees including Dysoxylum sp and tawa. - Avifauna biodiversity is high. One endangered bird species occurs within the project area (mao), and one critically endangered bird species is likely to use the project area (manumea). The Vaisigano catchment represents one of the few areas where mao and manumea remain on Upolu. The area is also used by other species which have near threatened threat rankings including the Samoan

triller, red headed parrot finch, the Samoan fruit bat and the Samoan broad bill.

Habitat One endangered and one critically endangered Yes required for bird species is present within the Vaisigano the survival of catchment. Mao: are found throughout the critically footprint and the surrounding area. The recovery endangered or plan states that mao may be present in smaller endangered numbers than the manumea. species Manumea: have been recorded near the proposed site. Appropriate habitat for the species is present in the footprint. The Vaisigano catchment is identified as a “key site” for Manumea. Area having As above. The area is also used by other species Yes special which have near threatened threat rankings significance including the Samoan triller, the Samoan fruit bat, for endemic or red headed parrot finch and the Samoan broad restricted bill. range species It however does not have particular significance to these species apart from being habitat that they utilise. Sites that are No evidence was found that the site is critical for No evidence was critical for the migratory species. found that the site is survival of critical for migratory migratory species. species

Areas Brown noddies and white terns breed in the area. No. supporting The site does not globally No globally significant concentrations of support globally significant congregatory species were identified. significant concentrations concentrations of or numbers of congregatory congregatory species. species Areas with No evidence was found that the site supports a No. unique unique assemblage of species. The site does not to assemblages No evidence was found that the site is not my knowledge of species or associated with key evolutionary processes. support unique that are assemblages of associated species or key with key evolutionary evolutionary processes. processes or provide ecosystem services. Areas having Cultural: the Vaisigano catchment is identified by Yes. biodiversity of MNRE as a “key site” for the conservation and The catchment in significant long term survival of manumea. The proposed which the project is social, development is situated within this catchment. situated has been economic or The manumea is the national bird of Samoa and identified as a key cultural of chiefly importance. conservation site for importance to the critically local endangered communities manumea, the national bird of Samoa, and has significant cultural importance to the community. Areas either The site has no legal protection. Yes. legally Mao and manumea are two trigger species for the protected or classification of this area as a KBA. officially proposed for protection

4.1.2 5.2 Effects on vegetation Approximately 24.4 hectares of vegetation clearance would occur to construct the proposed dam. This calculation does not include the additional clearance to increase roads and for additional infrastructure. A full assessment of clearance could be

calculated once this information is available. The effects of vegetation clearance will be the localised loss of native and exotic trees, shrubs and groundcovers, loss of habitat, and reduction in habitat values. Indirect effects of vegetation clearance will include increased edge effects caused by the creation of additional cleared forest. Edge effects are changes to vegetation and community composition caused through increased sunlight, wind, temperature and associated disturbance caused by the development. This can impact larger areas than the actual zone of clearance, and cause weed and pest invasion, and windblown trees. The significance of edge effects will largely be determined by the degree of disturbance in the forest interior. Because of the predicted edge effects and in particular the increased invasive weed penetration expected, additional mitigation or restoration efforts should be factored in to mitigation activity. An associated indirect effect of vegetation clearance will be increased access for people, which can potentially result in further forest clearance, logging, hunting pressure and agriculture. This impact will be determined by the ease of access to the forest both in the short and long term. Loss of forest habitat during construction can be minimized by reducing the width of forest clearance for construction and adjusting the alignments to avoid as many mature native trees as possible, particularly those that are essential food sources for the avifauna of concern. Strict security measures could potentially reduce but cannot eliminate the risk of further clearance from site workers or the local community. Replanting of native vegetation prior to, during and upon completion of construction can mitigate for the removal of vegetation. The earlier the planting was to occur the more that the time lag effect (discussed below) would be reduced. Other than areas occupied by essential infrastructure, all disturbed areas should be replanted on completion of construction with native trees. In addition, planting could potentially occur in other parts of the catchment, or in the adjacent catchment, to create new habitat in areas where vegetation has historically been cleared or been invaded by invasive species such as rubber trees.

4.1.3 5.3 Effects on avifauna As discussed in section 5.1 above, the overall level of the project effects for mao and manumea is likely to be ‘Very High’ (‘Very High’ ecological value x ‘High’ magnitude of effects – referring to Table X) without mitigation. The overall effect on the Samoan thrillers and broad bill is likely to be moderate to low even without mitigation. This difference is because this habitat is likely to be less critical for these species and the majority of their populations are likely to be outside the footprint. Each of the effects on manumea and mao is addressed in detail in the tables below along with possible mitigation or offsetting measures for each effect (Tables 6 and 7). Appropriate design can assist in reducing the environmental impact of the Project. The measures taken to reduce the environmental footprint of the proposed project could in part be through optimisation of site selection and design. Monitoring will be required to ensure that mitigation measures are effective. The mitigation package would have to be carefully designed and implemented with appropriate government and village consultations to ensure buy in and support for proposed activities. Careful consideration would also be needed in regards to the effects of timelags since mitigation activities such as tree planting can take >5 years prior to fruit being available as alternative food sources. This consideration is a key outstanding impact which would require further detailed information on the proposed construction and vegetation time scale plans.

Standard environmental management and replanting of cleared areas is insufficient to mitigate for the loss of habitat for threatened and endangered species. To enable the clearance of habitat known to be used by the mao and manumea, the remaining habitat must be protected and enhanced to increase the carrying capacity for these birds. Table 6. Description of the magnitude of effect on Mao and changes with recommended mitigation applied Effect on mao Magnitude of Possible Suggestions for mitigation/offsetting impact by the timescales proposed for duration development of effects

without mitigation

Loss of home range Very High Permanent Long term pest control (rats, cats) breeding habitat that in new source sites has a key life cycle Restoration of habitat at alternative function and sites (Time lag impact) associated population effects Where possible leave large trees

and habitat.

Development of nursery with more native trees

Move some large trees to an alternative site.

Avoidance of some key large trees

Loss/ Mortality of High Long term Establish a new source population adult birds and and aid breeding chicks Translocation of birds to new sites e.g. American Samoa

Avoidance of some key large trees

Reduction in habitat Very high Permanent Restoration to increase potential and feeding area on habitat including mistletoe areas site Avoidance of some key large trees

Loss of buffering Very high Permanent Restoration of habitat and and loss of extent extensive weed control especially of key habitat and in adjacent areas increased invasion of weedy plants

Fragmentation and High Permanent Ensure that fragmentation loss of connectivity is reduced through to other sites connective restoration

Increase in threats medium Permanent Long term pest control for cats and from pest species beside sites rats

Reduction in Very High Permanent Ensure restoration and pest control access to valley occurs in valley sites which are refugia sites with refugia cyclones Disturbance e.g. Medium- The Temporary Limit disturbance as much as through increased clearance of possible in key areas. human access, vegetation, construction combined with vehicles, noise the noise and activity from construction activities, and presence of workers, will lead to the localised disturbance and loss of habitat.

Table 7. Description of the magnitude of effect on manumea/tooth billed pigeon and changes with proposed mitigation Effect on Manumea Magnitude of Possible Suggestions for impact by the timescales mitigation/offsetting proposed for duration development of effects without mitigation

Loss of home range Very High Permanent Long term pest control (rats, cats) breeding habitat that in potential source sites. has a key life cycle function and Restoration of habitat at associated population alternative sites (Time lag effects impact)

Where possible leave large trees and habitat.

Development of a nursery with more native trees

Move some large trees to an alternative site. Mortality of territorial High Permanent Improve breeding success and adult birds with habitat survival with pest control. loss

Reduction in habitat Very high Permanent Restoration of habitat to increase and feeding area on potential habitat site

Loss of buffering and Very high Permanent Restoration of habitat and weed loss of extent of key control especially in adjacent habitat and increased areas invasion of weedy plants

Fragmentation and High Permanent Ensure that fragmentation loss of connectivity to is reduced through other sites connective restoration

Increase in threats medium Permanent Long term pest control for cats, from pest species rats and pigs

Reduction in access to Very High Permanent Ensure restoration and pest valley refugia sites control occurs in 2-3 valley sites with cyclones which are considered refugia sites

Hunting effects Very High Permanent- Strengthen pigeon hunting bans temporary in collaboration with hunters and the villages

Employ hunters to undertake pest control and be village advocates for the restoration project

Disturbance e.g. Medium- The Temporary Have protocols in place to reduce through increased clearance of disturbance as much as possible. human access, vegetation, construction vehicles, combined with the noise noise and activity from construction activities, and presence of workers, will lead to the localised disturbance and loss of habitat.

Table 8. Description of the magnitude of effect on other passerines and changes with proposed mitigation Effects on other passerines Magnitude of impact Suggestions for by the proposed mitigation/offsetting development without mitigation

1) Loss of home range and Medium-low Long term pest control breeding habitat that has a (rats, cats) in potential key life cycle function and source sites. associated population effects as well as fragmentation of Restoration of habitat at habitat alternative sites (Time lag (permanent) impact) Disturbance during 2) Development of a nursery construction with more native trees (temporary) 3) Habitat reduction and Where possible leave large increased invasion of weedy trees and habitat. plants (permanent) 4) Increase in threats from pest species (permanent) 5) Loss of individual breeding pairs

4.1.4 5.4 Effects on bats The critical effect on fruit bats is likely to be through loss of feeding habitat and loss of protected refugia area during cyclones. The loss of these sites will be permanent. Though both Tongan and Samoan bats utilise were seen in the study footprint are the species of concern is the Samoan fruit bat. The Tongan bats are listed as of least concern and are not range restricted. Possible mitigation could include habitat restoration and pest control to increase food resources especially during the time lag period. Restoration should include restoration which targets key food trees. Restoration sites should include restoration in a valley refugia sites. Enhancing habitat could also be effective by controlling invasive plants such as rubber trees.

5. 6 Avoidance, remediation, mitigation and offsetting/compensation requirements

5.1 6.1 Mitigation Strategy As stated above the mitigation of adverse effects associated with the Alola Dam project on avifauna and bat habitat is such that there will be significant adverse residual effects which cannot be directly avoided, remediated or mitigated. These residual effects will require design and implementation of a substantial biodiversity offsetting or compensation package to ensure there is no net loss of biodiversity values. BBOP (2009) have outlined 10 key principles which should be applied to projects where this situation occurs (refer to Table 9 below). It is recommended that a detailed offset/compensation package is designed in accordance with these principles. Quantification of the type, extent and duration of the offset package will

require further detailed application of offsetting modelling and preparation of several specific mitigation and monitoring plans. Table 9 BBOP Biodiversity Principles

BBOP Principles

Principle 1 No net loss: A biodiversity offset should be designed and implemented to achieve in situ, measurable conservation outcomes that can reasonably be expected to result in no net loss and preferably a net gain of biodiversity.

Principle 2 Additional conservation outcomes: A biodiversity offset should achieve conservation outcomes above and beyond results that would have occurred if the offset had not taken place. Offset design and implementation should avoid displacing activities harmful to biodiversity in other locations.

Principle 3 Adherence to the mitigation hierarchy: a biodiversity offset is a commitment to compensate for significant residual adverse impacts on biodiversity identified after appropriate avoidance, minimisation, and on-site rehabilitation measures have been taken according to the mitigation hierarchy.

Principle 4 Limits to what can be offset: There are situations where residual impacts cannot be fully compensated for by the biodiversity offset because of the lasting vulnerability of the biodiversity affected.

Principle 5 Landscape context: A biodiversity offset should be designed and implemented in a landscape context to achieve the expected measurable conservation outcomes taking into account available information on the full range of biological, social, and cultural values of biodiversity and supporting an ecosystem approach.

Principle 6 Stakeholder participation: in areas affected by the development project and by the biodiversity offset, the effective participation of stakeholders should be ensured in decision- making about biodiversity offsets, including their evaluation, selection, design, implementation and monitoring.

Principle 7 Equity: A biodiversity offset should be designed and implemented in an equitable manner, which means the sharing among stakeholders of the rights and responsibilities, risks, and rewards associated with a development project and offset in a fair and balanced way, respecting legal and customary arrangements. Special consideration should be given to respecting both internationally and nationally recognised rights of indigenous people and local communities.

Principle 8 Long-term outcomes: the design and implementation of a biodiversity offset should be based on an adaptive management approach, incorporating monitoring and evaluation, with the objective of securing outcomes that last at least as long as the development project’s impacts and preferably in perpetuity. Criterion 8-2: Adaptive monitoring and evaluation approaches shall be integrated into the Biodiversity Offset Management Plan to ensure regular feedback and allow management to adapt to changing conditions and achieve conservation outcomes on the ground. Indicator 8-2-1: A risk-monitoring protocol is in place and followed to identify any risks (such as climate change, population pressure, land-use change) that could affect achievement of proposed conservation outcomes. Indicator 8-2-2: Offset conservation outcomes and milestones are independently audited and project responds to audit recommendations in a timely manner. Indicator 8-2-3: Monitoring and evaluation protocols provide regular feedback on implementation

progress and results and are used to document, correct, and learn from problems (e.g. adaptive management). Guidelines for assessing and auditing biodiversity offset management plans against BBOP PCI criteria are available (see reference).

Principle 9 Transparency: The design and implementation of a biodiversity offset, and communication of its results to the public, should be undertaken in a transparent and timely manner.

Principle 10 Scientific information and, where applicable, traditional knowledge shall be utilised when designing and implementing the offset.

A number of measures are required to avoid, remedy or mitigate potential adverse effects on terrestrial ecology. These include:

• A Vegetation Clearance and Restoration Management Plan; • An Avifauna/bat Management and Monitoring Plan; • Weed hygiene and surveillance protocols; and • Pre and Post-construction monitoring of key bird and bat species adjacent to the Dam footprint with suitable trigger points to provide contingency habitat enhancement if unforeseen adverse effects are shown to occur. The details of these mitigation and monitoring measures can be given effect by appropriately drafted conditions, and should be prepared before construction commences. In addition to measures to avoid, remedy or directly mitigate for potentially affected flora and fauna and their habitats, two Mitigation and Offset packages are recommended to offset any residual adverse effects on ecologically significant terrestrial fauna and flora and their habitats. These are: A Alola Dam Restoration Buffer and Catchment Restoration Plan; and B Alola Dam Threatened Species Habitat Protection and Enhancement Plan. The above programmes would result in a number of significant ecological benefits, including: Intensive, targeted pest control in forest habitat within the catchment, habitat restoration and enhancement around the new dam and within its sub-catchment; assisting landowners to protect and manage forest, and habitat would need to be restored for mao and manumea as well as the other passerines and bats effected; and developing and ongoing monitoring programme contributing towards projects that will enhance the knowledge of mao, Manumea, other threatened bird species and Samoan fruit bats, as well as their habitats.

5.1.1 6.2 Vegetation mitigation

Recommended management and mitigation measures should include:

1. Vegetation and tree clearance shall be minimised to the extent necessary to allow for the construction; 2. The site office and works yard should be established in existing open, already modified areas avoiding the unnecessary clearance of vegetation; 3. All construction access roads should utilise existing access roads as much as possible, to avoid the unnecessary clearance of additional vegetation/habitat; 4. The design and construction of the dam should avoid as many mature native trees (especially Dysoxylum sp.) as possible and determine if it is possible to relocate mature trees. 5. Prior to construction commencement, all mature native trees within the project footprint should be identified, clearly marked and protected to avoid damage and works affecting habitat unnecessarily;

6. Vegetation clearance should be conducted outside of the main nesting period for mao. Immediately prior to any clearance a survey by an experienced expert who can make sure Mao breeding is not occurring during the activity should occur. 7. Native trees (including Dysoxylum sp.) and shrubs should be replanted in along the construction corridor to revegetate bare soil and replace vegetation removed prior to construction. 8. A Restoration Plan should be prepared for the site specifying the number, species, size and location of native trees and shrubs to be planted. 9. Restoration plantings should be native trees and shrubs that would naturally be found within the catchment. 10. Invasive species shall not be used for replanting. 11. Restoration plantings shall utilise a diversity of native species, but include no less than 20% of the various Dysoxylum sp found in the catchment. 5% of all trees planted should be aoa and 15% other species which are beneficial feeding habitat to manumea and 15% additional feeding habitat beneficial to mao. Restoration planting should include enhancement planting of epiphytes such as native mistletoe. A step wise restoration program is therefore likely to be needed. A plan should also be developed for enhancing habitat this should include:

1. Planting maintenance for a minimum of ten years, by way of monthly checks for survival, manual or chemical weed and invasive tree control, and replacement planting. 2. Felling of vegetation and planting of crops by construction workers shall be prohibited. 3. The Contractor shall maintain strict security measures to prohibit access to the site and surrounding catchment outside normal hours of construction. 4. Following the completion of construction, the sides of all access roads should be distributed with topsoil and revegetated with native trees and plants. Any access roads retained should be the minimum width required to allow for continued maintenance of the site.

5.1.2 6.3 Avifauna mitigation & monitoring

The preparation of an Avifauna Management and Monitoring Plan (AMMP) would be needed for the project to cover monitoring, pest control, education and adaptive management for mao and manumea. The AMMP plan should include:

1. Details of a survey plan prior to construction, during the works, and post-construction, it would also include a survey for mao and manumea distribution within the entire catchment. This is to determine the distribution

and abundance of birds and to monitor the overall impact of development on these species.

2. For in the breeding seasons prior to construction commencement, during the works, and post-construction, undertake a comprehensive pest control programme, focussing on three species- rats, cats and pigs. For mao reducing rat numbers and protecting nests (through trapping and steel ringing trees), reproductive success and adult and fledgling survival will be increased. This will lead to increased recruitment and could offset some of the effects of the proposed development. Reducing numbers all three pest species is likely to be beneficial for Manumea since it will increase fruiting availability and potentially increasing reproductive success. Other species species of concern will also be positively affected with increased reproductive success by pest control, especially as a result of rat targeted control.

3. Vegetation clearance should be conducted at the beginning of the dry season and outside of the main nesting period for mao (June to October (Stirnemann,2017)).

4. No construction activity or vegetation clearance should occur within 150 metres from an occupied nest of an endangered or critically endangered species.

5. Engagement and education of the local community. This could include the employment of members of the community in planting, pest control and monitoring activities, providing an economic incentive to protect the birds.

6. In areas buffering the site habitat enhancement should also occur by reducing invasive trees e.g. rubber trees

5.1.3 6.3 Bat mitigation & monitoring

Habitat would also need to be enhanced and restored for the Samoan fruit bat targeting appropriate habitat restoration. Mitigation should be targeted for another potential valley refugia site. Calculations on the extent of restoration needed would have to consider the timelag needed for plants to reach fruiting maturity. For this reason a 1-1 ratio is not considered appropriate. Hunting of bats is strongly related to access to sites. There are a number of ways that this could be tackled. These have been reviewed for Manumea and pigeons in Stirnemann (2017)but these techniques all also apply to reducing fruit bat consumption and hunting.

6. References

ADB, (2009) Safeguard Policy Statement

Birdlife International. (2012). Gymnomyza samoensis. The IUCN Red List of Threatened Species. [accessed 19 September 2019] Business and Biodiversity Programme Offsets Programme (BBOP) 2009. Biodiversity Offset Design Handbook, BBOP, Washington, D.C., ISBN 978-1- 932928-31-0. Butler, D. (2013). CEFP Final Project Completion Report. Nelson: David Butler Associates Ltd. Butler, D., & Stirnemann, R. (2013). Lessons Learned Technical Series 25: Leading the recovery of two of Samoa’s most threatened bird species, the tooth-billed pigeon (Manumea) and the mao (Ma’oma’o) through ecological research to identify current threats. Apia: Conservation International. Conservation International et al. (2010). Priority Sites for Conservation in Samoa: Key Biodiversity Areas. Apia: Conservation International, Ministry of Natural Resources and Environment, Secretariat of the Pacific Regional Environment Programme IUCN. (2012). IUCN Red List of Threatened Species. Gland: International Union for Conservation of Nature and Natural Resources. MNRE. (2006). Recovery plan for the Ma’oma’o or Mao (Gymnomyza samoensis): Samoa’s largest forest : 2006-2016. Ministry of Natural Resources and Environment, Apia MNRE (2006) Recovery plan for the manumea or tooth-billed pigeon (Didunculus strigirostris): 2006- 2016. Ministry of Natural Resources and Environment, Apia

MNRE. (2009). Convention on Biological Diversity: Samoa's 4th National Report 2009. Apia: Ministry of Natural Resources and Environment.

MNRE. (2014). Conservation Leadership Program Project Final Report Project ID: 06141613 Samoa's Little Dodo - Saving the Tooth-Billed Pigeon. Apia: Ministry of Natural Resources and Environment.

SPREP, (2012) Rapid Biodiversity Assessment of Upland Savai’i, Samoa, SPREP

Stirnemann R. L. , I. A. Stirnemann, D. Abbot, D. Biggs & R. Heinsohn, (2017). Interactive impacts of by catch take and elite consumption of illegal wildlife, Biodivers Conserv Stirnemann R. L. , Potter, M., Butler, D., and Minot, E., (2015) Compounding effects of habitat fragmentation on bird’s nests, Austral Ecology 40: 974–981. Stirnemann R. L. , Potter, M., Butler, D., and Minot, E., (2016a) The breeding biology of the Mao an endangered large tropical forest passerine, Bird Conservation International: 1-14 Stirnemann R. L., Potter, M., Butler, D., and Minot, E., (2016b) Slow life history traits in an endangered tropical island bird, the Ma’oma’o, Bird Conservation International Wattling, D. (2003). A Guide to the Birds of Fiji and Western Polynesia: Including American Samoa, , Samoa, Tokelau, Environmental Consultants, Pg 272

7. Appendix I: Vegetation types - species lists

Appendix I Table 1. Partial species list, phenology and habitat importance for the Mixed Plantation: Predominantly plantation with some large indigenous and exotic trees Phenological Scientific name Common name observations on Importance footprint Some flowering Contains mistletoe Decaisnea Albizia sp. Tamaligi on site sp. Critical food source for Mao Carica papaya Papaya Panama rubber Castilla elastica tree Cestrum nocturnum Ali'I o le po Cocos nucifera Niu Food source for Mao (pers com Decaisnea sp. Mistletoe Some Flowering Stirnemann pers obs) Food source for Manumea Dysoxylum sp. Maota (Stirnemann) Mountain ginger, Some flowering Food source for Mao (pers com Etlingera cevuga avpui vao on site Stirnemann) Food source for Mao pers com Erythrina variegata Gatae (Stirnemann) Merremia peltata Musa sp. Fa'i Homalanthus nutans Mamala Food source for Pigeons Theobroma cacao Cacao

Appendix I Table 2. Partial species list, phenology and habitat importance for Secondary Forest: Predominantly forest with a mixture of indigenous and exotic trees Scientific name Common name Phenological Importance observations within footprint Albizia sp. Tamaligi Some Contains mistletoe Decaisnea flowering sp. Food source for Mao (Stirnemann pers obs) Amyema artensis Tapuna Angiopteris evecta Giant fern Cananga odorata Moso’oi Flowering Pigeons eat fruit (Whistler 2004) Canarium vitiense Ma’ali

Castilla elastica Panama rubber tree Cinnamomum verum Tinamoni Cordyline fruticosa Ti vao Decaisnea sp. Mistletoe Some Food source for Mao (Stirnemann Flowering pers obs)

Dysoxylum maota Maota Fruiting Food source for Manumea nearby (Stirnemann pers obs) Dysoxylum samoense Tufaso Food source for Manumea (Stirnemann pers obs) Etlingera cevuga Mountain ginger, Flowering on Food source for Mao (Stirnemann avpui vao site pers obs) Fagraea berteroana Pua lulu Fruiting and Fruit and flowers eaten by Mao flowering (Stirnemann pers obs) Ficus sp. Mati Fruiting An important food source for nearby Manumea (Stirnemann pers obs) Flacourtia rukam Filimoto Fruits are eaten by pigeons (Whistler 2004) Flueggea flexuosa Poumuli Fruits are eaten by birds (Whistler 2004) Funtumia elastica African rubber tree Heliconia laufao Laufao Flowers and fruit food source for Mao (Stirnemann pers obs) Morinda citrifolia Nonu Nectar is favourite food of (Whistler 2004) Myristica inutilis ‘Atone Fruiting Fruit eaten by pigeons (Whistler 2004) Homalanthus nutans Mamala Pigeons sometimes eat the leaves (Whistler 2004) Palaquium stehlini Gasu Fruit eaten by bats (Whistler 2004) Planchonella garberi ‘Ala’a Fruits are an important food source for pigeons and bats (Whistler 2004) Planchonella samoensis Mamalava Fruits are a major food source for pigeons and bats (Whistler 2004) Pometia pinnata Tava Syzygium dealatum Asi vai The seasonal fruits are eaten by pigeons and doves (Whistler 2004) Syzygium inophylla Asi toa Major food source for bats and pigeons (Whistler 2004)

Syzygium samarangense Nouveau Fruiting Eaten by pigeons and bats (Whistler 2004) richii Malili Flowering Fruits are valuable food for near site pigeons (Whistler 2004) entry Trichospermum richii Makosina

8. Appendix II

Additional Information on Threatened and At Risk Flora and Fauna Species Recorded within a 10 km Radius around the Scheme Area.

Photo 6. An example of high quality habitat found within 3km of the footprint.

High quality habitat is found within 2km from the proposed site (Photo 6). Within the 2 km area is also some large and likely old trees which should be protected this including large Tawa trees. One key tree is over 9 meters in diameter.