Appendix 5-4

OAHU ARMY NATURAL RESOURCES PROGRAM MONITORING PROGRAM

VEGETATION MONITORING RESULTS FOR THE ACHATINELLA MUSTELINA ESU-E ENCLOSURE AT PALIKEA, 2018

INTRODUCTION

The Oahu Army Natural Resources Program (OANRP) documents vegetation cover change at the “Palikea North” Achatinella mustelina ESU-E predator resistant enclosure at Palikea Management Unit (MU) (Figure 1) as a measure of vegetation rehabilitation success in association with the enclosure’s restoration plan (OANRP 2017). The goal of the restoration plan is to achieve a native dominated community favorable for A. mustelina as well as Drosophila substenoptera, Drosophila montgomeryi, and Chasiempis ibidis. The primary objective of monitoring is to asses if vegetation cover goals are met. Native vegetation cover goals include:

>50% for 0 – 1 m above ground level (AGL) after one year and beyond >50% for 1 – 2 m AGL after two years and beyond >50% for >2 m AGL by five years, >75% by 10 years, and >90% by 15 years >75% for total AGL cover by two years and beyond

Figure 1. Location of the Palikea North Snail Enclosure at Palikea MU.

A secondary objective is to assess if weed cover goals in the plan are met. Goals include maintaining < 10% weed cover, and having zero tolerance for a number of species (Blechnum appendiculatum, Ehrharta stipoides, Nephrolepis brownii, Paspalum conjugatum, and Drymaria cordata). Other taxa are specified as control targets (Clidemia hirta, Cyclosorus dentatus, Cyclosorus parasiticus, Passiflora edulis, Passiflora suberosa, Phytolacca octandra, Psidium cattleianum, Rubus rosifolius, and Schinus terebinthifolius, as well as non-native grasses and Asteraceae). Vegetation monitoring occurred in April 2016 prior to non-native vegetation removal and enclosure wall construction (OANRP 2016), in September 2017 following completion of the enclosure wall, and in September 2018 one year post- construction completion. Monitoring will occur annually for the first five years post-construction, after which the interval may be extended to every two to three years.

METHODS

Percent cover: Point intercept monitoring was used to assess changes in percent cover of native and non-native taxa in the understory and canopy. All species “hit” at points along transects were recorded for understory and canopy vegetation. A 5 millimeter diameter pole was used to determine “hits” in the understory (live vegetation that touches the pole, including , branches and trunks) along an outstretched measuring tape at regular intervals. Vegetation “hits” in the understory were recorded from 0 – 1 m AGL and 1 – 2 m AGL. A laser pointer held against the pole was used to determine laser “hits” in the canopy (above 2 m AGL) at these same intercept points, where the point fell within the perimeter of a ’s canopy. Locations where no vegetation was intercepted were recorded as non-vegetated. Approximately 500 points were planned based on a priori analysis of a sample size necessary to detect a 10% change with a power of 0.90 using G* Power Version 3.1.9.2. Locations of the sampled points are not permanent. Transects were spaced 5 m apart and oriented east/west, with point intercepts located every 1 m along transects (n = 542 in 2016, n = 501 in 2017, and n = 492 in 2018). Approximations of percent cover were obtained from the proportion of “hits” among all intercepts. Analysis included Pearson’s chi-square tests for cover change and Kruskal-Wallis tests with Bonferroni adjusted post-hoc pairwise comparison for canopy openness change using IBM SPSS Version 25. Only absolute cover changes > 10% were analyzed to mitigate the probability of detecting a change when none exists (Type I error), and α = 0.05 was used for significance determinations. Additional point intercepts were obtained during the 2016 monitoring (prior to wall construction) and were excluded from the analysis as they were located outside the enclosure wall. Exact wall placement had not been determined at that time.

Canopy openness: Hemispherical photography was used to monitor changes in canopy openness. Photographs were taken using a fish-eye lens at 2 m AGL, aimed 180° from the forest floor every 10 m on alternating transects in 2016, and every 5 m along all transects in 2017 and 2018. Analysis included ANOVA with Tukey’s post-hoc comparisons for differences in canopy openness over time in hemispheric photographs using Gap Light Analyzer (GLA), Version 2.0.

Supplemental data: UAV imagery was taken during the course of non-native vegetation clearing and in conjunction with vegetation monitoring following completion of the enclosure construction. Five permanent photopoints were established (marked with permanent galvanized pipe) for visual documentation of sub-canopy change in each cardinal direction at each point. Photos were taken during clearing (January 2017), post-clearing (February 2017), during the early stages of construction (May 2017), after the initial outplantings (January 2018), and after one year post-completion of the wall (October 2018). During the course of vegetation monitoring, a species diversity list was created documenting all species that happened to be observed, but not intercepted. The list will help document change in the presence or absence of species that have low cover, or are uncommon, and therefore less likely to be documented during point intercept monitoring. An Onset HOBO U23-001 data logger was installed on site in April 2018 to document hourly temperature and relative humidity, the results of which are not included in this report.

RESULTS

In 2016, prior to non-native vegetation clearing that occurred in advance of the wall construction, the enclosure area contained dense vegetation (17% mean canopy openness) with nearly continuous non- native vegetation (98% total cover), and a smaller amount of scattered native vegetation (32% total cover) (Table 1). Non-native vegetation primarily consisted of Psidium cattleianum (78%) and Schinus terebinthifolius (39%) above 2 m AGL, Clidemia hirta (20%) and P. cattleianum (16%) from 1 – 2 m AGL, and P. cattleianum (20%) and C. hirta (13%) from 0 – 1 m AGL (Table 2). Native vegetation was mainly Metrosideros polymorpha (10%) and Freycinetia arborea (5%) above 2m AGL, F. arborea (2%) from 1 – 2 m AGL, and Nephrolepis exaltata subsp. hawaiiensis (14%) from 0 – 1 m AGL. Though native vegetation cover was relatively low overall, there was twice as much native diversity compared to non-native diversity, with many native taxa anecdotally observed but not intercepted during monitoring (Table 3)

After removal of non-native vegetation and enclosure wall completion in September 2017, there was a significant reduction in non-native vegetation, resulting in 0% total non-native cover, as well as a significant increase in canopy openness (median openness = 100%; p = 0.000). However, there was also a significant reduction in native cover, to 9% total cover, primarily consisting of the remaining M. polymorpha . Though non-native cover was not intercepted during monitoring, several non-native taxa were anecdotally observed, most notably the vigorously recruiting pioneer species Phytolacca octandra. Weed cover goals were met, though two zero tolerance species and a number of control targets were among the anecdotally observed taxa.

Following the expansion of native vegetation from natural regrowth and recruitment, as well as the addition of nearly 1500 outplants, >10,000 sown seeds, and dozens of transplants (Table 4), in the first year after enclosure construction completion, there was a significant increase in native vegetation within the 0 – 1 m AGL stratum to 24% cover by September 2018, surpassing the native cover that occurred in that strata prior to weed removal and construction. While native cover was predominated by ferns in the understory strata prior to clearing (Table 5), by the end of the first year post-construction, there was a diversity of growth forms, with trees predominating, and the fast growing tree albidus most prevalent, along with N. exaltata subsp. hawaiiensis, Coprosma longifolia, Bidens torta, Kadua affinis, and Scaevola gaudichaudiana. Nearly all outplanted/sown/transplanted taxa were intercepted during monitoring, and all but one (Santalum freycinetianum var. freycinetianum, which had only four outplants, and whose fates were undetermined) were otherwise observed anecdotally. However, the native cover goal for the first year was not met. Weed cover goals were maintained with only 6.1% overall cover, all of which was in the 0 – 1m stratum. Two zero tolerance taxa persisted, and a number of new control targets were present in conjunction with an increase in non-native species diversity. While non-native taxa growth forms were primarily trees and shrubs with only a small amount of grass prior to clearing, weed cover was predominantly grass after the first year following construction completion. Canopy openness remained high and did not differ significantly from the prior year (median openness = 100%; p = 1.000).

The dramatic changes in vegetation cover following removal of non-native vegetation, and notable changes in the first year following construction completion and initiation of both passive an active restoration, are quite apparent for total vegetation cover in the UAV imagery (Figure 2) as well as visual representation of sub-canopy vegetation in photopoint images (Figures 3-7).

Table 1. Vegetation cover by stratum prior to clearing (2016), post-construction (2017), and one year post-construction (2018). Pearson's chi-square tests were used for cover change between 2016 and 2017, and between 2017 and 2018. Statistically significant results are in boldface. Arrows indicate increase (↑) or decrease (↓) in cover. 2016 2017 2018 2016 to 2017 2017 to 2018 n = 542 n = 501 n = 492 P X2 P X2 0-1 m Native 18.1 2.8 24.4 0.000↓ 63.472 0.000↑ 99.173 Non-native 37.3 0.0 6.1 0.000↓ 231.568 0.000↑ 31.500 Non-vegetated 49.8 97.2 71.1 0.000↑ 293.788 0.000↓ 127.379 1-2 m Native 6.1 1.4 2.2 0.000↓ 15.537 0.322 0.981 Non-native 36.3 0.0 0.0 0.000↓ 224.501 NA NA Non-vegetated 59.6 98.6 97.8 0.000↑ 233.407 0.322 0.981 >2 m Native 16.8 6.8 6.9 0.000↓ 24.697 0.938 0.006 Non-native 94.8 0.0 0.0 0.000↓ 936.764 NA NA Non-vegetated 2.0 93.2 93.1 0.000↑ 871.932 0.938 0.006 Total vegetation AGL Native 32.1 9.0 28.9 0.000↓ 83.902 0.000↑ 64.180 Non-native 98.2 0.0 6.1 0.000↓ 1003.722 0.000↑ 31.500

Table 2. Species percent cover by stratum and year. Pearson's chi-square tests were used for taxa with >10% absolute cover change, all of which occurred between 2016 and 2017. Native taxa are in boldface. Arrows indicate increase (↑) or decrease (↓) in cover. Species 2016 2017 2018 2016 to 2017 0-1 m n = 542 n = 501 n = 492 p X2 Acacia koa 0.0 0.0 0.2 Asplenium contiguum 0.9 0.0 0.0 Bidens torta 0.0 0.0 2.2 1 Blechnum appendiculatum 0.6 0.0 0.0 Cheirodendron trigynum 0.0 0.2 0.8 Cibotium chamissoi 0.0 0.0 0.2 Clidemia hirta 13.1 0.0 0.6 0.000↓ 75.976 Coprosma foliosa 0.0 0.2 0.0 Coprosma longifolia 0.0 0.2 3.3 Cyclosorus parasiticus 0.2 0.0 0.0 Cyperus polystachyos 0.0 0.0 1.6 1 Ehrharta stipoides 0.9 0.0 0.2 Freycinetia arborea 0.6 1.0 1.8 Kadua affinis 0.0 0.0 2.2 Metrosideros polymorpha 0.4 0.2 0.0 Microlepia strigosa 2.8 0.0 0.2 Nephrolepis exaltata subsp. hawaiiensis 14.0 1.0 3.9 0.000↓ 67.260 Oxalis corniculata 0.0 0.0 0.4 1 Paspalum conjugatum 4.4 0.0 4.1 Passiflora suberosa 0.2 0.0 0.0 sandwicensis 0.0 0.0 0.6 Phytolacca octandra 0.0 0.0 0.6 Pipturus albidus 0.0 0.0 6.5 Pisonia brunoniana 0.0 0.0 0.6 Pluchea carolinensis 0.0 0.0 0.2 Psidium cattleianum 20.3 0.0 0.4 0.000↓ 122.423 Psychotria mariniana 0.0 0.0 0.2 Rubus rosifolius 2.4 0.0 0.2 Scaevola gaudichaudiana 0.0 0.0 2.2 Table 2 (continued). Species 2016 2017 2018 2016 to 2017 0-1 m, cont. n = 542 n = 501 n = 492 p X2 Schinus terebinthifolius 0.4 0.0 0.0 Wikstroemia oahuensis var. oahuensis 0.0 0.0 0.2 1-2 m Antidesma platyphyllum 0.2 0.0 0.0 Cheirodendron trigynum 0.2 0.2 0.0 Cibotium chamissoi 0.4 0.2 0.6 Clidemia hirta 19.9 0.0 0.0 0.000↓ 119.951 Coprosma longifolia 0.0 0.6 0.2 Freycinetia arborea 2.0 0.0 0.0 Kadua affinis 0.6 0.2 0.4 Metrosideros polymorpha 0.9 0.2 0.0 Microlepia strigosa 0.6 0.0 0.0 Morella faya 0.2 0.0 0.0 Nephrolepis exaltata subsp. hawaiiensis 1.1 0.0 0.0 Pipturus albidus 0.0 0.0 1.0 Psidium cattleianum 15.5 0.0 0.0 0.000↓ 91.056 Psychotria mariniana 0.2 0.0 0.0 Rubus rosifolius 0.2 0.0 0.0 Schinus terebinthifolius 1.1 0.0 0.0 Wikstroemia oahuensis var. oahuensis 0.2 0.0 0.0 >2 m Cheirodendron trigynum 1.1 0.2 0.2 Cibotium chamissoi 0.4 0.0 0.0 Clidemia hirta 4.2 0.0 0.0 Coprosma longifolia 0.2 1.0 0.4 Freycinetia arborea 5.4 0.2 0.0 Grevillea robusta 1.1 0.0 0.0 Kadua affinis 0.6 0.0 0.4 Metrosideros polymorpha 9.6 5.4 5.9 Morella faya 4.4 0.0 0.0 Passiflora edulis 0.6 0.0 0.0 Psidium cattleianum 77.7 0.0 0.0 0.000↓ 688.332 Psychotria mariniana 0.4 0.0 0.0 Scaevola gaudichaudiana 0.4 0.0 0.0 Schinus terebinthifolius 39.1 0.0 0.0 0.000↓ 263.541 1 Zero tolerance weed

Table 3. All taxa intercepted (I) or anecdotally (*) observed during monitoring prior to clearing (2016), during baseline (2017) and 1-year post-construction (2018) in any strata from 0 - 1 m, 1 -2 m, or >2 m above ground level. Note: The anecdotally observed area in 2016 (prior to wall construction and final wall placement determination) included area outside the enclosure wall, and may include taxa not present within the bounds of the enclosure.

Native 2016 2017 2018 Non-native 2016 2017 2018 Acacia koa I Ageratum conyzoides * * Andropogon virginicus * 1 Antidesma platyphyllum I Blechnum appendiculatum I Asplenium caudatum * Buddleja asiatica * Asplenium contiguum I * * Casuarina equisetifolia * Athyrium microphyllum * Clidemia hirta I * I Bidens torta I Conyza bonariensis * Broussaisia arguta * * Crassocephalum crepidoides * Cheirodendron trigynum I I I Cyclosorus parasiticus I * * Cibotium chamissoi I I I Epidendrum x obrienianum * Coprosma foliosa * I Ehrharta stipoides1 I * I Coprosma longifolia I I I Emilia sonchifolia * Cyperus polystachyos I Gamochaeta purpurea * Dianella sandwicensis * * * Grevillea robusta I Dodonaea viscosa * Leucaena leucocephala * Dryopteris fusco-atra * * Melinis minutiflora * Dryopteris glabra * Morella faya I * Freycinetia arborea I I I Oxalis corniculata I 1 Ilex anomala * * Paspalum conjugatum I * I Kadua affinis I I I Passiflora edulis I Labordia kaalae * * Passiflora suberosa I * Lepisorus thunbergianus * * * Phlebodium aureum * Metrosideros polymorpha I I I Phytolacca octandra * I Microlepia strigosa I * I Pityrogramma austroamericana * Myrsine lessertiana * Pluchea carolinensis I Nephrolepis exaltata subsp. hawaiiensis I I I Polystachia concreta * Peperomia membranacea * * Psidium cattleianum I * I Peperomia tetraphylla * * * Rubus rosifolius I I Perrottetia sandwicensis I Schinus terebinthifolius I * * Pipturus albidus I Unknown sp. (Scoparia dulcis?) * Pisonia brunoniana I Youngia japonica * * Psilotum nudum * * Psychotria hathewayi * * Psychotria mariniana I * I Pteridium aquilinum * Scaevola gaudichaudiana I * I Smilax melastomifolia * * * Streblus pendulinus * * * Urera glabra * Vandenboschia davallioides * Waltheria indica * Wikstroemia oahuensis var. oahuensis I * I Total native diversity 29 25 30 Total non-native diversity 14 9 25 1 Zero tolerance weed

Table 4. Species outplanted, sown, and/or transplanted in association with native vegetation restoration during the first year following enclosure construction completion. Total number of outplants, seeds, and transplants are listed by date. Seed numbers were not tracked for every seed sow conducted; these are listed as undetermined, or U.

05 08 17 25 08 11 27 12 - 01 - 01 - 01 - 03 - 04 - 06 ------

2017 2018 2018 2018 2018 2018 2018 Total Outplants Acacia koa 24 24 Alyxia stellata 3 3 Bidens torta 140 6 146 Cheirodendron trigynum 18 17 83 50 168 Coprosma longifolia 134 24 92 40 290 Freycinetia arborea 3 3 Ilex anomala 4 8 12 Kadua affinis 375 29 8 412 Perrottetia sandwicensis 24 2 26 Pipturus albidus 12 12 Pisonia brunoniana 60 1 99 160 Psychotria mariniana 37 1 38 Santalum freycinetianum var. freycinetianum 4 4 Scaevola gaudichaudiana 9 9 Urera glabra 52 12 53 14 131 Wikstroemia oahuensis var. oahuensis 33 33 Total outplants 843 116 381 131 1471 Seed sows Bidens torta U ca. 10,000 >10,000 Pipturus albidus U U U U Scaevola gaudichaudiana 24 24 Total seeds sown U U >10,000 >10,024 Transplants Cibotium chamissoi 65 65 Total transplants 65 65

Table 5. Vegetation percent cover by growth form within stratum prior to clearing (April 2016), post-construction (September 2017), and one year post-construction (September 2018). Native 2016 2017 2018 Non-native 2016 2017 2018 0-1 m 0-1 m Fern 17.7 2.0 5.7 Fern 0.7 0.0 0.0 Herb 0.0 0.0 2.2 Grass 5.4 0.0 4.3 Sedge 0.0 0.0 1.6 Herb 0.0 0.0 0.4 Shrub 0.0 0.4 5.7 Shrub 15.1 0.0 1.6 Tree 0.4 0.4 11.0 Tree 20.7 0.0 0.4 Vine 0.2 0.0 0.0 1-2 m 1-2 m Fern 4.1 0.2 0.6 Shrub 20.1 0.0 0.0 Shrub 0.2 0.6 0.2 Tree 16.6 0.0 0.0 Tree 2.0 0.6 1.4 Vine 0.0 0.0 0.0 >2 m >2 m Fern 5.7 0.2 0.0 Shrub 4.2 0.0 0.0 Shrub 0.6 1.0 0.4 Tree 94.1 0.0 0.0 Tree 11.6 5.6 6.5 Vine 0.6 0.0 0.0

Figure 2. UAV imagery of the Palikea North Snail Enclosure area in the initial stages of non-native vegetation clearing (top left), following completion of the enclosure wall (top right), one year post- completion of the wall (lower left), and 1.5 years post-completion (lower right).

Figure 3. Photopoint 1 images over time with views in each cardinal direction prior to non-native vegetation clearing (June 2016), post-clearing (February 2017), during the early stages of construction (May 2017), after the initial outplantings (January 2018), and one year post-completion of the wall (October 2018).

Figure 4. Photopoint 2 images over time with views in each cardinal direction prior to non-native vegetation clearing (June 2016), post-clearing (February 2017), during the early stages of construction (May 2017), after the initial outplantings (January 2018), and one year post-completion of the wall (October 2018).

Figure 5. Photopoint 3 images over time with views in each cardinal direction prior to non-native vegetation clearing (June 2016), post-clearing (February 2017), during the early stages of construction (May 2017), after the initial outplantings (January 2018), and one year post-completion of the wall (October 2018).

Figure 6. Photopoint 4 images over time with views in each cardinal direction prior to non-native vegetation clearing (June 2016), post-clearing (February 2017), during the early stages of construction (May 2017), after the initial outplantings (January 2018), and one year post-completion of the wall (October 2018).

Figure 7. Photopoint 5 images over time with views in each cardinal direction prior to non-native vegetation clearing (June 2016), post-clearing (February 2017), during the early stages of construction (May 2017), after the initial outplantings (January 2018), and one year post-completion of the wall (October 2018).

DISCUSSION

Native cover goal

The initial reduction in native cover that occurred in association with non-native vegetation clearing at the Palikea North enclosure is not unprecedented, as similar reductions in native vegetation happened during clearing of dense P. cattleianum at the Maile Flats chipper site in Kahanahaiki, resulting in a frequency decline for the native vine Alyxia stellata (OANRP 2016). This was also expected to occur at Palikea North given the presence of F. arboretum interwoven in dense P. cattleianum, and the fragile nature of ferns, which were the dominant native component of the understory. Trimming of F. arborea was necessary either because it was growing on non-native trees, or to curtail vegetation problematic to predatory snail (Euglandina rosea) searches/removal, as these snails may easily remain hidden within the foliage. There was also an unfortunate and unexpected loss of some smaller M. polymorpha canopy trees that became vulnerable upon exposure to strong winds. Loss of much native understory vegetation, particularly Nephrolepis exaltata subsp. hawaiiensis, occurred in the process of felling and removal of large amounts of non-native tree material, as well as the necessary raking and removal of litter to facilitate predatory snail elimination. At Ohikilolo Upper MU, F. aborea was observed to recover from severe ungulate browsing, and it is expected to similarly rebound at Palikea North. Indeed, it was anecdotally observed to have filled in within the first year post-construction. Impact to the site was even more severe than the Maile Flats site as the ground was also raked in association with E. rosea removal. This left the area more vulnerable to drying out, as soil moisture retention was likely compromised in absence of leaf litter and detritus.

The goal of 50% native cover from 0 – 1m AGL for the first year was not met, partly resulting from the delayed onset of outplantings, seed sows and transplants, in association with completion of E. rosea removal protocols. The timeline for cover goals may also be somewhat overly optimistic. It was challenging to plan a realistic timeline, having no prior evidence from which to make predictions. However, considering the very low starting point of native cover (3%) in that stratum following enclosure construction, considerable progress has been made towards the goal as a result of both passive (regrowth and expansion of existing vegetation as well as natural recruitment) and active (outplantings, seed sows, and transplants) restoration of native vegetation.

It was too soon to expect to see much change in native vegetation in the 1 - 2m stratum resulting from restoration efforts. It is anticipated that vegetation present from 0 – 1mAGL observed after the first year will have a marked vertical expansion into the 1 – 2m AGL stratum by the end of the second year. However, given the slower than expected progress towards 50% cover from 0 – 1m AGL, it is possible that the 50% cover goal for 1 – 2m AGL by the second year may be similarly overly optimistic. Having > 25% cover from 1 – 2m AGL by two years may be more reasonable. Likewise, the goals of > 75% overall cover by two years may also be overly ambitious. With additional inputs and expansion of existing vegetation, having > 50% overall cover may be more realistic by two years.

Similarly, it was too soon to expect to have new vegetation from restoration efforts reach > 2m AGL by the end of the first year. Substantial change was not expected for the existing canopy trees that survived the initial vegetation clearing. Photopoint images show that some trees became more foliated. This increased leafiness over time may provide more shade and moisture, and have a cooling effect on the surrounding area beneficial to the outplants, seed sows, and transplants.

The shift in native growth forms from fern- to tree-dominated was not surprising, as the active restoration strategy during the early years is to establish native trees expeditiously. Ferns are not highly desirable understory components, as they may obstruct E. rosea searches, and any restoration of understory ferns (except for tree ferns) will remain passive.

Weed cover goals

Though there was higher weed diversity after the first year post-construction compared with pre- clearing, weed cover goal of < 10% was achieved and maintained. Higher weed diversity was not surprising, as dense P. cattleianum cover likely kept diversity low, while abrupt changes in light level and the sudden creation of open disturbed habitat is prone to influx of weedy pioneer taxa. The change from tree and shrub dominated weeds to primarily grass weeds was not unexpected, as grass is challenging to control and grows quickly in open sunny areas. There were, however, some unexpected colonizers, such as Casuarina equisetifolia. It is anticipated that as native cover continues to fill in, conditions will be less prone to pioneer weed invasion, and native cover will suppress at least some weed growth.

The majority of non-native species present after the first year post-construction were target weeds, and the taxon with the highest cover was a zero tolerance weed (P. conjugatum). The enclosure is weeded regularly, at least once quarterly, to maintain low cover given the continual influx of weeds, with particular emphasis on controlling target and zero tolerance taxa (OANRP 2018a). The enclosure was not weeded for two months prior to the most recent monitoring, and the monitoring results are indicative of weedy ingress that may incur between control efforts.

RECOMMENDATIONS AND MANAGEMENT RESPONSE

The impetus for making goals was to trigger additional restoration actions if goals were not met. To make progress towards restoration goals, more outplantings, seed sows, and transplants will facilitate vegetation expansion within the 0 -1m AGL stratum by the second year, and further expansion in the 1 – 2 m and >2 m AGL strata in the following years. Continued regular weed control efforts may mitigate competition from weeds for space, water, and nutrients.

A 10 m grid of trails within the enclosure was installed using whisker markers to facilitate predator removal, restoration efforts, and snail monitoring. This also has the added benefit of preventing trampling while restoration efforts are underway. Many naturally recruiting seedlings, such as M. polymorpha, were anecdotally observed, especially in locations that tend not to be trampled, such as along the inside edge of the enclosure wall. It is important that staff continue to use the trail system to prevent trampling of seedlings and roots, and allow for stabilization of the topsoil and re-formation of the humus layer that was depleted in association with E. rosea removal protocols.

While restoration actions are not new for the program, this restoration site is the first to have specific aggressive goals within a narrow timeframe, given the pressing need for suitable protected habitat for A. mustelina. As such it is an excellent learning opportunity for planning, execution, and practical timelines for fully restored habitat and habitat readiness for use by endangered animals. Qualifications for the level of restoration suitable for the release of A. mustelina have not been studied. Though dense multi- layered host vegetation is presumed ideal, partially restored vegetation may be sufficient for releasing snails. Stable or increasing A. mustelina populations may also be used a measure of vegetation rehabilitation success.

Overcrowding and an unexplained die-off episode of ESU-E snails in the laboratory and rapid declines in wild populations prompted an accelerated release of lab snails (OANRP 2018b). While most of the vegetation after one year consisted of scattered small individuals, there were clusters of vegetation deemed potentially suitable for A. mustelina, having dense and nearly continuous cover, dominated by F. arborea, a known snail host plant. Laboratory snails (26) were released into the most dense vegetation cluster in December 2018 (see Chapter 5). Additional snails were translocated from wild populations (21) in April, 2019. The population remained stable, as very few deaths occurred, and there was at least one birth. The release area was bounded by a low plywood wall with an internal electric barrier to prevent snails from traversing out into areas of sparse vegetation where they may encounter environmental stress, and to facilitate monitoring of survival and mortality within a confined area. Restricting snails to a smaller area also maximizes the opportunity for snails to encounter one another and to potentially reproduce. The electronic barrier further ensured that snails do not escape over the wall of the larger enclosure. Shade cloths and an irrigation system were installed to enhance shade and moisture levels. From these initial snail releases, it appears that early stages of vegetation restoration may be sufficient for snail releases in confined areas with an appropriate level of vegetation cover and host plant presence, supplemental shade, and moisture.

While restoration efforts are primarily associated with management of A. mustelina, they are also meant to benefit D. substenoptera, D. montgomeryi, and C. ibidis. Though use of the enclosure by these taxa will not likely occur for some years, it will be meaningful for the future management of those taxa to understand at what stage of vegetation rehabilitation that utilization is initiated. This will be particularly useful for planning future Drosophila restoration efforts. Host for D. substenoptera (Cheirodendron trigynum) and D. montgomeryi (Urera glabra) were included the restoration outplantings at the enclosure. Plans for baseline monitoring should be initiated, given the inconspicuous nature of this taxon and the expertise required for identification. This may be incorporated into the existing Drosophila monitoring that occurs in close proximity to the enclosure. While C. ibidis management efforts do not include vegetation restoration, such efforts at times occur in areas occupied by that taxon, and understanding how quickly these birds return to the area for nesting, foraging, etc., will better inform our understanding of restoration impacts. Should birds eventually nest within the predator-proof enclosure, they will have added protection from rats as compared with the MU trapping grid that occurs outside the enclosure, which strongly reduces rat presence, but does not completely eliminate it. The construction of the enclosure had a considerable impact on the territory of a C. ibidis pair, however they persisted on its periphery exploiting remaining habitat, and successfully produced one fledgling in the 2018 breeding season (OANRP 2018c). While monitoring of C. ibidis at Palikea is not currently planned for the future, this taxon is very conspicuous and easily recognizable by minimally trained staff, and may be anecdotally observed by staff working in the area.

Modifications to the restoration goals should be considered, including more realistic native cover goals over time, any necessary changes to the target weed list, and restoration focused on specific areas within the enclosure for use in future snail releases.

REFERENCES

Oahu Army Natural Resources Program. 2016. Appendix 3-7 Vegetation monitoring of Achatinella mustelina ESU-E enclosure, 2016 pre-clearing results in 2016 Status Report for the Makua and Oahu Implementation Plans. http://manoa.hawaii.edu/hpicesu/DPW/2016_YER/A3-07.pdf.

Oahu Army Natural Resources Program. 2016. Appendix 3-8 Results of Kahanahaiki chipper site vegetation monitoring five years after initial clearing in 2016 Status Report for the Makua and Oahu Implementation Plans. http://manoa.hawaii.edu/hpicesu/DPW/2016_YER/A3-08.pdf.

Oahu Army Natural Resources Program. 2017. Appendix 5-5 Palikea North Restoration Plan in Status Report for the Makua and Oahu Implementation Plans. https://pcsuhawaii.org/projects/oanrp/reports/2017/A5-5.pdf.

Oahu Army Natural Resources Program. 2018a. Oahu Army Natural Resources Program Database. Accessed November 20, 2018.

Oahu Army Natural Resources Program. 2018b. Appendix 5-1 Management Plan for Achatinella mustelina ESU-E Initial Release of Excess Laboratory Snails at the Ekahanui Temporary Enclosure and the Palikea North Enclosure in 2018 Status Report for the Makua and Oahu Implementation Plans. https://pcsuhawaii.org/projects/oanrp/reports/2018/A5-1.pdf.

Oahu Army Natural Resources Program. 2018c. Chapter 5: Rare Vertebrate Management in 2018 Status Report for the Makua and Oahu Implementation Plans. https://pcsuhawaii.org/projects/oanrp/reports/2018/06.pdf.