The northern striped gecko (Toropuku inexpectatus) is an arboreal gecko endemic to Aotearoa | New Zealand, formally describedin 2020 andprimarily knownfromone monitoredpopulationonthe Coromandel Peninsula. Little is known about the northern striped geckos' use of habitat, which impedes monitoring, and thus limits effective conservation management. We evaluated the geckos' habitat use via Very High Frequency radio telemetry of 32 individual geckos for three months in the summer of 2024 and compared data to the available habitat. We also investigated potential observer biases during systematic searching (visual searching at night), currently the preferred monitoring method. The geckos used a wide variety of forest habitat, preferring edge habitats and shrub-like vegetation. While inactive during the day, and in close proximity to the forest edge, geckos sheltered under the leaf litter on the forest floor. While away from the forest edge, geckos moved into the forest canopy and did not utilise the forest floor. We uncovered observer biases in systematic search techniques, favouring observation of geckos near forest edges and often on or nearthe fernkiokio (Parablechnum novae-zelandiae). Systematic searches only captured a portion of gecko habitat use and largely failed to locate geckos in the complex vegetation that was preferred by them. The limitation of systematic searching indicates the potential for undiscovered populations in the remote forests of the Coromandel Peninsula. Overall, our results contribute to a better understanding of the ecology of the northern striped gecko and where to find them, potentially enabling improvements in surveys and monitoring and, ultimately, conservation of one of the most elusive geckos of Aotearoa.
Seed dispersal distances drive the rate of spread for invasive plants and are strongly influenced by dispersal mechanisms. We investigated weed seed dispersal mechanisms and dispersal potential in New Zealand. We collated data on dispersal mechanisms for 318 species that produce seed in New Zealand. We modelled maximum seed dispersal distances of 286 species for which data were available using dispersal mechanism, growth form, seed release height, and seed mass. Internal dispersal by vertebrates was the most frequent mechanism, both among weed species (56%, 60% of which had fleshy fruits) and all weed species & times; dispersal mechanism interactions (34%). Wind was the next most frequent mechanism (39% of species, 24% of mechanisms), followed by water (29%, 18%), attachment (16%, 10%), ant (8%, 5%), and ballistic dispersal (5%, 3%). Most weed species (56%) have only one known dispersal mechanism, 37% are dispersed by multiple mechanisms, and 7% have no known mechanism. Non-standard mechanisms (those not predicted by plant morphology) were used by 44% of weed species and comprised 51% of all mechanisms. Modelled maximum seed dispersal distances show that most weed species have the potential for long-distance dispersal: 70% of species were predicted to be dispersed > 100 m, 17% > 1 km, and 0.3% > 10 km. Dispersal distances were greatest for vertebrate-dispersed seeds (mean = 973 m), then seeds dispersed by wind (146 m), ants (75 m), and ballistic mechanisms (14 m). Many weed species are dispersed by non-standard mechanisms (especially vertebrates) that have the potential to move seeds much further than standard mechanisms and thus accelerate range expansion. We predict that non-standard mechanisms are even more frequent than reported to date, as they have been rarely studied in New Zealand. Examining the combined effect of multiple mechanisms is critical to understanding the historic and future spread of weeds.
This pilot study aims to quantify the impacts of roosting birds on the pest exclusion fenced area of Rotopiko wetland ecosystem in the Waikato region of New Zealand. Water samples were collected from five monitoring stations, encompassing four stations within the fenced area (three along a drain running through the fenced area into the lake, and one from the fenced lake itself), and one from a nearby unfenced lake, not affected by roosting. Soil samples were collected from three monitoring stations: two inside the fenced area, with varied roosting density, and one outside. Sampling was conducted once in each season, during winter (August 2021), spring (November 2021) and summer (February 2022). For each sampling time, a triplicate sample was taken at each station. Within the fenced area, increased nutrient loading in water and soil, and microbial growth in water, were observed in areas with the highest roosting pressure. Water tested where the drain enters the lake showed an up to 44-fold increase of total nitrogen, a 750-fold increase in faecal coliforms, and a 915-fold increase in Escherichia coli compared to water collected at the point where the drain enters the fenced area. Comparison between the lakes showed that the total nitrogen concentration in the fenced lake was 16 times higher than in the non-fenced lake, while the number of faecal coliforms and E. coli were over 30 times higher at the fenced lake. This study, although limited in sample size and statistical power, sheds light on the impacts of roosting birds on the Rotopiko ecosystem.
Ephemeral wetlands are naturally uncommon ecosystems that have become threatened in Aotearoa | New Zealand and include many glacially derived kettle holes (kettles) of the South Island. These kettles can be easily overlooked as wetlands, particularly when in a dry phase, thereby increasing their vulnerability. To aid conservation planning and management of this threatened ecosystem, we identified and mapped all detectable individual kettles in the South Island (n = 1802) and used geographic and survey data to evaluate patterns in their distribution, floristic diversity, land tenure, changes in land tenure over time, legal status, and the occurrence of Threatened plants and lichens. Ninety-five percent of kettles were found in Waitaha | Canterbury, mostly in Te Manahuna | Mackenzie Basin, and most were small (< 0.6 ha) and close to other kettles. Altogether, kettles occupy just 0.004% of Aotearoa's land area. Sixty-eight percent of kettles were found on either freehold land or pastoral lease land, with 10% of these protected by conservation or Queen Elizabeth II covenants; 29% of kettles were on public conservation land. Of the 29%, over half (60%) were on Stewardship Land. Therefore, kettles generally have limited protections that are dependent on their land managers. We completed plant and lichen surveys for 269 of the mapped kettles. We recorded 443 plant and lichen taxa; 128 were exotic species, 314 were indigenous, and one had uncertain biostatus. Occurrence of plants in Threatened orAt Risk categories did not differ by land tenure or legal status of the kettles, so preservation of these wetlands needs to occur across all land tenures to ensure representative protection of kettles and their botanical values. Pairing desktop mapping with a flora survey shows that kettles are remarkable plant biodiversity hotspots, especially given they occupy such a small land area in Aotearoa.
Long-term monitoring is pivotal to understanding population trends in threatened species and quantifying the impact of conservation interventions. Effective monitoring methods are generally underdeveloped or absent for taxa inhabiting the alpine zone. In Aotearoa | New Zealand (NZ), the alpine zone covers c. 11% of the land mass and supports a diverse fauna of at least 34 native lizard species. We developed and tested the effectiveness of footprint tracking tunnels and rock lifting for monitoring two cryptic alpine gecko taxa, cascade geckos (Mokopirirakau "Cascades") and orange-spotted geckos (Mokopirirakau "Roy's Peak"), in NZ's SouthernAlps during the austral summer. Field sampling occurred within defined strip transects and included two variants of rock lifting (one highly constrained involving lifting just 30 pre-selected marked rocks per transect; the other involving lifting all rocks judged suitable) and baited footprint tracking tunnels. All sampling methods resulted in low capture rates, plus highly variable detection rates for tracking tunnels. Rock turning counts of cascade geckos were consistently higher than those of orange-spotted geckos, and marked rocks returned counts 2.5-6.3 times higher than other rocks lifted within the transects. We used simulations to assess the ability of our trialled methods to detect a 20% change in gecko counts (hand searching via rock lifting) or detection rates (tracking tunnels) over a 10-year period, across 10 years for rock lifting and 15 years for tracking tunnels. None of our methods were capable of reliably detecting a rate of change equivalent to a 20% increase or decline, even if sampling effort was increased. Our results highlight the difficulties in developing effective monitoring methods for species that are sparse or difficult to detect. We conclude that alpine geckos will require longer monitoring timeframes than their lowland counterparts due to their slow life histories and cryptic nature.
Fenced ecosanctuaries provide critical refuges for native species vulnerable to predation and have driven major biodiversity gains. Understanding whether they can also support sustainable populations beyond their boundaries is an important next step. While previous studies have documented native species dispersing beyond ecosanctuaries, few have utilised methods that explicitly quantify dispersal and the establishment of resident birds. We quantified dispersal of a kakaruai | South Island robin (Petroica australis) population, translocated to Orokonui Ecosanctuary in 2010, by evaluating site occupancy, persistence in previously colonised areas, and establishment of new territories within a site. Surveys were conducted during the kakaruai breeding season in 2018, 2019, and 2022. In 2018, five fixed transects were each surveyed twice. This was expanded to 11 transects in 2019 and 2022. Between 2018 and 2022, the number of sites occupied by kakaruai steadily increased, with four new sites colonised in the final survey year. Three sites showed persistent occupancy over multiple years, and both the number of individual birds and the number of territories within established sites increased over time, possibly as individuals settled near conspecifics. Data from 2019 and 2022 were then used to generate a model-averaged analysis incorporating site, year, connectivity, and habitat type (native or mixed) as key predictors of kakaruai presence. The strongest predictor of kakaruai presence was proximity to the ecosanctuary, with presence declining significantly beyond 1500 metres. Our research provides a valuable model for assessing initial dispersal success at the population level by using species-specific monitoring approaches rather than general bird counts. Demonstrating dispersal beyond the fence is a critical first step toward extending the ecological benefits of sanctuaries into the wider landscape, informing strategies that enhance connectivity and long-term population viability.
Genomics can provide conservation-relevant insights into population size, genetic diversity, and connectivity withrelatively little sampling effort. However, forfrogs and other smallvertebrates, invasive genetic sampling (e.g. toe-clipping) can impact animals' long-term survival and welfare. Minimally-invasive genetic sampling methods may be required to facilitate robust conservation genomic studies of small vertebrates. In Aotearoa | New Zealand, previous genetic studies of the three native Leiopelma spp. frogs have relied almost entirely on toe-clips or whole specimens, creating an ethical barrier to genomic studies. Here, we trialled duplicate buccal swabbing as a minimally-invasive genetic sampling method on the small-bodied pepeketua, Hamilton's frog (Leiopelma hamiltoni) and Hochstetter's frog (Leiopelma hochstetteri). DNAyield was primarily influenced by our duplicate swabbing method rather than by frog body size, condition, or species. The second swab had a significantly higher DNA yield (721 +/- 104 ng), more than twice that of the first swab (317 +/- 49 ng). We compared the utility of buccal swabs and toe clips for genotyping-by-sequencing (GBS) in Hamilton's frogs and assessed the suitability of buccal swabs for long-range PCR in Hochstetter's frogs. Buccal swabs were as effective as toe-clips for GBS, with the proportion of missing single-nucleotide polymorphisms not significantly different between sampling methods. However, only 7/20 buccal swabs amplified successfully with long-range PCR. Our results indicate that genetic sampling of Leiopelma spp. could move towards buccal swabbing as a standard method. Toe-clipping and euthanasia of individuals should be limited to situations where high molecular weight DNA is crucial for analysis (e.g. genome assembly) or whole voucher specimens are needed for taxonomy.
Radio telemetry is a valuable technique to gain insights into the ecology of a species by enabling repeated observations of an individual, regardless of its activity or detectability. The miniaturisation of transmitters has allowed a wide variety of taxa to be studied using radio telemetry, including amphibians. In Aotearoa | New Zealand, all three extant endemic Leiopelmatid frogs are threatened and require ecologically informed conservation management. Radio telemetry has previously been used to explore the movement ecology of the two terrestrial Leiopelma species. However, the movement ecology of the semi-aquatic Hochstetter's frog (Leiopelma hochstetteri) remains poorly understood. To determine whether a simple waistband harness design was suitable for L. hochstetteri, we conducted a field study at Mahakirau Forest Sanctuary, Coromandel Peninsula, New Zealand. While assessing harness suitability, we made repeat observations of individuals with fully functional transmitters (n = 9) for up to 16 days, providing insight into the species' movement ecology. The longest distance moved between observations was 16.9 m. Minor abrasion occurred in the hip joints of ten frogs. Following refinements to the waistband length to reduce abrasion, radio telemetry will potentially be a beneficial tool to gain information needed for conservation planning.
Control of introduced mammalian predators is important for the protection of pekapeka | lesser shorttailed bat (Mystacina tuberculata) populations, but toxins used to control these predators may also pose a risk to bats when they forage on invertebrates that have consumed toxic bait or if bats directly consume bait. We studied the survival of a population of lesser short-tailed bats at Pikiariki, Pureora Forest Park, between 2012 and 2024, over which time conservation managers varied the methods and timing of toxin application for rat control to reduce secondary poisoning risk to bats. Reducing the duration of rat control operations and timing them to occur prior to the bat breeding season successfully reduced exposure of the bats to toxins. Although rat control operations effectively reduced relative rat abundance in the short term, there was no relationship between duration of rat control operations and annual mean rat tracking rates (the mean tracking rate for each year). Neither annual mean rat tracking rates nor the amount of time toxin was present in bait stations explained very much variation in bat survival from year to year. With an overall mean rat tracking rate across all years of 27 %, mean annual survival of the adult female lesser short-tailed bat population across all years was high (0.86), resulting in growth of the population. More frequent rat tracking data could be collected to test whether relative rat abundance at specific times of the year can be used to set a more precise operational target for rat control to further benefit bats. We recommend continuing to monitor the bat population as well as the relative abundance of rats and other introduced mammalian predators to improve the understanding of factors influencing bat survival. While an overall mean rat tracking rate of 27 % may allow this lesser short-tailed bat population to grow, other, more vulnerable, forest species requiring lower residual relative rat abundance targets should be considered in future management plans.
Actively attracting rats to trapping and baiting areas would potentially reduce the costs and labour required for controlling rats over large areas of forest. In this pilot study we explored the potential of using an ultrasonic lure to attract rats in two forest study areas: Tuateawa and Te Hoe. Data was collected from the Tuateawa study area over 21 discrete 24 h periods using tracking tunnels and cameras. Results indicated that rats were present more often at sites where a 50 kHz Norway rat ultrasonic vocalisation (USV) was played compared to sites where it was not played, although significance could not be attributed based on the low numbers detected. Interestingly, mice were detected at the Tuateawa study area more often at sites where the USV was played compared to sites where it was not played, with differences approaching a significant level of confidence (p = 0.059). At the Te Hoe study area, data was collected over a three-month period using A24 self-resetting traps and cameras. Results indicated that the weekly numbers of rats killed was higher (p = 0.025) at sites where USVs were played than at sites where USVs were not played. These findings indicate that rat USVs can attract rats and mice to trapping areas in New Zealand forest settings and support the view that including USVs in trapping methods could potentially enhance the efficiency of trapping and hence reduce the costs associated with current methods.
In Aotearoa | New Zealand, conversion of low-yield high hill-country pasture farms to m & amacr;nuka (Leptospermum scoparium) forests to support the harvesting of high quality monofloral m & amacr;nuka honey can provide an economic benefit for rural landowners. However, the effects of such conversions on the local ecosystem and biodiversity are largely unknown. In a cross-sectional (space-for-time substitution) case study, we undertook biodiversity surveys at a single central North Island property that contained four sites representing different time points in the transition from pasture to m & amacr;nuka forest: pasture (34 ha), three-year-old (36 ha) and five-year-old (23 ha) planted m & amacr;nuka, and > 30-year-old naturally regenerated m & amacr;nuka forest (57 ha). Terrestrial invertebrate surveys using window traps and bat and bird audio surveys were performed at each site. Freshwater health assessments were undertaken at three stream sites that were independent of the terrestrial sites. We found that invertebrate and bird abundance and diversity were lower in the three-year-old planted m & amacr;nuka than in the pasture site. However, both invertebrate and bird abundance and diversity were higher in the five-year-old planted m & amacr;nuka than the pasture, and were comparable to the > 30-year-old naturally regenerated m & amacr;nuka forest in terms of community composition and diversity. Calls from the critically endangered long tailed bat (Chalinolobus tuberculatus) were detected more frequently at the five-year-old and > 30-year-old naturally-regenerated sites than at the other sites. Compared to the pasture catchment, macroinvertebrate scores and other stream health measures were higher in streams whose surrounding land use was native forest, naturally regenerated m & amacr;nuka, or planted m & amacr;nuka. Although site age was not replicated, our results suggest that there could be considerable positive changes in biodiversity that occur between three to five years after planting m & amacr;nuka, with the five-year-old m & amacr;nuka in our study supporting a more diverse and abundant community that was similar to that of the > 30-year-old naturally regenerated m & amacr;nuka forest. Overall, our study provides supporting evidence that planting of m & amacr;nuka forests in former pasture likely leads to a positive change in the ecosystem, however replication at different locations is required to confirm these case-study observations. Results from such surveys may aid in developing management systems to further enhance ecological outcomes of planted m & amacr;nuka forests.
Globally, rapid human-caused climate change is posing one of the greatest emerging pressures on species and ecosystems. Uncertainty remains about how species will react to changing climate, particularly how vulnerable or adaptable and resilient they will be to changes and whether they can move to remaining favourable habitats. Here we report on a rapid trait-based climate change vulnerability assessment for all bat taxa resident in Aotearoa New Zealand. We use 16 traits across three dimensions (sensitivity, low adaptive capacity, exposure) that could make bats highly vulnerable to climate change, review the potential consequences on their long-term population viability and recommend research required to understand vulnerability and potential adaptive management requirements. We assessed all five known endemic bat taxa under two greenhouse gas (2040, 2090). One taxon (the greater short-tailed bat Mystacina robusta) was categorised as Highly Vulnerable by 2040 under RCP4.5, and the other four taxa demonstrated Latent Risk profiles, indicating they should be monitored closely to determine whether their potential responses to projected climate change happen earlier than 2090. All taxa were categorised as Highly Vulnerable by 2090 under the high emission RCP8.5 scenario. Although no overall conclusion can be reached about whether climate change will have positive, negative or mixed consequences for bats, the likely negative consequences include increased risk of predation, reduced availability and quality of roosting and foraging habitats, and increased thermal stress. Future research should focus on understanding the costs and benefits of changing climate on productivity and survival of bats and determining if management responses are desirable and feasible.
Introduced mammalian predators have had dramatic impacts on the ecosystems of oceanic islands. While conservation strategies have been developed to suppress or eradicate them in a wide range of unpopulated habitats, their management in human-dominated landscapes is less advanced. Here, we assess the efficacy of urban predator control using two years (four sessions) of mammal monitoring data in before-after controlimpact (BACI) experiments in & Omacr;tepoti Dunedin and Kirikiriroa Hamilton. Results of the two BACI experiments provided no evidence for an effect of intensified predator control on rat, mouse, or possum abundance. Short experimental timeframes and low rat numbers in the before condition (especially for rats at non-treatment sites) may have made it difficult to identify effects. Further research is required to understand relationships between urban predator control and pest mammal abundance, and how control may be optimised to maximise cost-effectiveness.
Wetlands are a critical, though vulnerable, global carbon store, but the carbon stored in vegetation has not been quantified at a national scale for New Zealand wetlands. We undertook a literature review to assess vegetation carbon density in wetlands and used meta-analysis to estimate means and uncertainty for both vegetation structural classes and wetland type. We then used a combination of derived vegetation carbon densities alongside spatial extrapolation to estimate the amount of carbon stored in wetland vegetation in New Zealand. Our area of interest was the "Wetland-Vegetated non forest" class mapped by the Land Use Map (LUM), which is the authoritative layer for New Zealand carbon accounting. Within our area of interest, we used weighted aggregated land cover classes signifying vegetation to calculate a grand mean density (C ha-1) for New Zealand wetlands mapped by the LUM in 2008, 2012, and 2016, and total vegetation carbon stocks for each of those years. We found that among wetland vegetation structural classes, total above-ground carbon varied from a mean of 4.8 Mg C ha-1 for sedgelands to 23.8 Mg C ha-1 for tall mangroves. For wetland types, total above-ground carbon varied from a group mean of 3.6 Mg C ha-1 for pakihi to 19.5 Mg C ha-1 for mangroves. Estimates of below-ground biomass were uncommon (n = ten estimates from four studies) but the limited data available suggest that below-ground carbon density was less than total above-ground carbon density for three herbaceous structural classes (below-ground divided by above-ground ranged from 0.70 for reedland to 0.86 for rushland), while for mangroves below-ground density varied: below-ground divided by total above-ground was 0.99 for tall mangroves and 2.09 for dwarf mangroves. We estimated a total carbon stock of 5 868 710 (1 467 178-10 270 243) Mg C for our area of interest, based on a density of 27.080 (6.68-47.48) Mg C ha-1, and an area of 216 717 ha in 2016, the most recent estimate available. Using selected literature, we confirmed our estimates were broadly consistent with international values. There is a clear information gap regarding carbon densities in forested wetlands in New Zealand, and little ability to assess vegetation by wetland type due to the lack of spatial data around structural classes in New Zealand. We set out some recommendations to address this and other data gaps.
& Amacr;wheto, also known as the vegetable caterpillar, is a taonga (treasure) of Aotearoa | New Zealand, produced when endemic Ophiocordyceps fungi infect ghost moth (Hepialidae) larvae and produce fruiting bodies extending from the infected larvae to above the forest floor. Despite its cultural importance, & amacr;wheto ecology is not well documented or understood. Partnering with kaitiaki (guardians), we paired m & amacr;tauranga (M & amacr;ori knowledge)-informed soil DNA assays with systematic field surveys to resolve & amacr;wheto ecology in a montane native forest. ITS2/18S gene metabarcoding detected Ophiocordyceps sequences in 92% of random soil samples, however & amacr;wheto were found in higher density clustered at a subset of three of these sites (and at 25% of randomly sampled sites). & Amacr;wheto occurred exclusively in tall red-beech (Fuscospora fusca) forest with partial canopy (c. 75%), deep litter, high bryophyte cover, and minimal bare ground (< 10%). Indicator analysis highlighted p & omacr;k & amacr;k & amacr; (Elaeocarpus hookerianus) and Hall's t & omacr;tara (Podocarpus laetus) as positive associates; occurrence was also more likely in west facing higher altitude sites. Sanger sequencing of stromata and moth larvae cadavers was used to identify Ophiocordyceps robertsii and its sole host, Dumbletonius unimaculatus, with low intraspecific genetic variation. Stroma persisted for > 10 months; motion-detection camera monitoring showed frequent deervisits caused negligible disturbance. These findings demonstrate that, although O. robertsii is widespread in soil, successful & amacr;wheto formation depends on specific micro-habitats and host distribution. Integrating molecular detection with m & amacr;tauranga M & amacr;ori offers a reliable, non-invasive tool for locating & amacr;wheto and provides a baseline for kaitiakitanga-led conservation.
The Western honey bee (Apis mellifera L.) is an important species for crop pollination and honey production. New Zealand has a major wealth-creation opportunity through the production and export of honeys sourced from native flowers. However, honey bees are not native to New Zealand, and the impacts of commercial honey production and honey bee apiaries on native ecosystems are largely unknown. In this study, we used DNA metabarcoding of bulk flowers and bulk Malaise trapped invertebrate samples alongside conventional specimen identifications to compare invertebrate communities in m & amacr;nuka (Leptospermum spp.) shrublands with and without honey bee apiaries. Diptera, Hemiptera, Hymenoptera, and Lepidoptera each had 44-47% lower amplicon sequence variant (ASV) richness on flowers from apiary sites compared to non-apiary sites, suggesting honey bees disrupt flower visits by these other invertebrates. Conversely, Thysanoptera (thrip) ASV richness was 64% higher on flowers from apiary sites compared to non-apiary sites, suggesting transport of thrips between flowers by honey bees. Apiary sites also had lower abundances of conventionally identified specimens of Lepidoptera and a pool of unidentified small invertebrates compared to non-apiary sites. In contrast, DNA metabarcoding of Malaise-trapped bulk invertebrates suggested that honey bee effects on these communities may be site-dependent, with differing richness of Hymenoptera, Lepidoptera, and Orthoptera ASVs between site pairs but not between sites with and without apiaries. Multivariate analyses of community composition also suggested that honey bees had consistent effects on flower communities, but not on bulk invertebrate communities. We recommend future studies with greater replication of site pairs along key environmental gradients to complement and increase interpretability of the results presented here.
Rapid alterations in plant and animal phenology driven by global climate change and rising temperatures can have far-reaching consequences for cultural and ecological systems. We documented changes, and mechanisms behind the changes, in fruit biomass and phenology in mixed podocarp-hardwood forests in the central North Island of New Zealand since the 1950s using the traditional knowledge (herein referred to as m & amacr;tauranga) of Indigenous M & amacr;ori (Tahoe Tuawhenua and Ng & amacr;ti Whare) forest practitioners. We also explored the likely cascading consequences of alterations in fruiting systems for forest frugivores, below-ground subsystems, local livelihoods, and language. Practitioners reported that the frequency of fruiting, crop size, and individual fruit size and quality declined significantly over the last 75 years, with the timing of fruit ripening often delayed by three months. A general warming trend, combined with less frequent and later frosts, and increased weather variability such as strong wind events, were identified as mechanisms that have impacted fruiting. Alterations in fruit biomass and phenology have also adversely impacted the body condition, breeding rates, and abundance of kerera | New Zealand pigeon (Hemiphaga novaeseelandiae) and feral pigs (Sus scrofa), two important food species for the community in the forest. Practitioner estimates of historic fruit inputs suggest that there were substantial nitrogen (N) returns to the below-ground subsystem from fruit that were 1-2 orders of magnitude higher than those reported from forest fruit fall elsewhere, and which are now severely reduced. The reduction in nutrient inputs and simple carbohydrates (e.g. sugars) in the fruit pulp to the soil is likely to have greatly impaired forest soil microbial activity, earthworm densities, and soil decomposition processes and rates, and ultimately the supply of soil nutrients for trees. Retention of m & amacr;tauranga and nomenclature related to fruit are important for comprehension of past ecological states, and for future customary management goal setting and decision-making.
