Successional processes induced by fires on the northern offshore islands of New Zealand

A modern pollen-vegetation data set of 46 samples is presented from subantarctic Campbell Island, 600 km south of the New Zealand mainland. The sampled vegetation includes all major community types: maritime turf and grassland, sedge flushes, dwarf forest, scrub, cushion bog, tussock grassland, and high altitude graminoid turfs and tundra. Macrophyllous forbs—characteristic plants of subantarctic islands—are common throughout. Most taxa have highly restricted pollen dispersal, largely due to the short stature of the vegetation and the high proportion of insect-pollinated species. Percentages of pollen or spores of the dominant taxa have a significant positive correlation with the percent vegetation cover of the corresponding species, the exceptions being the widespread ferns Polystichum vestitum and Blechnum spp., and the ubiquitous macrophyllous forb, Bulbinella rossii .The relationship between the vegetation cover of a given taxon and its pollen representation was usually not strong enough to give confidence in a quantitative reconstruction based on pollen frequency alone. However, the broad vegetation groupings have characteristic pollen and spore spectra clearly related to the abundance of their dominant plant species. Detrended correspondence analysis of the pollen spectra grouped most sites according to their source vegetation type and generated a pattern similar to that of vegetation data analysed in a similar fashion in previous studies of the island. This study, together with recent work on Auckland Island pollen and spore representation, has resulted in a combined modem palynological data base of more than 100 sites for the New Zealand subantarctic islands.     

Comparative study of the effect of distance on the daily and hourly pollen counts in a city in the south-western Iberian Peninsula

The airborne pollen concentration in the city of Badajoz was measured in two locations 2.9 km apart. The measurements were taken from January to June between 2009 and 2012 using Hirst-type volumetric aerobiological samplers. One sampler was placed at the Faculty of Science (FS) and the other at the Agricultural Engineering School (AES) of the University of Extremadura, Spain, on terraces located 16 and 6 m above ground, respectively. The two sets of hourly and daily pollen concentrations were compared regarding the meteorological parameters and the distribution of local pollen sources. A total of 46 pollen types were counted, with a mean total concentration of 43 pollen grains/m³ in the winter and 336 pollen grains/m³ in the spring. In the winter, pollen grains from (in decreasing order) Cupressaceae, Fraxinus–Phillyrea, Urticaceae spp., Alnus glutinosa and Urtica membranacea types represented 77 % of the total. In the spring, 89 % of the total was represented by pollen grains from (in decreasing order) Quercus, Poaceae, Olea, Pinaceae and Plantago. The FS/AES ratio was 0.57 in the winter and 1.31 in the spring. While a Wilcoxon test applied to the daily total concentration data showed statistically significant differences between the two sites, a correlation study based on the Spearman coefficient showed statistically significant correlations in both the winter and spring. The results were similar when comparing the separate pollen types, except for Urticaceae spp., which showed no statistically significant correlation. The meteorological data studied showed a statistically significant correlation with the daily concentrations. A comparison of the hourly concentration data showed no correspondence with the time of maximum concentration. The local distribution of pollen sources explained some of the differences found between the two sites. Overall, the results indicate that a single aerobiological sampler may be sufficiently representative to register the daily pollen grain data of an urban area of approximately 3 km or greater in diameter, although it underestimates the influence of heterogeneity in the distribution of the local flora.     

Spore rain in relation to regional sources and beyond

While patterns of spore dispersal from single sources at short distances are fairly well known, information about ‘spore rain’ from numerous sources and at larger spatial scales is generally lacking. In this study, I sampled spore rain using a novel method consisting of 0.25–0.5 m² cotton cloth traps at nine sites in the boreo‐nemoral vegetation zone in eastern Sweden during two seasons, using Sphagnum spores as a model. Traps were located in various landscapes (mainland, islands). Additional trapping was done in an arctic area (Svalbard) without spore production. Spore densities were tested against distance from the nearest source and area of sources (open peatlands) within different radii around each site (5, 10, 20, 50, 100, 200, 300, 400 km). The cloth method appeared reliable when accounting for precipitation losses, retaining approximately 20–60% of the spores under the recorded amounts of precipitation. Estimated spore densities ranged from 6 million m^?2 and season within a large area source, via regional deposition of 50 000–240 000 spores m^?2, down to 1000 m^?2 at Svalbard. Spore rain for all sites was strongly related to distance from the nearest source, but when excluding samples taken within a source peatland, the amount of sources within 200 km was most important. Spores were larger at isolated island sites, indicating that a higher proportion originated from distant, humid areas. Immense amounts of Sphagnum spores are dispersed across regional distances annually in boreal areas, explaining the success of the genus to colonise nutrient poor wetlands. The detectable deposition at Svalbard indicates that about 1% of the regional spore rain has a trans‐ or intercontinental origin. The regional spore rain, originating from numerous sources in the landscape, is probably valid for most organisms with small diaspores and provides a useful insight in ecology, habitat restoration and conservation planning.     

Surface pollen-vegetation relationships on the Atlantic seaboard: South Uist, Scotland

Studies of modern pollen rain from remote islands have raised a number of interesting issues concerning the spatial precision of present‐day pollen spectra in relation to their parent plant community types. This paper examines the relationships and degree of correlation between a sequence of contemporary vegetation types, environments and their associated surface pollen spectra from a transect across the island of South Uist in the Outer Hebrides of Scotland. Paired data on both contemporary vegetation and associated surface pollen assemblages have been collected and analysed using methods of numerical classification and ordination. In general terms, the modern pollen rain on South Uist reflects the major changes in vegetation pattern and the major community types fairly closely. The major boundary between the alkaline machair sand dune communities and the various acidic upland vegetation types is particularly clear. However, both variability in the vegetation and the effects of the strong prevailing westerly and south‐westerly winds tend to blur the boundaries of the various communities within each of these larger categories. On average 86.6% of the palynomorphs come from in‐community quadrat sources, while only 1.3% are from off‐island sources. The limited present‐day tree distribution on the transect is discussed in the context of the more widespread distribution of arboreal pollen. Overall, there is a strong numerical correspondence between vegetation, pollen and environmental variables. The various problems inherent in examining surface pollen spectra are reviewed.     

Taxon-related pollen source areas for lake basins in the southern Alps: an empirical approach

The pollen/vegetation relationship in broadleaved forests dominated by Castanea sativa was analysed using an empirical approach. The pollen content of surface sediments of three lake basins of different sizes (6.3, 22.2, and 101.2 ha) in Ticino (southern Switzerland) was used for a comparison with the surrounding vegetation. We surveyed the vegetation around the two small lakes, Lago di Origlio and Lago di Muzzano, and estimated the relative crown coverage of tree species. The regional vegetation outside the lake catchment (ca. >1 km) was determined with the data from the first Swiss National Forest Inventory. For the third large lake, basin of Ponte Tresa, we used only this latter approach for comparison with pollen data. We compare uncorrected and corrected pollen percentages with vegetational data that were processed with distance-weighting functions. To assess the degree of correspondence between pollen and vegetation data we define a ratio pollen/vegetation, which allows a comparison at the taxon level. The best fit between total pollen load and vegetation is reached for a distance from the lake shore of ca. 300 m for Lago di Origlio (150×350 m in size) and of ca. 600 m for Lago di Muzzano (300×750 m in size). Beside these general patterns, our analysis reveals taxon-specific pollen dispersal patterns that are in agreement with results from previous studies in northern Europe. Ratios of species with local (proximal) and long-distance (distal) pollen dispersal provide evidence that pollen dispersal mechanisms can influence the size of the taxon-related pollen source area, from small (100–400 m) to large (>5 km) for the same lake. The proportion of distal species increases with increasing lake size, highlighting the predominance of atmospheric pollen transport. We conclude that the large species-related differences in pollen source areas have to be taken into account when the provenance at a site is estimated and discussed.     

Influences of sea level changes and volcanic eruptions on Holocene vegetation in Tonga

Here, we investigate Mid- to Late-Holocene vegetation changes in low-lying coastal areas in Tonga and how changing sea levels and recurrent volcanic eruptions have influenced vegetation dynamics on four islands of the Tongan archipelago (South Pacific). To investigate past vegetation and environmental change at Ngofe Marsh (‘Uta Vava’u), we examined palynomorphs (pollen and spores), charcoal (fire), and sediment characteristics (volcanic activity) from a 6.7-m-long sediment core. Radiocarbon dating indicated the sediments were deposited over the last 7700 years. We integrated the Ngofe Marsh data with similar previously published data from Avai’o’vuna Swamp on Pangaimotu Island, Lotofoa Swamp on Foa Island, and Finemui Swamp on Ha’afeva Island. Plant taxa were categorized as littoral, mangrove, rainforest, successional/ disturbance, and wetland groups, and linear models were used to examine relationships between vegetation, relative sea level change, and volcanic eruptions (tephra). We found that relative sea level change has impacted vegetation on three of the four islands investigated. Volcanic eruptions were not identified as a driver of vegetation change. Rainforest decline does not appear to be driven by sea level changes or volcanic eruptions. From all sites analyzed, vegetation at Finemui Swamp was most sensitive to changes in relative sea level. While vegetation on low-lying Pacific islands is sensitive to changing sea levels, island characteristics, such as area and elevation, are also likely to be important factors that mediate specific island responses to drivers of change.     

Phosphorus Deposition in a Low-Phosphorus Landscape: Sources,Accuracy and Contribution to Declines in Surface Water P

Atmospheric deposition can be an important source of phosphorus (P) to watersheds in geologically low-P landscapes like the Precambrian Shield, in south-central Ontario. Recent observations in this region of declining total P (TP) concentrations in streams and lakes, as well as P limitation in vegetation, prompted this investigation into the potential contributing role of atmospheric deposition. Long-term (1980–2014) measurements of bulk TP deposition at four monitoring stations, as well as wet-only deposition at one of the sites (2012–2014) and pollen measurements in 2014, were used to assess the potential contribution of biogenic sources to TP deposition in the region. Ratios of phosphorus to potassium (P:K) in historical bulk deposition and variability in P deposition among the four sites indicated that continuously open, bulk deposition collectors are enriched in ‘local’ biogenic sources of P, particularly in the summer, fall and spring months. Furthermore, in 2014 pollen could account for as much as 30% of bulk TP deposition, and pine trees (Pinus strobus) were the most important source of pollen (>75% of collected pollen). In contrast, winter bulk deposition was the least affected by local, biogenic sources, and likely best reflects net P inputs to this region from long-range (that is, external to the watershed) sources. Catchment input–output budgets for P that use measured annual bulk deposition totals as inputs are generally strongly positive (indicating net retention/accumulation of P within watersheds), whereas annual budgets calculated using extrapolated winter deposition are in approximate balance. Balanced P budgets are more consistent with regional observations of declining surface water TP and P limitation in terrestrial vegetation. Importantly, there was no significant decline in winter TP deposition over the period of record (1980–2014), which suggests that changes in regional deposition are not responsible for declines in surface water TP. Although we recommend extrapolating winter bulk deposition measurements to approximate annual atmospheric TP input to terrestrial ecosystems in non-dusty, oligotrophic landscapes such as this, local sources of P, particularly pollen, are likely important seasonal sources of P nutrition to lakes.     

Quantitative representation of local forest composition in forest‐floor pollen assemblages

1 We compared modern pollen assemblages from 60 moss polster sites in northern New York with forest composition data within 20–120 m of the sites using extended R ‐value (ERV) models, which correct for non‐linearities arising from use of pollen percentage data. Our sites were concentrated in two regions, one dominated by Tsuga and hardwood ( Acer , Betula , Fagus ) forests, and the other by Tsuga , Pinus , Betula , Acer and Quercus forests.
2 Our results confirm that forest‐floor pollen assemblages are dominated by pollen originating from trees growing more than 20 m from the site of deposition. However, our results suggest that background pollen percentages were overestimated by Jackson & Wong in 1994, owing to unusually high Pinus pollen production in the year of their sampling.
3 Expansion of our vegetation sampling radius from 20 to 120 m resulted in modest but consistent improvement in model fit and a decrease in background pollen percentages.
4 ERV model parameters (slope and background) differed substantially between the two study regions, primarily owing to differences in background pollen productivity and dispersal from regional sources.
5 High background pollen percentages may lead to poor estimation of calibration parameters in regions of complex vegetation patterns. Expansion of the vegetation sampling radius to reduce the background component may lead to better parameter estimates.
6 Calibration of pollen–vegetation relationships requires definition of the vegetation term so that it approximates the vegetation sampled by the pollen assemblages. Critical challenges are to define better the appropriate vegetation sampling area and distance‐weighting functions for application to pollen–vegetation calibration.     

Habitat associations of parrots on the Wallacean islands of Buru, Seram and Sumba

Aim To investigate patterns of habitat use in nineteen populations of parrots on three Wallacean islands, and examine the degree to which habitats used by a species can be predicted across islands. Location The Wallacean islands on Buru, Seram and Sumba. Methods A logistic regression method was used to relate the presence/absence of parrot species at census plots to vegetation axes derived from ‘gross’ habitat parameters such as altitude, tree sizes, and percentages of vegetation coverage at different strata. Receiver Operating Characteristic Plots were then used to test predictions of habitat use by conspecific populations on the nearby and distant islands, and by different parrot species on the same island. Results Parrots on Buru and nearby Seram tended to be associated with open-canopied lowland forests, while on the distant and low-lying island of Sumba, the preference was for closed-canopy forests at higher altitudes. There were usually close affinities in habitat usage by conspecific populations on the two nearby islands. One exception was where the species occurred on the two islands as separate subspecies. Habitat preferences on Sumba, however, were so distinct that paradigms developed for Buru and Seram failed to predict distributions on this distant island. Main conclusions For populations on more distant islands, it will usually be better to predict habitat use by considering the habitat used by other parrot species occurring on that island, than a conspecific population from another region. Conservation research programmes which concentrate parrot study on just one or two islands in every region of Wallacea may be more successful than those which ‘run through’ several islands in a group, but which miss out whole regions.     

A comparison between modern pollen spectra of moss cushions and Cundill pollen traps

Pollen spectra of 23 Cundill pollen traps from 23 different sampling sites in Southwest Turkey are compared with the corresponding pollen spectra of moss cushions from the same sites. The Cundill pollen traps represent the modern pollen rain data from one year whereas the moss cushions represent the pollen rain of several years. The comparative study reveals some main differences between the two pollen entrapment media. The one-year entrapment medium (pollen trap) appears to be more sensitive for local (releve area 10x10 m) and regional (100-500 m, or a few kilometres for Olea europaea) vegetation. Pollen spectra of moss cushions are dominated by high pine pollen percentage values and hardly sense fine vegetation structures. The conclusion of this comparative modern pollen study allows to interpret fossil sediment spectra from the Near East in a more critical way. It is concluded that one should preferably sample rapidly deposited sediments for palynological analyses, as the resulting highresolution pollen diagrams will be most informative about the former vegetation patterns.     

Diurnal variation in airborne pollen concentrations of the selected taxa in Zagreb, Croatia

The number of individuals allergic to plant pollen has recently been on a constant increase. The knowledge of diurnal distribution and abundance of allergenic pollen types, their patterns and response to source position and weather is useful to correlate hay fever symptoms with the presence of allergenic pollen in the atmosphere. The aim of this study was to determine diurnal distribution of total airborne pollen, pollen of particular allergenic taxa, possible variation in diurnal pollen distribution at measuring sites placed at different heights, and effect of some meteorological parameters on airborne pollen concentrations. A 7-day Hirst-type volumetric pollen trap was used for pollen sampling. Qualitative and quantitative pollen analysis was performed under a light microscope (magnification x400). Total pollen of all plant taxa (Ambrosia sp., Betula sp., Cupressaceae, Urticaceae, Poaceae, Quercus sp., Fraxinus sp., Alnus sp., Corylus sp., Populus sp., Pinus sp., Picea sp.) observed showed a regular diurnal distribution at both sampling sites in both study years, with a rise in the pollen concentration recorded after 4.00 a.m. and 6.00 a.m., respectively. The peak pollen concentration occurred between 12.00 a.m. and 4.00 p.m., and the lowest diurnal pollen concentrations were recorded overnight. About 50% of the 24-h pollen concentration were released to the atmosphere between 10.00 a.m. and 4.00 p.m. The timing and size of diurnal peaks were closely related to high temperature, low humidity and south-west maximum wind direction.     

新疆天山北坡地区中晚全新世古生物多样性特征

选择新疆天山北坡不同海拔高度、不同植被带和不同沉积年代的大西沟、桦树窝子、小西沟、东道海子和四厂湖剖面计算孢粉复合分异度指数,并根据已发表的有关文献中关于粒度、磁化率、烧失量等多项分析的数值整合结果探讨了本区中晚全新世以来生物多样性的时空变化和环境特征。森林草原过渡带的古生物多样性远高于高山、亚高山草甸带和荒漠带,孢粉复合分异度较高,鉴定的孢粉科属数较多。3600~3200 aB.P.、1700~1400 aB.P.、1000~600 aB.P.和450 aB.P.以来4个时段的生物多样性偏高。大西沟地区3600~3200 aB.P.期间和四厂湖地区中世纪期间气候均较湿润,其古生物多样性也较高,可见古环境演变和古生物多样性的动态变化具有相对一致性,孢粉复合分异度指数在一定程度上可以反映气候与环境的变化。但由于孢粉鉴定的精确度较差、受外来花粉的干扰等多种因素影响,使得孢粉多样性指数在反映生物多样性和古环境方面存在一定的偏差。     

Lateglacial and Holocene vegetation change on the Isle of Skye: new data from three coastal locations

Detailed lithostratigraphical and biostratigraphical studies from three contrasting coastal sites on the Isle of Skye were undertaken to investigate spatial variations in vegetation development on the island since the Lateglacial. The pollen profiles were then compared with the published pollen data for Skye. The new data reveal broadly similar trends for the sites studied and show that these correspond well to the previously established chronology. A date of ca. 10110 radiocarbon years B.P. has been obtained for the rational rise in Corylus, a date of ca. 8850 uncal B.P. for the appearance of Ulmus and a date of ca. 6600 uncal B.P. for the rise in Alnus. However, significant variations were evident and these were due to (a) coastal proximity, which directly affects vegetation composition through the low but persistent presence of Armeria and Chenopodiaceae, (b) altitude, which determines the effect of rising groundwater that may be related to relative sea level rise, and (c) aspect and position in relation to central Cuillin mountain range, through exposure to SW winds that restrict the development of thermophilous woodland and encourage the acidification of heathlands and development of herb-rich grasslands from ca. 4000 B.P.. This study reveals that the range of variation caused by these local factors is as great as the variation found throughout the island. This has negative implications for the reconstruction of regional patterns of vegetation change but highlights the range of local habitats and wild resources available to early inhabitants of the island.     

The effects of island area, isolation and source population size on the presence of the grayling butterfly Hipparchia semele (L.) (Lepidoptera: Satyrinae) on British and Irish offshore islands

Records of Hipparchia semele on British and Irish islands have been modelled against island area, isolation (sea and land distance) and the size of the nearest potential source populations. All three variables have been found to contribute significantly to the presence or absence of H. semele on the islands. Isolation is a more significant predictor than island area. This result differs from the multiple species case where area was found to be a more important influence than isolation. Records on islands are also shown to depend on the size of populations at the nearest sources; this underpins the relationships identified for the multiple species case, first, between the number of species on islands and at nearest sources and, second, between the incidence of species on islands and at nearest sources. There are clear indications that smaller islands may become increasingly marginalized for H.semele; with ongoing habitat loss, because isolation increases and source populations become sparser, the probability of H. semele recolonizing islands also decreases.     

Historical arthropod diversity patterns direct rehabilitation targets for Robben Island,South Africa,a continental island in a biodiversity hotspot

Island species are susceptible to extinction through disturbances such as habitat transformation. Due to the small size and isolation of islands, species have limited options for refuges and recolonization, making their rehabilitation a conservation priority. Robben Island is a continental island, isolated from the mainland ca. 15 000 years ago, and has been degraded by humans and alien species for nearly 400 years. Mainland areas with similar vegetation should be good reference sites for the biological restoration of the island due to historical connectedness. However, very little information exists as to which species were lost. Here we aim to identify the best mainland sites to use as reference sites for Robben Island based on remaining arthropod diversity on the island. Sites found to be most similar in terms of arthropod diversity to Robben Island were sites north of Robben Island (Elandsbaai and Dwarskersbos) rather than the geographically closest locations. These sites therefore represent ideal reference sites for biological restoration of the island. We do not suggest the reintroduction of species from these localities, but rather Robben Island should be restored to match their vegetation height and cover.     

阴山山脉东段花粉通量及其与表土花粉比较研究

阴山山脉东段蛮汉山和大青山山顶草甸、桦木林、虎榛子灌丛、人工油松林及杂草草原5个植被类型花粉通量与表土花粉对比研究发现,不同植被类型及同一植被类型不同样点花粉通量和花粉浓度差异明显,捕捉器样品花粉源区面积小于表土样品,能较好地反映样品点周围植被组成,但受虫媒植物和局地植物花粉影响,与表土样品花粉组合差异明显;表土样品中松属、桦属、蒿属及藜科花粉含量多高于捕捉器样品,表明这些花粉不仅易于传播,而且在表土中具有较强的保存能力;禾本科花粉具低代表性,花粉保存能力较低.地层花粉分析中应引起注意.     

Subrecent spore–pollen spectra and modern vegetation from the Lena River Delta,Russian Arctic

This paper presents the results of a pollen analysis of 40 surface samples collected from various geomorphological levels such as low and high floodplains, surfaces and slopes of terraces, tops of island remnants, etc., in different parts of the Lena River Delta. The obtained spore–pollen spectra (SPS) were compared with the modern vegetation composition. The role of long-distance transported pollen and the pollen of local plant biocenoses in the subrecent SPS formation is shown. Modern vegetation is most adequately reflected in the spectra of samples from a height range from 11 to 52 m above river level. SPS of other samples contain up to 40% of long-distance transported pollen, mostly Pinus s/g Haploxylon. This results in a significant decrease in the percentage of local pollen. The data offer the possibility of improving the reliability of palaeoclimatic reconstructions based on palaeopalynological data for Arctic regions.     

吉林省敦化地区晚全新世泥炭沼泽孢粉组合特征及古植被

对大桥泥炭地的孢粉进行了分析和年龄测定.根据孢粉组合、地层特征和¹⁴C年龄将剖面划分为4个孢粉带,分别为2 195±70～2 045±70 yr B.P.松为主的松-云冷杉-榛-栎-苔草组合带、2 045±70～1 745±70 yr B.P.苔草为主的松-桦-榛 胡桃-苔草组合带、1 745±70～705±70 yr B.P.松-榛-鹅耳枥-苔草-毛茛组合带和705±70 yr B.P.～1 950 yr A.D.松-云冷杉-桦-苔草组合带.周围植被由温带山地针叶林（类似现今海拔1 100 m以上）、温带针阔混交林下部（类似现今海拔400～600 m）和温带针阔混交林中部（类似现今海拔600～800 m）过渡到温带针阔混交林上部（类似现今海拔800～1 100 m）.相应地,泥炭沼泽经历了孕育期、蓬勃发展期、继续扩张期和消亡期.     

A glance into the vegetation and settlement of the fortress islands of Suomenlinna,Helsinki.: Pollen analytical study

Pollen studies were carried out of the sediments from a small pond on Susisaari, the largest island of the fortress of Suomenlinna, Helsinki. Three cores with altogether 80 samples were analysed, but only one core being reported here and additional comments are given on the results from two other cores. As it is most probable that the material has been mixed in several constructional phases during the last 100-200 years, the results were interpreted as one unit representing aerial and local vegetation during approximately that period. The results were compared with the recent vegetation, with the history of imported species, and with three botanical inventories made on the islands during the 20th century. Differences in pollen production and in pollen dispersal properties of different species strongly affect the results.     

Non-native pollen found in short-tailed bat (

We analysed pollen in short-tailed bat guano samples from Rangataua Forest and from guano and pollen found in bat holding bags used in the Kaimanawa Range, central North Island. Fifty seven percent of the pollen from Rangataua was from a previously unrecorded source and was tentatively identified as Trachycarpus fortunei (Chinese windmill palm). The significant remaining pollen was identified as Collospermum (15%) and Nothofagus (14%) from Rangataua, and Collospermum (90%) and Nothofagus (6%) from Kaimanawa. While the presence of Collospermum from both sites is consistent with previous work, pollen from T. fortunei, an exotic palm growing near Rangataua Forest, has not previously been found in association with short-tailed bats. Despite nocturnal surveillance with automated bat detectors and infra-red video cameras, we failed to confirm bat visitation to these palms. Nothofagus is wind-pollinated and pollen extracted from samples taken from both sites is probably wind-borne contamination. A collation of data from all available studies on the pollen found associated with short-tailed bats throughout New Zealand reveals that flowers from just four plant groups appear to be regularly used by bats: Collospermum spp., Knightia excelsa, Metrosideros spp. and, apparently, T. fortunei.