Standardizing names applied to pollen and spores in New Zealand Quaternary palynology

Abstract

Pollen analysts often study pollen grains that cannot be easily attributed to a particular taxon. As a consequence, a pollen type may be recorded under various names, leading to uncertainty in their interpretation by others. This limits the usefulness of pollen data for inclusion in databases or meta-analyses. To minimize any such ambiguity, we briefly describe the pollen grains of New Zealand gymnosperms and angiosperms and assign an appropriate name to each. Tree ferns are also included, as they are almost ubiquitous in New Zealand palynomorph records, and are often misidentified and misrepresented. We also provide a review of the occurrence of all taxa in New Zealand pollen records, and of pollination modes, pollen production and dispersal.

Keywords:

• pollen grains
• Filicopsida
• Pinopsida
• Magnoliopsida
• pollen record
• pollination mode
• pollen production
• pollen dispersal

Introduction

Pollen analysis is a valuable tool in diverse fields such as plant ecology, vegetation history and climate history, oil exploration and glaciology, and melissopalynology to determine the nectar source in honey. In all these studies, it is important to identify the pollen grains and spores as accurately as possible. However, not all pollen grains can be assigned with certainty to a particular taxon. Factors influencing pollen production, pollen morphology, pollen dispersal, the habit and range of source plants, deposition and preservation of pollen and spores (Wilmshurst & McGlone 2005a, b) and the process of sample preparation can all influence the final decision. Access to a regional pollen collection or pollen atlas (Moar 1993) may resolve some problems, but there are always grains that cannot easily be identified, either because they have not been encountered before or because their characteristic features may be obscured or destroyed by corrosion (Wilmshurst & McGlone 2005a). For these reasons, palynologists often differ in the name they apply to a particular pollen grain or spore.

Easily recognized grains may be recorded in different ways for no apparent reason. Thus, pollen derived from a monotypic taxon in New Zealand, e.g. Ascarina lucida, is often recorded at the generic level in the same way as pollen derived from a genus with several or many species, e.g. Metrosideros. Pollen also may be referred to as a family, e.g. Myrtaceae, without any attempt to separate genera, and therefore treated in the same way as those which are so similar as to defy easy recognition, e.g. Poaceae. In some cases, the pollen of genera such as Leptospermum and Kunzea (Myrtaceae) cannot be distinguished; in others, pollen of some genera within a family such as Podocarpaceae can be easily recognized, e.g. Dacrydium cupressinum, whereas others cannot, e.g. Halocarpus. A review of pollen diagrams published over the last 30 years shows that as pollen grains became progressively better known, the precision of identification has improved. Despite this, the ambiguities we refer to persist. For example, sedge pollen is usually referred to as Cyperaceae, but as Moar and Wilmshurst (2003) have shown, it is possible to achieve a more precise identification of the sedges by recognizing different pollen types within the group.

A difficulty for some who use pollen analytical data arises when new taxonomic names are adopted, but not explained by the palynologist. An example of this problem is the almost universal use of the subgeneric name Fuscospora for pollen formerly recorded as Nothofagus fusca type, or the changes for the taxon once known as Dacrydium colensoi, then as Lagarostrobos colensoi, and currently as Manoao colensoi. It is important that an author clearly states the nomenclatural basis for names used in diagrams and any associated discussion.

The above comments apply as much to fern spores as they do to pollen grains. However, it is not intended here to discuss spores in any detail (covered by Large & Braggins 1991) except those of Cyatheaceae and Dicksoniaceae. The spores of these tree ferns are found in most sediments because they are highly resistant to decay, and are often well preserved, as in peat, or variously degraded, as in lake sediments or in organic soils (Wilmshurst et al. 1999; Wilmshurst & McGlone 2005a). Although they are distinctive enough, it is not always easy to separate them from other similarly shaped spores such as Adiantum. Accordingly, we include the Cyatheaceae and Dicksoniaceae in this discussion; comments are based on Large & Braggins (1991).

The purpose of this review is to suggest how best to refer to the pollen types and tree fern spores recorded in a New Zealand pollen diagram so as to avoid ambiguity, and to make for a more uniform presentation of the data. We restrict our treatment to late Quaternary material in order to avoid conflict with form genera proposed by those working with older sediments, particularly since these often have no known modern plant source to which they can be assigned. In large families, e.g. Cyperaceae, or in large genera, e.g. Hebe, not all taxa have been considered, an admission which extends to recently revised taxa, e.g. Montia, which now includes eight species (Heenan 2007) rather than the one, Neopaxia australasica, previously recognized, and to pollen of the Orchidaceae and Poaceae, which to date have not been studied in detail in New Zealand. We suggest the following conventions for nominative use in pollen diagrams:

1.	Use a specific name when there is only one species represented in the region or country, e.g. Dacrycarpus dacrydioides and Ascarina lucida, or when the pollen or spores of one species of a genus is clearly different from the rest, e.g. Nothofagus menziesii and Drosera arcturi.
2.	Use a generic or subgeneric name when the pollen or spores of different species within a genus or subgenus are so similar that separation is difficult or impossible under routine analysis. Examples are Epilobium, Fuscospora (a subgenus of Nothofagus that includes N. fusca, N. solandri s.l. and N. truncata), Metrosideros and Myrsine.
3.	Use a generic name and ‘type’ when two or more genera share the same pollen or spore type. Examples are Leptospermum type to include pollen of Leptospermum and Kunzea, and Phyllachne type to include Phyllachne and Forstera. In selecting the generic name in these instances, we follow previous palynological usage.
4.	Use specific name followed by ‘type’ when the pollen or spores of different species within a genus can be identified into groups based on distinctive characters. For example, Myosotis pollen can be distinguished into Myosotis uniflora type, Myosotis angustata type and Myosotis australis type (Moar 1993); likewise, Cyathea spores can be distinguished into Cyathea smithii type and Cyathea dealbata type.
5.	Use a tribe or family name for recording pollen or spores above generic level as appropriate, e.g. Lactuceae, Poaceae.

Although it is hoped that these conventions will lead to a more uniform presentation of pollen diagrams, it is not suggested that reference to species be omitted from any discussion when it is appropriate to do so. Thus a pollen diagram from Westland may record the presence of Metrosideros, but it needs to be stated that the probable source is Metrosideros umbellata, especially if it is abundant. When this pollen is recorded in high numbers from Auckland Island peat samples, the source can only be that of Metrosideros umbellata. However, a conservative approach is recommended for most standard pollen diagrams and all databases. Distinctions should go no lower than the taxonomic level at which they can be reliably and consistently made.

In the following review, the pollen of most gymnosperm and most angiosperm genera indigenous to New Zealand is considered, some in greater detail than others. Some have never been recorded in a pollen analysis, some only occasionally, some frequently, while others are ubiquitous. However, all types are considered here because we may have overlooked some rare records of certain pollen grains, and some presently unrecorded types may be recognized in the future.

The figure in parentheses following the generic name refers to the number of species in that genus. Generic and specific names generally follow the Allan Herbarium (2002–2006), but more recent treatments are followed for family names and taxa at a higher taxonomic rank (Heywood et al. 2007; Mabberley 2008). For this reason, Dianella, Herpolirion and Phormium are included in Hemerocallidaceae and Arthropodium and Cordyline are assigned to Laxmanniaceae because the pollen in these two groups are quite different. It is convenient for palynologists to retain the traditional families Callitrichaceae, Plantaginaceae and Scrophulariaceae following Heywood et al. (2007) because of the distinct differences in their pollen morphology. For some genera, alternative familial affinities are indicated in parentheses. We provide brief descriptions of pollen grains based mainly on Cranwell (1953), Moar (1993) and Pocknall (1981a, b, c), and of the spores of tree ferns (Large & Braggins 1991), together with the suggested name to be used for each pollen and spore type recorded. Where relevant, we also provide brief comments. Finally, for each taxon we provide a categorical summary for the pollen occurrence, pollination mode, type of pollen production and dispersal based on data from published pollen diagrams to assist with the interpretation of pollen diagrams. These categories are summarized for convenience in Table 1. Dispersal refers to occurrence of significant quantities of a given pollen type, not simply presence. It is noted that for easy comparison, pollen and spore preparations from sediments, and from modern reference material (either fresh or herbarium specimens), are both processed using the standard acid treatment of acetolysis. Finally, we would like to acknowledge the previous work on pollen distribution by numerous palynologists in New Zealand, in particular Macphail & McQueen (1983).

Table 1  Categories used to summarise the occurrence, pollination mode, production and dispersal mechanism for each pollen type discussed.

Criteria	Category
Occurrence	X=the absence of any record in pollen analyses
OX–4	1=an occasional record
	2=a regular occurrence
	3=a regular occurrence, often in considerable numbers
	4=ubiquitous types dependent upon the nature of source vegetation
Pollination mode	W=wind-pollinated taxa
W, I, V, A	I=pollination by insects or other invertebrates
	V=pollination by vertebrates, birds, or other animals
	A=water pollinated
Pollen production	1=low pollen production
P1–3	2=moderate pollen production
	3=prolific pollen production
Dispersal	1=restricted distribution
D1–3	2=a distribution within a few kilometres of the source
	3=grains that may be widely distributed, often over great distances

POLYPODIOPSIDA (Filicopsida)

Cyatheaceae

genus: Cyathea (7).

spores: Trilete; triangular in polar view; exine thick (2–3 µm) in all taxa except C. smithii and C. kermadecensis (1–2 µm); spores generally smooth after acetolysis; size variable.

names: Cyathea: C. smithii type (C. colensoi, C. kermadecensis, C. smithii); C. dealbata type (C. cunninghamii, C. dealbata, C. medullaris, C. milnei).

comments: C. kermadecensis and C. milnei are endemic to the Kermadec group. In most analyses, the C. smithii type usually refers to C. smithii. C. smithii spores are often mistakenly identified as Pteridium esculentum despite the fact their exine is more than twice as thick as P. esculentum. Like Pteridium, tree fern spores may be over-represented through reworking.

summary: O4, W, P3, D2.

Dicksoniaceae

genus: Dicksonia 5.

spores: Trilete; triangular in polar view; exine thick (2–3 µm), thickens abruptly at angles; under light microscope surface sculpture varies from smooth or scabrate (D. fibrosa), granulate–tuberculate (D. lanata) or obviously tuberculate–reticulate (D. squarrosa); in these last two, taxa sculpture may vary between distal and proximal surfaces.

names: Dicksonia fibrosa; D. lanata; D. squarrosa.

comment: Spores of Dicksonia squarrosa are most frequently recorded.

summary: O4, W, P3, D2.

PINOPSIDA

Araucariaceae

genus: Agathis (1).

pollen: Grains spheroidal, inaperturate; exine thick, minutely scabrate or papillate.

name: Agathis australis.

comment: Grains may be variously degraded during preservation, especially through loss of papillae or the tectum.

summary: O2, W, P3, D2.

Cupressaceae

genus: Libocedrus (2).

pollen: Grains small, thin walled and spheroidal; exine bears scattered papillae.

name: Libocedrus.

comments: Grains are often split into two halves and consideration should be given to counting these as half grains. The papillae may be sloughed off. Pollen of introduced Cupressaceae is difficult to separate from Libocedrus. In post-European sites, pollen should be recorded as Cupressaceae.

summary: O3, W, P2, D3.

Podocarpaceae

genera: Dacrycarpus (1), Dacrydium (1), Halocarpus (3), Lepidothamnus (2), Manoao (1), Phyllocladus (3), Podocarpus (4), Prumnopitys (2).

pollen: Grains of all taxa are bisaccate with the exception of the trisaccate Dacrycarpus dacrydioides. Of the bisaccate grains, those of Dacrydium cupressinum are easily recognized by the reduced size of the pollen sacs, which form a shallow frill around the body of the grain. The rather larger saccae of Halocarpus with thick parallel markings allow for easy identification of this genus as do the small, plain, often partly folded, saccae of Phyllocladus. Pollen of Podocarpus totara and the related species, P. hallii and P. nivalis, are easily recognized by strongly folded (rugulate) exine over the proximal cap. Grains of Prumnopitys ferruginea and P. taxifolia are often confused; the larger grains, clearly projecting saccae, with conspicuous folds at the point of insertion, and thick proximal cap often smooth in the centre, usually separate the former from the latter. Pollen of Lepidothamnus and Manoao are similar, but the fine, regular reticulum of Manoao contrasts with the more open, broken reticulum of both Lepidothamnus intermedium and L. laxifolium.

names: Dacrycarpus dacrydioides; Dacrydium cupressinum; Halocarpus; Lepidothamnus; Manoao colensoi; Phyllocladus; Podocarpus; Prumnopitys ferruginea; Prumnopitys taxifolia.

comments: Sometimes the pollen of Lepidothamnus, Manoao, Podocarpus and Prumnopitys are variously distorted and so may defy ready identification. In such cases, refer to as bisaccate-Podocarpaceae. Podocarp pollen types are always recorded when source plants are present in the surrounding vegetation. Lepidothamnus is only rarely recorded, perhaps because it often resembles Prumnopitys taxifolia when not well displayed. It is not a major component of present-day vegetation.

summary: Dacrycarpus dacrydioides O4, W, P2, D3; Dacrydium cupressinum O4, W, P3, D3; Halocarpus O4, W, P2, D2; Lepidothamnus O1, W, P2, D1; Manoao colensoi O1 (4), W, P2, D2; Phyllocladus O4, W, P3, D3; Podocarpus O3, W, P3, D3; Prumnopitys ferruginea O3, W, P3, D3; Prumnopitys taxifolia O4, W, P3, D3.

MAGNOLIOPSIDA

1. NYMPHAEALES

Hydatellaceae

genus: Trithuria (1).

pollen: Grains monosulcate or monoporate; exine very thin, sparingly perforate and bearing minute broad-based spinules difficult to resolve; longest axis ∼ 36 µm.

name: Trithuria inconspicua.

comment: These fragile grains are easily distorted and to date have not been recorded in a pollen analysis.

summary: OX, W, P1, D1.

2. MAGNOLIIDAE s.s.

Lauraceae

genera: Beilschmiedia (2), Cassytha (1), Litsea (1).

pollen: Grains echinate, delicate, often collapsed; intectate; minor differences not easily resolved under bright field light microscopy.

name: Beilschmiedia type.

comments: Grains recorded only a very few times, and then usually as single examples; possibly one of the most under-represented pollen types in the flora. See discussion in Macphail (1980).

summary (general): O1, I, P1, D1.

Monimiaceae

genera: Hedycarya (1), Laurelia (1).

pollen: Grains of Hedycarya are inaperturate tetragonal tetrads; those of Laurelia are dicolpate, finely reticulate monads, often splitting into halves.

names: Hedycarya arborea; Laurelia novae-zelandiae.

summary: Hedycarya arborea O1, I, P1, D2; Laurelia novae-zelandiae O1, I, P1, D1.

Piperaceae

genera: Macropiper (2), Peperomia (2).

pollen: Grains small. Macropiper monocolpate; colpus circular, its membrane bearing coarse exine fragments; exine faintly verrucate. Peperomia acolpate; exine coarsely verrucate.

names: Macropiper; Peperomia.

comments: Macropiper mostly rarely recorded, but occasionally frequent as on Tawhiti Rahi Island, Poor Knights Islands (Wilmshurst & Moar unpubl. data); Peperomia not recorded in pollen analyses.

summary: Macropiper, Peperomia O1, W, P1, D1.

Winteraceae

genera: Pseudowintera (4).

pollen: Grains tetrahedral tetrads; exine boldly reticulate; each tetrad member monoporate.

name: Pseudowintera.

comment: The most widely distributed species is P. colorata, and in most instances grains recorded in Holocene samples are probably derived from plants of this taxon.

summary: O1, I, P1, D1.

3. CHLORANTHALES

Chloranthaceae

genus: Ascarina (1).

pollen: Grains bilaterally symmetrical, monocolpate; exine intectate, clavate; clavae arranged in reticuloid pattern.

name: Ascarina lucida.

comment: The characteristic clavae may be partly or completely stripped from the exine; in subfossil preparations there may be a gradation from complete grains through to those from which the clavae are completely absent.

summary: O3, W, P3, D2.

4. MONOCOTYLEDONS

Arecaceae

genus: Rhopalostylis (2).

pollen: Grains monosulcate; sulcus long, narrow, but wider at ends; exine up to 2 µm thick on proximal surface; exine mostly smooth or may be slightly pitted or vaguely rugulate.

name: Rhopalostylis.

comments: Unlikely to be confused with other monosulcate pollen types. R. cheesemanii (= R. baueri) occurs in the Kermadec group (Prebble & Dowe 2008), whereas R. sapida is the species of the two main islands. In the South Island it occurs south to Banks Peninsula and Hokitika in Canterbury and Westland, respectively. It is also found on Chatham Island. In the North Island it is recorded from sites in northern North Island, Great Barrier Island (Horrocks et al. 2002), and Tawhiti Rahi of the Poor Knights Islands (Wilmshurst & Moar unpublished data). Most, if not all, records from the main islands refer to R. sapida.

summary: O1 (4), I (W), P3, D1.

Asphodelaceae (Xanthorrhoeaceae)

genus: Bulbinella (6).

pollen: Grains monosulcate; sulcus very long; exine rugulate, especially so in B. rossii; rugulae often absent on distal surface especially along sulcus margin.

names: Bulbinella rossii; Bulbinella.

comments: The only grain easily distinguished is that of B. rossii because of its much bolder rugulae. This grain is frequently recorded from Auckland and Campbell islands (McGlone 2002). Pollen of the other taxa, restricted to the main islands, is much less boldly rugulate and only occasionally recorded. Although there are subtle differences in the nature and distribution of the rugulae of these grains, identification is difficult. Rugulae, which are least well developed on the grains of B. modesta, are sometimes absent, which may provide a means of further separation. However, a decision needs to be made taking into account the location of the site and the number of grains recorded during an analysis.

summary: Bulbinella rossii O3, I, P2, D1; Bulbinella O1, I, P2, D1.

Asteliaceae

genera: Astelia (13), Collospermum (2).

pollen: Grains of Astelia and Collospermum cannot be confused. Although both Astelia and Collospermum grains are monosulcate, and in both the sulci are long, Astelia grains are elliptic in shape and bear spinules up to 2 µm long, whereas Collospermum grains are generally spheroidal, and the sulcus is often very broad. Unlike C. microspermum, the exine of C. hastatum is perforate and bears scattered very short, broad-based spinules.

names: Astelia; Collospermum hastatum; Collospermum microsporum.

comments: Both genera are recorded in pollen analyses, but never in great numbers. However, in the subantarctic Auckland and Campbell Islands, Astelia pollen, either A. linearis or A. subulata, may be an important element in pollen diagrams. Collospermum pollen, probably C. hastatum, is recorded occasionally from North Island sites.

summary: Astelia O1 (2), W (V), P2, D1; Collospermum hastatum, C. microsporum O1, W (V), P2, D1.

Centrolepidaceae

genera: Centrolepis (4), Gaimardia (1).

pollen: Grains ulcerate; the aperture rims formed by the irregular and abrupt ending of the exine; the exine is perforate (scrobiculate) and the scrobiculi are often joined by shallow canals.

names: Centrolepis type (mainland sites); Centrolepis for subantarctic sites.

comment: Gaimardia is not recorded from the subantarctic islands.

summary: O1 (3), W, P1, D1.

Colchicaceae

genus: Iphigenia (1).

pollen: Grains monosulcate; sulcus very long, sometimes wider at end; exine very thin, ∼ 1 µm, minutely reticulate, heterobrochate; shape elliptic to oval in polar view, longest axis up to 34 µm.

name: Iphigenia novae-zelandiae.

comment: Not recorded in pollen analyses.

summary: OX, I?, P1, D1.

Cyperaceae

genera: Baumea (7), Carex (80), Carpha (1), Cyperus (2), Desmoschoenus (1), Eleocharis (5), Fimbristylis (1), Gahnia (6), Isolepis (14), Lepidosperma (3), Machaerina (1), Morelotia (1), Oreobolus (3), Schoenus (9), Tetraria (1), Uncinia (32).

pollen: Two main types may be distinguished, spheroidal and variously cup-shaped or pear-shaped. Spheroidal grains occur in Baumea, Fimbristylis, Lepidosperma, Machaerina, Tetraria and Oreobolus. Of these, Fimbristylis pollen is recognized by its coarsely verrucate exine and poorly defined apertures; the exine of Baumea, Lepidosperma, Machaerina, Tetraria and Oreobolus is smooth, and of these, the apertures of Oreobolus are difficult to resolve, in contrast to those of the other four genera. The grains of all other genera are broadly rectangular, and usually either cup-shaped or pear-shaped. The most easily recognized are Eleocharis and Schoenus; the former by its very large elongate apertures, relative to grain size, and the latter by its narrow wedge shape. Pollen of E. sphacelata is distinguished from other species of that genus by its larger size, recorded as being between 35 and 46 µm long, in contrast to the length of the other species (24–31 µm). Of the other taxa, Isolepis and Uncinia grains are mostly cup-shaped, and with ± circular or elliptical lacunae. These characters separate them from the remaining genera, which possess short lacunae, never more than half the length of the grain (Carex), by elongate and narrow lacunae (Cyperus ustulatus), or by diffuse, often very narrow elongate lacunae (Carpha alpina). For further detail consult Moar & Wilmshurst (2003).

names: Baumea type (Baumea, Lepidosperma, Machaerina and Tetraria); Carex; Carpha alpina; Cyperus ustulatus; Eleocharis sphacelata; Eleocharis; Fimbristylis squarrosa; Gahnia type (Gahnia and Morelotia); Isolepis type (Isolepis and Uncinia); Schoenus type (Bolboschoenus, Desmoschoenus, Schoenoplectus and Schoenus); Oreobolus.

comments: These notes, especially regarding large genera such as Carex, are based on a limited number of taxa and should be seen as a guide only to identifying the pollen of this difficult group. Pollen of the Baumea type, Isolepis type, Eleocharis and Schoenus type are the most easily recognized; if there is any doubt, refer to Cyperaceae.

summary (general): O4, W, P2–3, D1.

Hemerocallidaceae (Xanthorrhoeaceae)

genera: Dianella (3), Herpolirion (1), Phormium (2).

pollen: Grains trichotomosulcate; exine reticulate (Phormium), or microreticulate (Dianella, Herpolirion); shape triangular or sub-triangular; Phormium tenax grains, 35–37 µm, are the largest; P. cookianum, 27–32 µm, slightly smaller; Dianella, 19–27 µm, and Herpolirion, 22–31 µm, are considerably smaller and may be difficult to separate.

names: Herpolirion type (Dianella and Herpolirion); Phormium (P. tenax and P. cookianum). Dianella pollen has been occasionally recorded (e.g. Mildenhall 1980b).

comments: Of these genera, Phormium is the most likely to be recorded. Phormium pollen is extremely under-represented in relation to its prominence in many wetland areas.

summary: Herpolirion type O1, I, P1, D1; Phormium O1, V (I), P2, D1.

Iridaceae

genus: Libertia (8).

pollen: Grains monosulcate; exine microreticulate; size variable between 27 and 40 µm.

name: Libertia

comment: Rarely recorded in pollen analyses (Mildenhall 1980b).

summary: O1, W (I), P1, D1.

Juncaceae

genera: Juncus (16), Luzula (11), Marsippospermum (1), Rostkovia (1).

pollen: Grains tetrahedral tetrads; each member of the tetrad bears a large pore occupying most of the distal surface; exine very thin, delicate; faintly verrucate.

name: Juncaceae.

comments: Acetolysis usually destroys the internal walls of the tetrad leading to its collapse and destruction. Rarely recorded in pollen analyses (Harris & Mildenhall 1980).

summary: O1, W, P2, D1.

Juncaginaceae

genus: Triglochin (2).

pollen: Grains inaperturate; finely reticulate, reticulum not easily resolved; shape ± spheroidal; size ∼ 20 µm.

name: Triglochin.

comments: The poorly resolved reticulum separates Triglochin from Potamogeton in which the reticulum is clearly resolved. Not recorded in pollen analyses.

summary: OX, W, P1, D1.

Laxmanniaceae (Asparagaceae)

genera: Arthropodium (3), Cordyline (5).

pollen: Grains monosulcate; sulcus very long. Exine of Arthropodium ∼ 1 µm thick, reticulate, heterobrochate; simplibaculate; longest axis ∼ 45 µm. Exine of Cordyline varies in thickness from < 1 µm (C. australis) to 2 µm (C. pumilio); surface undulating; surface of C. indivisa sparingly perforate, perforations small, visible only by lux obscuritas analysis (LO, see: Moore et al. 1991).

names: Arthropodium; Cordyline.

comments: Pollen of these two genera cannot be confused. The latter is more likely to be recorded, and then only occasionally.

summary: Arthropodium O1, I, P2, D1; Cordyline O1, I, P2, D1.

Luzuriagaceae (Alstroemeriaceae)

genus: Luzuriaga (1).

pollen: Grains monosulcate; sulcus often circular, occupying much of distal pole, margins irregular; exine surface areolate; shape spheroidal; size ∼ 32 µm.

name: Luzuriaga.

comment: Rarely recorded in pollen analyses (Mildenhall 1980a).

summary: O1, I, P1, D1.

Orchidaceae

comments: Detailed comment is reserved until ongoing revisions are completed (Brian Molloy personal communication). Pollen may occur as tetrads, monads, or as pollinia. All grains examined are reticulate; the exine of different taxa varies in thickness, and there is considerable variation in size of individual grains. Pollen can be identified to family level without difficulty, but they are rarely seen in the fossil record.

name: Orchidaceae.

summary: O1, I, P2, D1.

Pandanaceae

genus: Freycinetia (1).

pollen: Grains monoaperturate; aperture ± circular, grain pouch-shaped.

name: Freycinetia banksii.

comment: Sometimes recorded in high numbers, especially in Westland.

summary: O1, I (V), P3, D2.

Poaceae

genera: There are 34 genera native to New Zealand distributed between five subfamilies, Ehrhartoideae, Pooideae, Arundinoideae, Panicoideae and Chloridoidaea (Edgar & Connor 2000); pollen of Chloridoidaea have not been examined. Genera examined are: Microlaena (Ehrhartoideae); Achnatherum, Anemanthele, Austrostipa, Deschampsia, Deyeuxia, Elymus, Festuca, Hierochloe, Lachnagrostis, Poa, Trisetum (Pooideae); Chionochloa, Cortaderia, Rytidosperma (Arundinoideae); Isachne, Spinifex (Panicoideae). Some genera, e.g. Anemanthele, are represented in New Zealand by only one species, others by numerous species, e.g. Chionochloa and Rytidosperma.

pollen: Grains monoporate; pores with thickened annulus; exine ∼ 1 µm thick, tectate, baculate; bacula often difficult to resolve; at 1000× the tectum may carry separate spinules (e.g. Chionochloa), groups of spinules on ‘islands’ separated by shallow channels (e.g. Poa), or there may be an intermediate state in which some spinules are in groups, whereas others appear to be single (e.g. Cortaderia). The apparent ‘islands’ are interpreted to be the result of an undulating exine surface; shape spheroidal; size based on the limited number of genera examined varies between 23 and 49 µm.

names: Chionochloa type (all species of Chionochloa examined, Lachnagrostis filiformis); Poa type (species examined Deyeuxia aucklandica, Hierochloe fusca, H. brunonis, Poa breviglumis, P. cita, P. colensoi, Rumytidosperma caespitosa, R. unarede, Spinifex sericeus, Trisetum spicatum); Cortaderia type (Cortaderia spp., Hierochloe redolens, Rytidosperma clavatum and R. maculatum). Whenever in doubt, refer to Poaceae.

comments: A close examination of the exine surface, especially by phase contrast, offers some hope for a better understanding of the pollen grains, but studies by electron microscope, scanning electron microscope in particular, are needed to make any progress. Chionochloa species, and others, have clearly separated spinules although the degree of separation may depend upon the vagaries of preparation. Taxa in which ‘islands’ are most consistently seen occur in the genus Poa, although there is considerable variation between species. Pollen of 10 specimens named Poa colensoi collected by B. Molloy from different parts of Otago and Canterbury, and one from Mount Ruapehu, were examined. The exine ‘islands’ varied considerably; some were long and narrow, some were ± circular and some, especially on pollen from the Mount Ruapehu specimen, were broad and ± rectangular in shape. Long and narrow ‘islands’ have been described as ruguloid. Intermediate arrangements in which the spinules are aggregated into groups are found in the Cortaderia type.

summary (general): O4, W, P2, D3.

Potamogetonaceae

genus: Potamogeton (4).

pollen: Grains inaperturate; exine very thin, semitectate, baculate, reticulate; reticulum heterobrochate, delicate.

name: Potamogeton.

comment: Pollen grains of Potamogeton and Triglochin (Juncaginaceae) are similar but the easily resolved, delicate reticulum in Potamogeton separates it from Triglochin.

summary: O1, W (A), P2, D1.

Restionaceae

genera: Empodisma (1), Leptocarpus (1), Sporadanthus (2).

pollen: Grains monoporate; pores large, often ± circular; exine perforate (scrobiculate), undulate and marked by grooves or channels that emphasize the undulate nature of the exine, especially in Sporadanthus.

names: Empodisma type (Empodisma and Leptocarpus); Sporadanthus.

comments: Of these three genera, Sporadanthus is most easily recognized by its clearly undulate tectum, by the narrow clear grooves or channels separating the undulations, and by the numerous evenly distributed perforations (scrobiculi). There are subtle differences between Empodisma and Leptocarpus, but ecological considerations will often offer the best clue as to identification.

summary (general): O1, W, P2, D1.

Smilacaceae

genus: Ripogonum (1).

pollen: Grains monosulcate; sulcus extending along length of distal surface; exine microreticulate; muri very narrow, bearing very short spinules.

name: Ripogonum scandens.

comments: The spinules are not easily resolved under bright field light microscopy or even with phase contrast and require scanning electron microscopy study for confirmation. The pollen has been recorded as frequent in some Taranaki sites.

summary: O1, I (W), P1, D1.

Typhaceae

genera: Sparganium (1), Typha (1).

pollen: Grains monoporate, pore circular; exine slightly raised around pore to form a distinct ring; exine reticulate, muri of variable thickness and mostly duplibaculate.

names: Sparganium subglobosum; Typha orientalis.

comment: Typha and Sparganium may be distinguished by differences in exine structure and by differences in the pore margins.

summary: Sparganium subglobosum O1, W, P1, D1; Typha orientalis O1 (3), W, P3, D1.

Xeronemataceae

genus: Xeronema (1).

pollen: Grains monosulcate; exine microreticulate, muri coarse, duplibaculate.

name: Xeronema callistemon.

comments: Occasionally recorded on Poor Knights group (Wilmshurst & Moar unpublished data); pollen of Xeronema and Arthropodium (Laxmanniaceae) are similar although the duplibaculate muri of the former separate it from the latter.

summary: O1, V, P1 (2), D1.

5. EUDICOTS

Argophyllaceae

genus: Corokia (3).

pollen: Grains tricolporate, endoapertures H-shaped; exine faintly rugulate.

name: Corokia.

comment: Not noted in pollen analyses.

summary: OX, I, P1, D1.

Aizoaceae

genera: Disphyma (2), Tetragonia (2).

pollen: Grains tricolpate (Disphyma australe, Tetragonia tetragonioides) or mostly tetracolpate (D. papillatum, T. trigyna); tectum minutely spinulose (D. australe, Tetragonia spp.), reticulate (D. papillatum), perforate (D. australe); ectoapertures have raised margins producing a pouting effect in polar view, sometimes syncolpate (D. australe).

names: Disphyma australe type; D. papillatum type.

comment: Only rarely noted in pollen analyses.

summary (general): O1, I, P1, D1.

Alseuosmiaceae

genus: Alseuosmia (7).

pollen: Grains variable, mainly tricolporate, angulaperturate; tectum verrucate or rugulo-verrucate.

name: Alseuosmia.

comment: Rarely recorded (Mildenhall 1980b).

summary: O1, I, P1, D1.

Amaranthaceae

genus: Alternanthera (2).

pollen: Grains with ∼ 12 deeply recessed apertures, separated by narrow, finely spinulose muri.

name: Alternanthera.

comments: These grains cannot be confused with any other New Zealand pollen type; they are rarely recorded in a pollen analysis (Mildenhall 1980b). The two species in New Zealand are A. nahui and A. denticulata. Of these, the former is endemic, but the status of the latter is uncertain and may be naturalized.

summary: O1, W, P2, D1.

Apiaceae

genera: Aciphylla (40), Actinotus (1), Anisotome (16), Apium (1), Azorella (1), Centella (1), Chaerophyllum (4), Daucus (1), Eryngium (1), Gingidia (7), Hydrocotyle (10), Lignocarpa (2), Lilaeopsis (2), Scandia (2), Schizeilema (11), Stilbocarpa (3).

pollen: Grains easily recognized at the family level; identification sometimes possible at the generic level, and occasionally at the specific level.

Aciphylla, Anisotome and Scandia are similar, but Aciphylla pollen is distinguished by its obviously prolate shape and pointed poles, in contrast to the sub-prolate shape and rounded poles of Anisotome and Scandia.

Actinotus, Apium, Daucus and Lilaeopsis are easily confused, although the long ectoapertures of Actinotus and Apium (> two-thirds length of grain) contrast with those of Daucus and Lilaeopsis (two-thirds length of grain or less).

Centella and Eryngium are similar, although some grains of Centella may be tetra- or hexacolporate.

Chaerophyllum pollen may be recognized by the crimped appearance of the exine when seen in optical, equatorial view.

Gingidia and Lignocarpa pollen grains are characterized by markedly protruding apertures and are difficult to separate.

Hydrocotyle grains have obviously ‘offset’ polar exine, a characteristic shared with Azorella and with some Schizeilema. Small rhomboidal grains with pointed or rounded poles identify Azorella selago and H. moschata. Pollen of other taxa may vary in size, aperture number and surface pattern, but separation is difficult, especially as few grains are recorded in any pollen count. However, the larger grains of H. novae-zeelandiae may sometimes be recognized.

Schizeilema is identified by its strongly ‘bowed’ shape in equatorial view; polar exine in S. hydrocotyloides is strongly ‘offset’, and tetracolporate grains occur in S. allanii and S. colensoi.

Stilbocarpa grains are psilate, prolate and have narrow lalongate endoapertures fused to form an equatorial ring.

names: Aciphylla; Anisotome type (Anisotome, Scandia); Centella type (Centella, Eryngium); Chaerophyllum; Gingidia type (Gingidia, Lignocarpa); Hydrocotyle type (Hydrocotyle moschata, Azorella selago); Hydrocotyle pterocarpa type (Hydrocotyle dissecta, H. heteromeria, H. novae-zeelandiae, H. pterocarpa, H. sulcata); Lilaeopsis type (Lilaeopsis novae-zelandiae, L. ruthiana, Actinotus novae-zealandiae, Apium prostratum, Daucus glochidiatus); Schizeilema, Stilbocarpa.

comments: Aciphylla, Hydrocotyle type and Lilaeopsis type pollen are most frequently encountered. Although Azorella selago pollen is close to that of Hydrocotyle moschata type, most grains from Holocene sites are likely to be derived from Hydrocotyle moschata.

summary (general): O1, I, P2, D1.

Apocynaceae

genus: Parsonsia (3).

pollen: Grains heteropolar, diporate–tetraporate, but mostly triporate.

name: Parsonsia.

comment: Occasionally recorded; may be confused with Geniostoma (Loganiaceae).

summary: O1, I, P2, D1.

Araliaceae

genera: Meryta (1), Pseudopanax (14), Raukaua (3), Schefflera (1).

pollen: Typically the pollen is reticulate, subprolate, and the endoapertures are lalongate, relatively short and rectangular in shape. Pollen of Meryta sinclairii has the boldest reticulum with lumina up to 5 µm across and is distinct from other New Zealand Araliaceae. The reticulum of Raukaua type (Mitchell et al. 1997) is very small, often < 1 µm across, and sometimes difficult to resolve. Pollen of Pseudopanax chathamicus, P. crassifolius, P. discolor, P. ferox and P. linearis are close in general appearance to those of Raukaua, but the reticulum is easily resolved. Pollen of the remaining Pseudopanax species are bolder, and P. colensoi is recognized either by a markedly rugulate mesocolpium or by irregularly spaced verrucae and spines. The grains of the remaining species of Pseudopanax, including P. arboreus, come between the two extremes noted above, while Schefflera is generally smaller, prolate–spheroidal, and the rectangular endoapertures are sometimes H-shaped.

names: Meryta sinclairii; Pseudopanax colensoi; Pseudopanax arboreus type (P. arboreus, P. kermadecensis, P. laetus); Pseudopanax crassifolius type (P. chathamicus, P. crassifolius, P. discolor, P. ferox, P. linearis); Raukaua (R. anomalus, R. edgerleyi, R. simplex); Schefflera digitata. Refer to Araliaceae if uncertain.

comments: Pollen of the Pseudopanax arboreus type are most frequently observed. P. colensoi is sometimes seen in higher altitude sites from the South Island.

summary (general): O1 (2), I, P2, D2.

Asteraceae

genera: There are 32 genera distributed among five tribes, the Anthemideae, Astereae, Gnaphalieae, Lactuceae and Senecioneae.

pollen: Grains of Anthemideae and Lactuceae are easily recognized, but those of the Astereae, Gnaphalieae and Senecioneae are often difficult to separate.

Pollen of the Anthemideae are characterized by a two-layered exine comprising an inner layer of coarse bacula supporting a diffuse tectum carrying short, fine bacula, which in turn support a perforate, echinate tectum. Haastia grains (tribe Senecioneae) also have a two-layered exine, but the inner layer is amorphous and has no obvious baculi; these have not been recorded.

Lactuceae grains are fenestrate and characterized by large depressed areas (lacuneae), separated by high, narrow ridges of echinate exine. The complexity of the apertures, involving an ectoaperture, a mesoaperture, and an endoaperture, is not often clear. Of the six genera (Embergeria, Kirkianella, Microseris, Picris, Sonchus, Taraxacum) Embergeria is easily recognized by its long spines, up to 6 µm long, and by much reduced lacunae and exine ridges; it has never been recorded.

Pollen of some in the Astereae, including Olearia colensoi, O. lyallii, Pleurophyllum criniferum and P. hookeri, are characterized by short, close-set spines, and together comprise a well-defined group. Others, including Pleurophyllum speciosum, Damnamenia vernicosa, most Celmisia spp. and Pachystegia spp., characterized by acute, generally crowded spines 4–5 µm long, cannot easily be distinguished from each other.

Pollen of Raoulia and Ozothamnus (Gnaphalieae) are difficult to separate. Their ectoapertures are rarely more than half the length of the grain, which distinguishes them from pollen of other tribes, and their spines are usually no more than 2–3 µm long.

Senecioneae pollen, like the Lactuceae, possess compound apertures (Moar 1993); spines tend to be blunt, in contrast to the acutely tipped spines of other groups, and in many a hole is visible (e.g. Brachyglottis repanda); ectoapertures are mostly more than two-thirds the length of the grain, and the lalongate endoapertures are always very long, up to 14 µm or longer.

All other taxa are so similar that they cannot easily be distinguished and may be recorded as Asteraceae.

names: Leptinella type (Leptinella, Cotula); Embergeria grandifolia; Kirkianella type (all other Lactuceae); Olearia colensoi type (O. colensoi); Celmisia type (all Celmisia spp., Damnamenia vernicosa, Pachystegia spp., Pleurophyllum speciosum); Raoulia type (all Raoulia spp., Ozothamnus); Senecio type (all Senecioneae); Asteraceae for all others or whenever recognition is uncertain.

comments: Grains of the Anthemideae and Lactuceae are generally distinguished easily enough, but those referred to the other three tribes are more difficult and see notes above. Because Olearia lyallii is considered to have been introduced to the Auckland Islands early in the 19th century, pollen of the O. colensoi type recorded from the Auckland Islands before the European era may be referred to as Pleurophyllum. Embergeria has not been recorded.

summary (general): O3, I, P3, D2.

Avicenniaceae (Acanthaceae)

genus: Avicennia (1).

pollen: Grains tricolporate; endoapertures small, slightly lolongate with sinuous margins, often indistinct; exine up to 3 µm thick, reticulate, muri generally wide.

name: Avicennia marina.

comments: Unlikely to be confused with pollen of other taxa and occasionally recorded. Mildenhall and Brown (1987) have recorded it in mid-Holocene sites south of its present limit in the North Island.

summary: O1, I, P1, D1.

Balanophoraceae

genus: Dactylanthus (1).

pollen: Grains characterized by 9–12 irregularly distributed circular pores, each with an obvious annulus.

name: Dactylanthus taylorii.

comments: Pollen of this bat-pollinated plant is occasionally recorded in late Quaternary sediments (Macphail & Mildenhall 1980). For a rare, low-growing plant, it is surprising that it is recorded at all. Wilmshurst (1995) suggests that the pollen may be distributed in bat droppings.

summary: O1, V, P1, D1.

Bignoniaceae

genus: Tecomanthe (1).

pollen: Grains tricolpate; ectoaperture split in two or more places; exine boldly reticulate, reticulum heterobrochate; muri relatively coarse, duplibaculate.

name: Tecomanthe speciosa.

comments: Restricted to the Three Kings Islands. The splits across the ectoaperture are similar to those of Myoporum, but the bold reticulum makes confusion unlikely. Not recorded in a pollen analysis.

summary: OX, I, P1, D1.

Boraginaceae

genera: Myosotidium (1), Myosotis (34).

pollen: Grains of both genera are heterocolpate. Myosotidium hortensium pollen is very small, mostly < 11 µm long, in contrast to the larger, but still small grains of Myosotis which may be up to 22 µm long. Myosotis grains are divided into three groups: (a) Simple colpi are parasyncolpate forming a characteristic polar cap, e.g. M. uniflora. (b) Grains are flattened at the poles and constricted at the equator, e.g. M. angustata. (c) Grains are rounded at the poles and not constricted at the equator, e.g. M. australis.

names: Myosotidium hortensium; Myosotis uniflora type; Myosotis angustata type; Myosotis australis type.

comments: Of the Myosotis pollen grains those of M. uniflora type are the most easily recognized. If there is any doubt, refer to Myosotis. Myosotidium has been recorded from the Chatham Islands by Mildenhall (1994).

summary: Myosotis types O1, I, P1, D1; Myosotidium hortensium O1, I, P1, D1.

Brassicaceae

genera: Cardamine (7), Lepidium (9), Notothlaspi (2), Pachycladon (9), Rorippa (3).

pollen: Grains tricolpate; reticulate; bacula and lumina vary, up to 3 µm long and 3–4 µm across (Cardamine and Notothlaspi), bacula 2 µm long and lumina 2 µm across (Lepidium, Pachycladon, Rorippa), or bacula and lumina not more than ∼ 1 µm (Pachycladon cheesemanii); easily recognized at the family level.

name: Cardamine type (Cardamine, Notothlaspi); Lepidium type (Lepidium, Pachycladon Rorippa); Pachycladon cheesemanii.

comments: Brassicaceae pollen is not often recorded; the three pollen types are not easily distinguished. If in doubt, refer to Brassicaceae.

summary (general): O1, I, P1, D1.

Callitrichaceae (Plantaginaceae)

genus: Callitriche (4).

pollen: Grains small, thin walled; reticuloid or retipilate; vaguely colpoid.

name: Callitriche.

comments: Occasionally recorded; exine structure and the amphibious or terrestrial habit of most taxa suggest that pollination may be anemophilous (see Martinsson 1993) or geitonogamous (Osborn & Philbrick 1994). Molecular studies support inclusion in the Plantaginaceae, but here we follow Heywood et al. (2007) who on the basis of morphological characteristics prefer retention of the family Callitrichaceae. Pollen cannot be confused with Plantago or members of the Scrophulariaceae.

summary: O1, A?, P1, D1.

Campanulaceae

genera: Colensoa (1), Lobelia (11), Wahlenbergia (10).

pollen: Grains usually triporate (Wahlenbergia), or tricolporate (Colensoa, Lobelia). Exine bearing very short, evenly distributed spinules (Wahlenbergia); mostly finely reticulate in Colensoa and Lobelia, but rugulo-striate in L. fatiscens. Endoapertures lalongate in Lobelia, but H-shaped in Colensoa.

names: Colensoa physaloides; Lobelia; Wahlenbergia.

comments: Pollen grains of C. physaloides, restricted to the northern North Island, are not recorded; pollen of Lobelia is rarely recorded; and the pollen of W. gracilis is triporate–hexaporate, but mostly tetraporate. The occasional grains noted during pollen analysis are mostly triporate.

summary (general): O1, I, P1, D1.

Caryophyllaceae

genera: Colobanthus (16), Scleranthus (3), Spergularia (1), Stellaria (6).

pollen: Grains periporate except Spergularia, which are tricolpate; exine spinulose, except in Scleranthus and Spergularia; perforate in all taxa; number of pores in Stellaria <20, in Colobanthus and Scleranthus >30.

names: Colobanthus; Scleranthus; Spergularia media; Stellaria.

comments: Colobanthus and Scleranthus are close, but the absence of spinules in Scleranthus separates the two genera. Colobanthus is the most frequently recorded of the two genera.

summary (general): O1, I, P1, D1.

Celastraceae

genus: Stackhousia (1).

pollen: Grains reticulate; bacula short, often projecting from the floor of the lumina singly, in pairs, or in groups of three.

name: Stackhousia minima.

summary: O1, I, P1, D1.

Chenopodiaceae (Amaranthaceae)

genera: Atriplex (4), Chenopodium (3), Einadia (2), Sarcocornia (1), Suaeda (1).

pollen: Grains periporate, often with > 70 pores, but number varies considerably, e.g. Einadia has ∼ 60, Suaeda ∼ 80, and Sarcocornia may have as many as 90. These numbers are not helpful in pollen analysis because there is considerable variation in both pore number and size in any taxon, a difficulty compounded by the very few grains usually recorded in any sample.

name: Chenopodiaceae.

comments: Sarcocornia pollen is the most distinctive in terms of pore number and size. If site characteristics are appropriate, and the number of grains numerous enough, Sarcocornia quinqueflora may be used. Some pollen may be derived by long-distance transport from Australia, e.g. McGlone & Meurk (2000).

summary: O1, W, P1, D1.

Convolvulaceae

genera: Calystegia (3), Convolvulus (3), Dichondra (2), Ipomoea (2).

pollen: Grains either tricolpate (Convolvulus, Dichondra) or periporate (Calystegia, Ipomoea). Convolvulus is the larger of the tricolpate grains (polar axis 55–62 µm), and unlike the very thin, irregularly spinulose exine of Dichondra, the exine of Convolvulus is up to 5 µm thick, perforate and coarsely baculate. The periporate grains (Calystegia, Ipomoea) are spheroidal, very large, up to 91 µm for Ipomoea, with a thick (5 µm) perforate exine. Calystegia grains are spinulose, those of Ipomoea bear spines up to 8 µm long with very short spinules scattered in between.

names: Convolvulus; Dichondra; Calystegia; Ipomoea.

comment: Only Dichondra and Calystegia recorded and then only occasionally.

summary (general): O1, I, P1, D1.

Coriariaceae

genus: Coriaria (8).

pollen: Grains small, usually tricolporate; ectoapertures short, slit-like; endoapertures short, about as broad as long.

name: Coriaria.

comments: Tetracolporate grains may occur. The generally smaller grains, shorter apertures, and the absence of endocracks easily separate Coriaria from Coprosma.

summary: O1, W, P2, D2.

Corynocarpaceae

genus: Corynocarpus (1).

pollen: Grains dicolporate; heteropolar; elliptical in polar view.

name: Corynocarpus laevigatus.

comment: Rarely recorded.

summary: O1, I, P1, D1.

Crassulaceae

genus: Crassula (12).

pollen: Grains small; ectoapertures very long, sometimes fusing to form a polar cap; endoapertures H-shaped.

name: Crassula.

comment: Rarely recorded.

summary: O1, I, P1, D1.

Cucurbitaceae

genus: Sicyos (1).

pollen: Grains 7–9 colpate; exine minutely reticulate; tectum spinulose.

name: Sicyos australis.

comment: Not recorded in pollen analyses.

summary: OX, I, P1, D1.

Cunoniaceae

genera: Ackama (2), Weinmannia (2).

pollen: Grains small, reticulate; in Ackama the reticulum can only be resolved under oil immersion; in Weinmannia the reticulum is easily resolved at 400× magnification.

names: Ackama; Weinmannia.

comment: Pollen recorded in South Island sites may be treated as Weinmannia racemosa.

summary: Ackama O1, I, P2, D1; Weinmannia O4, I (W), P3, D2.

Donatiaceae (Stylidiaceae)

genus: Donatia (1).

pollen: Grains of Donatia novae-zelandiae are mostly tricolporate in contrast to the colpate grains (tri-, tetra- or pentacolpate) of Forstera, Oreostylidium and Phyllachne sometimes associated with it, but here placed in Stylidiaceae. In Donatia the often obscure endoapertures are H-shaped.

name: Donatia novae-zelandiae.

summary: O1, I, P1, D1.

Droseraceae

genus: Drosera (6).

pollen: Grains spinulose tetrads; in Drosera arcturi the tetrad is a compact, 8–12-porate unit; in all other species, the units of the tetrad are distinguished by stalked proximal surfaces, essentially the germinating apparatus.

names: Drosera arcturi; Drosera spatulata type (all other species of Drosera).

comments: Drosera arcturi pollen is easily distinguished from other Drosera pollen. Spine length varies in pollen of the other five species, but is not reliable enough as a character to permit separation. Only occasionally recorded.

summary (general): O1, I, P1, D1.

Elaeocarpaceae

genera: Aristotelia (2), Elaeocarpus (2).

pollen: Grains tricolporate, small; exine smooth in Elaeocarpus, faintly and finely reticulate in Aristotelia.

names: Aristotelia; Elaeocarpus.

comments: Elaeocarpus grains are smaller than those of Aristotelia. Aristotelia grains recorded in glacial sites are probably from A. fruticosa.

summary: Aristotelia O1, I, P1, D2; Elaeocarpus O2, I, P1, D2.

Elatinaceae

genus: Elatine (1).

pollen: Grains are of two forms; in the first the tectum has clustered perforations in mesocolpia; in the second the exine is finely and faintly reticulate.

name: Elatine gratioloides.

comment: Not recorded in pollen analyses.

summary: OX, I, P1, D1.

Ericaceae

genera: Acrothamnus (2), Androstoma (2), Archeria (2), Cyathodes (2), Dracophyllum (∼ 35), Epacris (2), Gaultheria (12), Leptecophylla (2), Leucopogon (5), Pentachondra (1), Sprengelia (1).

pollen: Grains tetrahedral tetrads. In most taxa the four members of the tetrad are fully developed (T-type grains). Of these Archeria and Dracophyllum subgenus Oreothamnus grains have very short ectoapertures (5–10 µm long), in contrast to the long or very long ectoapertures (14–32 µm) of the others. The exine of these T-type grains is smooth (Archeria, Acrothamnus colensoi, Pentachondra pumila), verrucate–echinate (Epacris alpina), finely reticulate (E. pauciflora) or faintly rugulate (Gaultheria); Dracophyllum grains have an intectate area in the mesocolpia. Sprengelia incarnata is recognized by its thick perforate exine often broken into coarse exinous patches. Pollen of the two subgenera of Dracophyllum, Dracophyllum and Oreothamnus, can be distinguished; the former by longer ectoapertures (20 µm) and straighter sides, e.g. D. menziesii, the latter by shorter ectoapertures and obviously lobed appearance, e.g. D. muscoides. Pentachondra pumila can be distinguished from Acrothamnus colensoi by shorter, broader ectoapertures and by more numerous and more easily resolved endocracks; the ectoapertures of Leucopogon colensoi may be bifurcate.

The tetrads of the remaining taxa are aborted in varying degrees. The three aborted components of Leucopogon fraseri are so reduced that the pollen may be mistaken for monads (S-type grains); they are heavily patterned with endocracks. By contrast, the three aborted grains of L. fasciculatus, although small, are usually obvious as separate entities.

Pollen of Androstoma empetrifolia and Cyathodes pumila generally occur as dyads because two members of the tetrad are aborted; the aborted cells are usually small, and scrobiculi are scattered over the exine. In Leptecophylla juniperina and L. robusta the aborted grains are not much smaller than the fertile grain and the exine is smooth.

names: Archeria; Cyathodes (Acrothamnus colensoi, A. suaveolens, Androstoma empetrifolia, Cyathodes dealbata, C. pumila); Dracophyllum or Dracophyllum and Oreothamnus; Epacris alpina; Epacris pauciflora; Leucopogon colensoi; Leucopogon fasciculatus; Leucopogon fraseri; Pentachondra pumila; Sprengelia incarnata.

comment: Mostly occasional, but some, e.g. Dracophyllum, may be frequent as in the Auckland, Campbell and Chatham islands (Mildenhall 1994; McGlone et al. 2000, 2007).

summary (general): O1, I, P1, D1.

Euphorbiaceae

genera: Euphorbia (1), Homalanthus (1), Poranthera (2).

pollen: Grains medium (Homalanthus, Poranthera) to large (Euphorbia); ectoapertures very long, fosse-aperturate in Homalanthus; endoapertures obscure (Euphorbia), large rectangular (Poranthera), or long, narrow, tapering (Homalanthus).

names: Homalanthus polyandrous; Euphorbia glauca; Poranthera.

summary (general): OX, I, P1, D1.

Fabaceae

genera: Canavalia (1), Carmichaelia (22), Clianthus (2), Montigena (1), Sophora (8).

pollen: Grains mostly prolate–subprolate; tricolporate, endoapertures lalongate; exine variously reticulate. Exceptions are Canavalia rosea and Sophora. Canavalia rosea grains are anisopolar, with smooth and very thick exine often broken into exinous islands at one pole; bacula short. Endoapertures in Sophora are lolongate. Montigena grains are similar to those of Sophora, but endoapertures tend to be lalongate. Most Carmichaelia grains are finely reticulate and cannot be distinguished although there are differences in size. However, C. carmichaeliae, C. crassicaulis, C. glabrescens, C. muritai, C. stevensonii, and C. torulosa are generally more boldly reticulate, especially C. stevensonii. Clianthus grains are generally larger than those of the other taxa.

names: Canavalia rosea; Carmichaelia (Carmichaelia), all species; Clianthus (Clianthus maximus, C. puniceus); Sophora type (Sophora, Montigena).

comments: The only recorded pollen types are those of Carmichaelia and Sophora. The more boldly reticulate grains of Carmichaelia have not been distinguished.

summary: Carmichaelia O1, I, P2, D1; Sophora type O1, V, P2, D1.

Fagaceae

genus: Nothofagus (5).

pollen: Grains easily recognized. Nothofagus menziesii pollen with ∼ 7 short apertures is stephanocolpate, angulaperturate, apertures not always easily resolved; the exine is very thin, and bears very short, more or less evenly distributed spinules; shape is peroblate. Grains of all other species are more robust, and smaller; apertures (5–8) are clearly defined, short, and thickened around the margins; exine up to 1.5 µm thick, bearing spinules; shape oblate, more or less circular in polar view. An analysis of aperture distribution (Harris 1956a, b; Newnham 1992; Hanks & Fairbrothers 1976) shows > 80% of all grains have either 6 or 7 apertures. N. fusca has approximately equal frequencies of both 6 or 7, whereas in the other taxa, ∼ 61–66% have 7 apertures. This may be useful to confirm the presence of N. fusca.

names: Fuscospora (subgenus includes N. fusca, N. solandri, N. cliffortioides, N. truncata); Nothofagus menziesii.

comments: Grains of the Fuscospora group (note: often mistakenly spelt Fuscaspora) are often over-represented, and some are found in almost all sites, including the subantarctic islands; Nothofagus menziesii grains are generally under-represented, but they may be well represented if source plants are locally dominant.

summary: Fuscospora O4, W, P3, D3; Nothofagus menziesii O4, W, P3, D2.

Gentianaceae

genera: Gentianella (30), Sebaea (1).

pollen: Grains easily recognized by their bold reticulate or striate–reticulate exine. Gentianella grains are larger (polar axis of G. bellidifolia 58–71 µm) than those of Sebaea (polar axis 25–29 µm).

names: Gentianella; Sebaea ovata.

comments: There is considerable variation in surface pattern (Moar 1993; Glenny 2004) and for this reason most records are best referred to Gentianella; in subantarctic records those from the Campbell Islands may be referred to G. antarctica, from the Antipodes Islands to G. antipoda, and from the Auckland Islands reference to G. cerina and G. concinna may be made. Sebaea ovata grains are rarely recorded.

summary (general): O1, I, P1, D1.

Geraniaceae

genera: Geranium (8), Pelargonium (1).

pollen: Grains tricolporate; ectoapertures in Geranium are very short (approximatley one-quarter length of the grain), contrasting with the very long ectoapertures of Pelargonium; exine in both very thick, semitectate; Geranium exine bears blunt bold processes; Pelargonium exine bears fine spinules.

names: Geranium; Pelargonium inodorum.

comment: Geranium is the only pollen so far recorded, and then only sparingly.

summary: Geranium O1, I, P1, D1.

Gesneriaceae

genus: Rhabdothamnus (1).

pollen: Grains tricolporate, angulaperturate; ectoapertures very long, endoapertures lolongate; exine finely reticulate; grains small.

name: Rhabdothamnus solandri.

comment: Not recorded in pollen analyses.

summary: OX, V, P1, D1.

Goodeniaceae

genera: Scaevola (1), Selliera (2).

pollen: Recognized by their coarse bacula which are often bifurcate and longer at the polar area. Scaevola is larger and bolder than Selliera.

names: Scaevola gracilis; Selliera radicans.

comment: Selliera is the only pollen type recorded, and then only occasionally.

summary: Selliera O1, I, P1, D1.

Griseliniaceae

genus: Griselinia (2).

pollen: Grains characterized by small lalongate, slit-like endoapertures and faintly and irregularly striate exine.

name: Griselinia.

summary: O2, I, P1, D1.

Gunneraceae

genus: Gunnera (5).

pollen: Grains tricolpate, markedly fossaperturate, finely reticulate.

name: Gunnera.

comment: Often recorded in late-glacial sediments.

summary: O1, I, P2, D1.

Haloragaceae

genera: Gonocarpus (4), Haloragis (1), Myriophyllum (4).

pollen: Three pollen types are recognized: (a) Grains tricolpate to pentacolpate; exine up to 2 µm thick; tectum generally smooth. Found in all species of Gonocarpus and in Haloragis erecta. (b) Grains tricolpate to pentacolpate; exine up to 5 µm thick; tectum surface roughened. Sometimes occurs in Gonocarpus aggregatus, Haloragis erecta and Myriophyllum robustum. (c) Grains triporate to hexaporate, pores generally protruding; exine 1–2 µm thick, surface smooth or verrucate. Characteristic of all species of Myriophyllum.

Myriophyllum grains are either triporate (M. propinquum, M. votschii) or mainly tetraporate to hexaporate (M. pedunculatum, M. triphyllum). Two pollen types are produced by M. robustum. The first is similar to those of M. pedunculatum and the second similar to type (b) noted above.

names: Gonocarpus type (all Gonocarpus spp. and Haloragis erecta); Haloragis type (pollen of Gonocarpus aggregatus, H. erecta and Myriophyllum robustum as described under type (b)); Myriophyllum type (all species of Myriophyllum).

comments: Good indicators of wet habitats. It has been suggested that the thick-walled grains noted under type (b) above may be unviable (Moar 1993).

summary (general): O1, W, P2, D1.

Hypericaceae

genus: Hypericum (4).

pollen: Grains of medium size; ectoapertures bordered by smooth margins; exine finely reticulate.

name: Hypericum.

summary: O1, I, P2, D1.

Ixerbaceae

genus: Ixerba (1).

pollen: Grains tetraporate or pentaporate, mostly the latter; angulaperturate; exine smooth, thinner along endoapertures to form a visible border.

name: Ixerba brexioides.

summary: O1, I, P1, D1.

Lamiaceae

genera: Mentha (1), Scutellaria (1).

pollen: Mentha grains usually hexacolpate; apertures very long, sometimes syncolpate; exine finely reticulate. Scutellaria pollen tricolpate; aperture membrane often ruptured in two or more places; exine reticulate.

names: Mentha cunninghamii; Scutellaria novae-zelandiae.

comments: Mentha cunninghamii is occasionally recorded; Scutellaria has not been recorded in pollen analyses.

summary: Mentha cunninghamii O1, I, P1, D1.

Lentibulariaceae

genus: Utricularia (3).

pollen: Grains tricolporate (Utricularia dichotoma) or pentacolporate (U. lateriflora); ectoapertures in tricolporate grains very long, shorter in pentacolporate grains; endoapertures small and slit-like.

names: Utricularia lateriflora; Utricularia dichotoma.

comments: Rarely, if ever recorded. Pollen of U. australis was not available for description.

summary (general): O1, I, P1, D1.

Linaceae

genus: Linum (1).

pollen: Grains large (70–75 µm); pericolpate (up to 12 apertures), very thick exine; baculi very short, of two sizes.

name: Linum monogynum.

comment: Rarely recorded.

summary: O1, W, P2, D1.

Loganiaceae

genera: Geniostoma (1), Mitrasacme (2).

pollen: Grains of Geniostoma triporate; exine ∼ 2 µm around pore; surface smooth in optical section. Mitrasacme tricolporate; exine finely reticulate. Grains of both taxa small.

names: Geniostoma rupestre; Mitrasacme.

comments: Geniostoma and Parsonsia (Apocynaceae) grains are similar and may be confused. The smaller isopolar grains of Geniostoma with their thick exine, ∼ 1 µm, separate them from those of Parsonsia. Geniostoma rupestre rarely recorded. Mitrasacme not recorded in pollen analyses.

summary (general): O1, I, P1, D1.

Loranthaceae

genera: Alepis (1), Ileostylis (1) Peraxilla (2), Trilepidea (1), Tupeia (1).

pollen: The small oblate or suboblate pollen of Ileostylis and Tupeia cannot be confused with the medium to large peroblate grains of Alepis, Peraxilla and Trilepidea.

The spiny exine of Tupeia separates it from Ileostylis of which the exine surface is faintly rugulate or smooth.

The peroblate grains of Alepis, Peraxilla, and Trilepidea are variously striate, markedly triangular with often concave sides, and with long, usually syncolpate ectoapertures.

names: Ileostylis micranthus; Peraxilla type (Peraxilla spp., Alepis, Trilepidea); Tupeia antarctica.

comments: Alepis, Peraxilla and Trilepidea pollen may be distinguished on the basis of the exine pattern, but variation is considerable often making this separation difficult. Peraxilla colensoi is probably most commonly encountered in South Island material; Trilepidea adamsii, now extinct, was restricted to northern districts of the North Island. Peraxilla type and Tupeia antarctica are the most commonly recorded pollen, but not in great numbers.

summary: Ileostylis micranthus O1, I, P1, D1; Peraxilla type O2, V, P2, D1; Tupeia antarctica O1, I, P1, D1.

Malvaceae

genera: Hibiscus (2), Hoheria (7), Plagianthus (2).

pollen: Grains medium (Hoheria, Plagianthus) to very large (Hibiscus); tricolporate–pentacolporate (Hoheria, mainly tetracolporate; Plagianthus, mainly pentacolporate) or periporate (Hibiscus); spinulose, spines ∼ 35 µm long in Hibiscus, up to 5 µm long in Hoheria, and rarely > 2 µm long in Plagianthus.

names: Hibiscus; Hoheria; Plagianthus.

comment: Hoheria angustifolia sometimes has narrow, needle-shaped spines, e.g. from Peel Forest, Canterbury (Moar 1993), and such grains could be referred to H. angustifolia in any discussion. Although spine length may vary considerably, it is possible to separate Hoheria and Plagianthus grains. Spines and ectexine are sometimes stripped from the grains; these are referable to Hoheria or Plagianthus on the basis of aperture number. Pollen of the northern Hibiscus cannot be confused with other pollen types.

summary: Hibiscus OX, I, P2, D1; Hoheria O2, I, P2, D2; Plagianthus O2, IW, P2, D2.

Meliaceae

genus: Dysoxylum (1).

pollen: Grains isopolar or subisopolar, usually tetracolporate; ectoapertures very short.

name: Dysoxylum spectabile.

comment: Rarely recorded. Pollen dispersal may be restricted because the inflorescence is cauliflorous.

summary: O1, I, P1, D1.

Menyanthaceae

genus: Liparophyllum (1).

pollen: Grains peroblate, triangular in polar view, tectum generally boldly rugulate or rugulo-striate; ectoapertures parasyncolpate.

name: Liparophyllum gunnii.

comment: The bold exine pattern clearly separates these grains from those of Myrtaceae.

summary: O1, I, P1, D1.

Moraceae

genus: Streblus (3).

pollen: Grains diporate; exine very thin, slightly thickened at pore margins.

name: Streblus.

summary: O1, W, P2, D2.

Myoporaceae (Scrophulariaceae)

genus: Myoporum (1).

pollen: Grains tricolpate; colpi characterized by up to two lalongate fissures; exine reticulate.

name: Myoporum.

comments: Myoporum laetum grains are larger (polar axis 35–40 µm) than those of M. debile (polar axis 26–30 µm), which apparently is now extinct (Webb et al. 1988); apertures are similar to those of Tecomanthe speciosa (Bignoniaceae). Grains of Myoporum are occasionally recorded.

summary: O1, I, P2, D1.

Myrsinaceae (Primulaceae)

genera: Elingamita (1), Myrsine (11).

pollen: Grains of Myrsine have short apertures, usually four, and a smooth exine. Elingamita johnsonii pollen is tricolporate and faintly reticulate.

names: Elingamita johnsonii; Myrsine.

comments: Myrsine pollen is often well represented in late-glacial sediments and in mainland sites may include either the prostrate M. nummularia or the shrub M. divaricata; in post-glacial sites any of the mainland tree or shrub taxa may be represented. In the subantarctic Auckland and Campbell islands the source taxon is M. divaricata; in Chatham Island sites the source taxa may be either M. chathamica or M. coxii. Although there are subtle differences, pollen of the different taxa cannot easily be distinguished. Pollen of Elingamita johnsonii has not been recorded in pollen analyses.

summary: Elingamita johnsonii OX, P1, D1; Myrsine O4, I/W?, P3, D3.

Myrtaceae

genera: Kunzea (2), Leptospermum (1), Lophomyrtus (2), Metrosideros (12), Neomyrtus (1), Syzygium (1).

pollen: Grains peroblate, tricolporate, usually parasyncolpate except for Lophomyrtus and Neomyrtus; angulaperturate; angles more or less rounded except in the acutely angled Syzygium; exine surface usually smooth or faintly marked, but clearly verrucate in Lophomyrtus and Neomyrtus; Leptospermum and Kunzea have the smallest grains at ∼ 15–20 µm in equatorial plane, all others ∼ 24 µm or greater. Metrosideros excelsa and M. robusta grains tend to be larger and more robust than those of other Metrosideros species.

names: Leptospermum type (Kunzea, Leptospermum); Metrosideros, Metrosideros robusta type; Neomyrtus type (Lophomyrtus, Neomyrtus); Syzygium maire.

comments: Because Kunzea was formerly included in Leptospermum, and the pollen grains of the two genera are indistinguishable, the latter name is chosen for the pollen type. Lophomyrtus pollen is sometimes more boldly verrucate than that of Neomyrtus, but because the distinction is often not clear, the pollen type is named after the more widely distributed Neomyrtus. Of the Metrosideros species represented in New Zealand, M. excelsa, M. robusta, and M. umbellata are most likely to be encountered during a pollen analysis. Occasionally a larger and more robust grain is identified as M. robusta type, especially in North Island sites. When southern sites are being studied, as in Westland, the probability that the pollen is derived from M. umbellata, especially if it is abundant, should be noted in the text. In material from the Auckland Islands grains could be identified as M. umbellata, although if grains are infrequent the possibility of long-distance transport, including Eucalyptus from Australia, must be considered.

summary: Leptospermum type (Kunzea, Leptospermum) O4, I, P3, D1; Metrosideros or Metrosideros robusta type (M. excelsa and M. robusta) O3, I (V), P3, D2; Neomyrtus type (Lophomyrtus, Neomyrtus) O1, I, P3, D1; Syzygium O1, I, P2 (3), D1.

Nyctaginaceae

genus: Pisonia (1).

pollen: Grains pericolpate, 12–15 aperturate; tectum often minutely spinulose, perforate.

name: Pisonia brunoniana.

comments: Grains not recorded in pollen analyses. Superficially similar to Montia but distinguished by much shorter and broader apertures.

summary: OX, I?, P1, D1.

Oleaceae

genus: Nestegis (4).

pollen: Grains tricolpate; apertures short, recessed, exine reticulate.

name: Nestegis.

comments: Grains regularly recorded in North Island samples; in the South Island mostly in interglacial samples. The grains have been confused with those of Brassicaceae and Callitriche.

summary: O2, I, P2, D2.

Onagraceae

genera: Epilobium (30), Fuchsia (3).

pollen: Grains with protruding compound apertures; thick spongy tectum; viscin threads often present. Epilobium pollen usually shed as loosely aggregated tetrads; individual grains tricolporate; Fuchsia always monads, dicolporate.

names: Epilobium; Fuchsia.

comment: The differences between pollen of species in both genera are too slight to allow greater precision of identification.

summary: Epilobium O1, I, P1, D1; Fuchsia O1, V, P1, D1.

Oxalidaceae

genus: Oxalis (4).

pollen: Grains tricolpate; finely reticulate; tectum in O. magellanica bears unevenly spaced spinules.

name: Oxalis.

comment: Not recorded in pollen analyses.

summary: OX, I, P1, D1.

Passifloraceae

genus: Passiflora (1).

pollen: Grains, large, hexacolpate, colpi in three pairs; exine finely reticulate.

name: Passiflora tetrandra.

summary: O1, V, P2, D1.

Pennantiaceae

genus: Pennantia (2).

pollen: Grains tricolporate, often fossaperturate in polar view; endoapertures short, slit-like, exine surface generally smooth, but sometimes roughened.

name: Pennantia.

comments: May be confused with Griselinia but distinguished by the smooth or roughened exine surface in contrast to the finely striate surface of Griselinia; rarely recorded.

summary: O1, I, P2, D1.

Pittosporaceae

genus: Pittosporum (22).

pollen: Grains tricolporate; ectoapertures very long; exine smooth; tectum with scattered perforations. Exine reticulate in P. dallii; uneven and undulate in P. pimeleoides.

name: Pittosporum.

comments: Not recorded in great numbers; neither P. dallii nor P. pimeleoides recorded in pollen analyses.

summary: O1, I, P1, D1.

Plantaginaceae

genus: Plantago (9).

pollen: Grains periporate, pores up to seven, small; tectum verrucate; shape spheroidal.

name: Plantago.

comment: Mostly recorded in glacial or late-glacial sediments.

summary: O1, W, P2, D2.

Polygonaceae

genera: Muehlenbeckia (5), Persicaria (1), Rumex (2).

pollen: Grains of the three genera are easily distinguished. Muehlenbeckia and Rumex grains are tricolporate, but the thicker exine of Muehlenbeckia (∼ 2 µm) and its striate–reticulate pattern separates it from the thinner, faintly reticulate grains of Rumex. Persicaria grains are periporate, the exine is thick and the pores deeply recessed.

names: Muehlenbeckia; Persicaria decipiens; Rumex.

summary: Muehlenbeckia O1, I, P3, D1; Persicaria decipiens O1, I, P1, D1; Rumex O1, W, P2, D1.

Portulacaceae

genera: Hectorella (1); Montia (8).

pollen: Hectorella grains tricolpate, colpi broad, very long; tectum bearing short scattered spinules and perforate. Montia grains pericolpate, colpi broad, 12–30 in number; tectum bearing widely spaced minute spinules, and perforate.

names: Hectorella caespitosa; Montia.

comment: Pollen grains of all species of Montia are similar and variable colpi numbers make separation difficult; usually recorded in subalpine or late-glacial sediments. There are similarities with Pisonia pollen, but confusion is unlikely because of the marked differences in distribution and plant communities. Hectorella pollen not recorded in pollen analyses.

summary: Hectorella caespitosa OX, I, P1, D1; Montia O1, I, P2, D1.

Primulaceae

genus: Samolus (1).

pollen: Grains small, finely reticulate, lumina larger in mesocolpia.

name: Samolus repens.

comment: Rarely recorded.

summary: O1, I, P1, D1.

Proteaceae

genera: Knightia (1), Toronia (1).

pollen: Grains triporate, angulaperturate; Knightia finely reticulate, Toronia exine perforate, with scattered gemmae.

names: Knightia excelsa; Toronia toru.

comment: Knightia is the most frequently recorded, although never in great numbers.

summary: Knightia excelsa O1, V, P2, D1; Toronia toru O1, V, P1, D1.

Quintiniaceae (Paracryphiaceae)

genus: Quintinia (3).

pollen: Grains small, tetracolporate or pentacolporate, mostly the latter, ectoapertures very long, endoapertures small, elliptical; exine very thin, smooth.

name: Quintinia.

comment: Generally only occasionally recorded, but sometimes abundant.

summary: O1, I, P1, D2.

Ranunculaceae

genera: Anemone (1), Ceratocephalus (1), Clematis (10), Myosurus (1), Psychrophila (2), Ranunculus (45).

pollen: Grains of Ceratocephalus, Clematis, Myosurus, Psychrophila and many species of Ranunculus, e.g. R. grahami, are tricolpate; other species of Ranunculus are pericolpate, e.g. R. nivicola. Anemone grains are tricolpate or hexacolpate. Exine characters are variously spinulose, perforate and, in the case of Ranunculus the surface is either slightly undulate, e.g. R. nivicola, or obviously undulate, e.g. R. carsei. For discussion on the differences in Ranunculus pollen see Moar (1993). The tricolpate grains of Ceratocephalus, Myosurus and Psychrophila bear very short, more or less evenly distributed spinules, and cannot easily be distinguished.

names: Psychrophila type (Ceratocephalus, Myosurus, Psychrophila); Ranunculus type (Anemone, Clematis, e.g. R. grahamii, R. biternatus); Ranunculus carsei type (e.g. R. carsei, R. cheesemanii); Ranunculus nivicola type (e.g. R. nivicola, R. insiginis).

comments: Variation within and between the genera make precise identification difficult and using the name Ranunculaceae would be acceptable for most situations. Pollen is mostly recorded in late-glacial sediment, but is never recorded in great numbers.

summary (general): O1, I, P2, D1.

Rhamnaceae

genera: Discaria (1), Pomaderris (7).

pollen: Grains of Discaria are generally tricolporate, angulaperturate; ectoapertures very long; endoapertures small, more or less circular; exine smooth or faintly verrucate or scabrate. Pomaderris kumeraho grains are similar, but ectoapertures are obviously costate and endoapertures are H-shaped. The pollen grains of other Pomaderris species are characterized by as many as eight, often distorted, ectoapertures.

names: Discaria toumatou; Pomaderris.

comments: Although it may be difficult to separate Pomaderris kumeraho from Discaria toumatou when viewed in polar view, in practice this is not a major difficulty because Pomaderris is mostly restricted to northern areas of the North Island. Pomaderris phylicifolia extends southwards into Canterbury, but only in scattered localities. Pomaderris pollen from Australia may occur in surface samples, especially on the western slopes of the Southern Alps. Both are generally only rarely recorded, but Discaria is sometimes frequent, e.g. Cowan & McGlone (1991).

summary (general): O1, I, P2, D1.

Rosaceae

genera: Acaena (18), Geum (7), Potentilla (1), Rubus (5).

pollen: Grains mostly tricolporate, but Geum pollen is tetracolporate or pentacolporate except for the tricolporate G. uniflorum. Acaena and Potentilla have a well-defined operculum over each ectoaperture, but are easily distinguished by differences in exine pattern – rugulate in Acaena, striate in Potentilla. Rubus pollen is striate, but apertures lack opercula, as do Geum pollen, which are often sub-rectangular or square in polar view and possess rugulate or spinulose exine.

comment: Grains most frequently recorded are those of Acaena and Rubus.

names: Acaena; Geum; Potentilla anserinoides; Rubus.

summary: Acaena O1, W–I?, P2, D1; Geum O1, I, P1, D1; Potentilla O1, I, P2, D1; Rubus O1, I, P3, D1.

Rousseaceae

genus: Carpodetus (1).

pollen: Grains tetrahedral tetrads; ectoapertures and endoapertures of about equal length; exine often faintly rugulose. Key character separating Carpodetus from other tetrad pollen is the presence of ectoapertures on the internal walls.

name: Carpodetus serratus.

comment: Often recorded in low numbers.

summary: O1, I, P1, D1.

Rubiaceae

genera: Coprosma (60), Galium (3), Nertera (8).

pollen: Grains of Galium are small, pentacolpate or hexacolpate, and minutely spinulose. Coprosma and Nertera grains are usually tricolporate, but some, especially Coprosma perpusilla, are tetracolporate; ectoapertures are short and narrow; endoapertures are broad and often long; endocracks are common, and in Nertera often numerous and anastomosing; in Nertera the exine is often gemmate.

names: Coprosma; Galium; Nertera.

comments: Sometimes the characteristic gemmae of Nertera pollen are lost; identification may be possible if endocracks are clearly visible. Tetracolporate grains, not often recorded, may reflect the presence of Coprosma perpusilla, especially in material from the subantarctic islands.

summary: Coprosma O3, W, P3, D3; Galium O1, I, P1, D1; Nertera O1, W, P3, D2.

Rutaceae

genera: Melicope (2), Phebalium (1).

pollen: Grains tricolporate, reticulate; Phebalium is considerably smaller than Melicope, has a finer reticulum and sometimes an H-shaped endoaperture is obvious.

names: Phebalium nudum; Melicope.

summary (general): O1, I, P2, D1.

Santalaceae

genera: Exocarpus (1), Mida (1).

pollen: Grains of Exocarpus are characterized by short ectoapertures and by the fusion of endoapertures to form a continuous equatorial ring. In contrast, Mida pollen is heteropolar, and ectoapertures are parasyncolpate at one pole, bifurcate at the other.

names: Exocarpus bidwillii; Mida salicifolia.

comments: Although the fused endoapertures of Exocarpus are similar to those of Stilbocarpa (Araliaceae) the smaller grains of Exocarpus (polar axis 21–23 µm), and their shorter ectoapertures, separate them from those of Stilbocarpa (polar axis 31–36 µm). Pollen not recorded in pollen analyses.

summary: Exocarpus bidwillii OX, I, P1, D1; Mida salicifolia OX, I, P1, D1.

Sapindaceae

genera: Alectryon (1), Dodonaea (1).

pollen: Grains of Alectryon are characterized by a finely striate–reticulate exine, in contrast to the scabrate exine of Dodonaea, which sometimes exhibits parallel folds in the mesocolpia, and slightly protruding lalongate endoapertures.

names: Alectryon excelsus; Dodonaea viscosa.

summary: Alectryon excelsus O1, I, P2, D1; Dodonaea viscosa O1, W, P3, D2.

Sapotaceae

genus: Pouteria (1).

pollen: Grains mainly rectangular; tetracolporate; flattened at pole and nearly square in polar view.

name: Pouteria costata.

summary: OX P2, P1, D1.

Scrophulariaceae

genera: All genera here are included in Scrophulariaceae, but alternative families as indicated by recent phylogenetic studies are also shown in parentheses. Chionohebe (5), Gratiola (3), Hebe (c. 80), Heliohebe (6), Leonohebe (5), Ourisia (13), Parahebe (17), (Plantaginaceae); Euphrasia (15) (Orobanchaceae); Jovellana (2) (Calceolariaceae); Glossostigma (2), Mazus (2), Mimulus (1) (Phrymaceae); Limosella (2) (Scrophulariaceae).

pollen: Grains mostly variously reticulate and colpate. Exceptions are Glossostigma, Gratiola and Jovellana. Glossostigma grains are minutely spinulose and perforate, mostly tetracolpate, and in the case of G. elatinoides the polar exine is markedly thickened. Gratiola and Jovellana pollen is colporate; Gratiola nana is usually tricolporate, but may be tetracolporate, in contrast to G. sexdentata which is usually tetracolporate, but may be tricolporate. Jovellana is mostly tricolporate.

Pollen grains of Hebe, Heliohebe, Chionohebe, Leonohebe, and Parahebe are very similar – all are finely reticulate or reticulate–striate in varying degrees, a characteristic that separates them from other genera in the family. These five genera are considered part of Veronica s.l. by other authors.

Euphrasia grains are larger, at the upper end of the medium size range, mostly tricolpate; colpus membrane often ruptured; tectum minutely reticulate. Mimulus and Limosella grains are small to medium in size, semitectate and reticulate.

Grains of Mazus and Ourisia are close, but the former are larger, up to 42 µm, in contrast to the 25–30 µm for Ourisia. Both are clearly reticulate with Mazus generally having the coarser reticulum.

name: Euphrasia; Glossostigma; Gratiola type (Gratiola, Jovellana); Hebe type (Chionohebe, Hebe, Heliohebe, Leonohebe, Parahebe); Limosella type (Limosella, Mimulus); Ourisia type (Mazus, Ourisia).

comments: Of these pollen grains those of Euphrasia and the Hebe type are the most frequently recorded; the others rarely so or not at all. Euphrasia is occasionally recorded in high numbers, e.g. Dickson (1972); those of the Hebe type always in such low numbers that more precise identification would be difficult.

summary: Euphrasia O1, I, P2, D1; Glossostigma, Gratiola type; Limosella type O1 I, P1, D1; Hebe type O2, I, P3, D2; Ourisia type OX, I, P2, D1.

Solanaceae

genus: Solanum (3).

pollen: Grains tricolporate (Solanum americanum, S. aviculare) or mostly tetracolporate (S. laciniatum); vestibulate.

name: Solanum.

comment: Grains not recorded in pollen analyses.

summary: OX, I, P2, D1.

Stylidiaceae

genera: Forstera (5), Oreostylidium (1), Phyllachne (3).

pollen: Pollen of Forstera, Oreostylidium and Phyllachne are spinulose; in all taxa the spinules are usually very short and best resolved by LO analysis (Moore et al. 1991); grains may be tricolpate as in Forstera sedifolia, tricolpate–pentacolpate as in Phyllachne species, or tetracolpate–pentacolpate as in Oreostylidium subulatum, which is distinguished by a perforate exine.

names: Phyllachne type (Phyllachne, Forstera); Oreostylidium subulatum.

summary (general): O1, I, P1, D1.

Tetrachondraceae

genus: Tetrachondra (1).

pollen: Grains tricolporate, small; tectum smooth, but rugulate in mesocolpium, perforate elsewhere.

name: Tetrachondra hamiltonii.

summary: OX, W? P1, D1.

Thymelaeaceae

genera: Kelleria (8), Pimelea (29).

pollen: Grains display croton pattern.

name: Pimelea type.

comments: There is some variation between taxa; numbers of grains recorded are few and usually only from sites reflecting cool conditions.

summary: O1, I, P2, D1.

Tiliaceae (Malvaceae)

genus: Entelea (1).

pollen: Surface exine either with supratectal reticulum or rugulate; tectum perforate.

name: Entelea arborescens.

summary: O1, I, P1, D1.

Urticaceae

genera: Australina (1), Boehmeria (1), Elatostema (1), Parietaria (1), Urtica (6).

pollen: Grains mostly triporate, but Elatostema grains are diporate, and Parietaria grains are mostly tetraporate.

name: Urtica type (all taxa).

comment: Urtica grains are most likely to be recorded, but only rarely.

summary: Urtica O1, W, P2, D1.

Verbenaceae

genera: Teucridium (1), Vitex (1).

pollen: Grains tricolpate; Teucridium parvifolium characterized by a finely verrucate tectum and well-developed operculum, which may be destroyed by processing; Vitex by a smooth, perforate, sometimes faintly striate tectum.

names: Teucridium parvifolium; Vitex lucens.

comments: Teucridium not recorded; cannot be confused with Rosaceae pollen.

summary: Teucridium parvifolium OX, W, P1, D1; Vitex lucens O1, W, P1, D1.

Violaceae

genera: Melicytus (10), Viola (3).

pollen: Grains generally similar except for the larger size of Viola, polar axis up to 30 µm long, in contrast to the more usual 21 µm for Melicytus. In both genera, the tectum appears smooth, or nearly so in Bright Field (BF), but may be finely reticulate (Melicytus ramiflorus), or rugulate; perforate (M. crassifolius) under scanning electron microscopy.

names: Melicytus; Viola.

comment: Melicytus is more frequently recorded than Viola, but neither is common.

summary (general): O1, I, P1, D1.

Viscaceae (Santalaceae)

genus: Korthalsella (3).

pollen: Grains small, exine in mesocolpium much thicker than polar exine.

name: Korthalsella.

summary: O1, I, P1, D1.

Acknowledgements

We are indebted to Peter Heenan, Landcare Research, for his care in checking taxonomic detail and for his comments in respect of these. We also thank Rewi Newnham for comments on the draft manuscript, two anonymous referees for their helpful comments and Christine Bezar, Landcare Research, for her careful scrutiny of the text. This research was supported by the New Zealand Foundation for Research, Science and Technology through the Ecosystem Resilience OBI.