Interpreting the Pollen Record - Reconstructing the Deforestation of New Zealand
Creating pollen diagrams
Once the pollen slides have been counted, then pollen diagrams can be made, to express the results in a meaningful way. The pollen counts are converted to percentages of the total terrestrial pollen grains counted. Because sediments accumulate over time, and depth is a proxy for age, then the best way to assemble the pollen counts from each depth of a core is to plot them out as a long histogram, where the stratigraphically lowest and oldest samples are at the base, and the youngest, surface samples at the top. This way, the changes in vegetation cover at a site can be expressed over time.
We use radiocarbon dating to tell us how old the sediments are, and to pinpoint exactly when changes in the pollen diagram occurred. There are excellent detailed web pages explaining the radiocarbon dating technique by the Waikato and Oxford University radiocarbon dating laboratories /www.radiocarbondating.com. This site explains very clearly the importance of dating samples; how different materials such as wood, peat, soil, lake sediments, seeds etc. can be dated; how samples are analysed; and the many applications of radiocarbon dating. Volcanic ashes or tephras are also used in the central and northern North Island to help age peat profiles, and, although they can only provide limiting dates for events, they provide a valuable indication of the accuracy of associated radiocarbon dates in a profile.
Interpreting the pollen record
A pollen diagram can tell us about what the vegetation was like in the past. Two examples are shown here to illustrate how we can detect initial human impacts from the pollen and sediment records. Our first example is of a pollen record derived from a lake sediment core from Lake Rotonuiaha, near Putere, Hawke’s Bay, North Island (from Wilmshurst 1997 and Wilmshurst et al. 1997). The pollen diagram here only shows the most important species for the sake of simplicity, but commonly, an entire pollen diagram may have more than 100 species or plant groups represented.
Zone 1 (pre-deforestation): The pollen is dominated by tall podocarp forest trees (including rimu, mataī, kahikatea, tōtara), with a wide range of hardwood trees (including rewarewa Knightia excelsa, hīnau Elaeocarpus dentatus, titoki Alectryon excelsus, maire Nestegis) shrubs, climbers, and ferns and reveal that this area was once covered with a tall, dense podocarp/hardwood forest. At the base of the zone, there is a large, but short-lived and rapid decline of pollen from tall forest taxa, at the same time as a thick layer of volcanic ash occurs in the core, together with a massive increase in charcoal, bracken and many seral species. This is a significant event in the history of the central North Island; it clearly shows us the impact the last eruption from the Lake Taupo volcano had on forests, even as far away as Lake Rotonuiaha.
The newly created clearings in the damaged forests (from ash damage, fallen trees, fires etc.) were rapidly exploited by plants characteristic of disturbed forests – these can all be detected in the pollen record (e.g. rewarewa Knightia excelsa,putaputaweta Carpodetus serratus, wineberry Aristotelia, tutu Coriaria arborea, coprosmas, pate Schefflera digitata, titoki Alectryon excelsus, bracken Pteridium esculentum and the tree fern wheki Dicksonia squarrosa). Climbers and epiphytic perching plants were also quick to exploit light gaps, shown by small increases of pollen from the shrub and small-tree parasites – perching epiphytes and vines and climbers that scramble over damaged canopies such as kiekie (Freycinetia baueriana), bush lawyer (Rubus) and supplejack (Ripogonum scandens).
The Taupo Eruption was probably the most explosive volcanic eruption in the world in the last 5000 years, and is known to have occurred about ad 200. This signal of massive disturbance seen in the Rotonuiaha core can be replicated in pollen records from sediment cores all around the central and mid-eastern North Island, as its impact on the landscape in this region was devastating. What is interesting about the forest composition after this huge volcanic eruption is that the canopy had closed over again within about 200 years after the eruption. Recovery was relatively rapid, and complete, returning to its original composition, despite the magnitude of the eruption.
Zone 2 (early Māori period): There is a sudden decline of pollen from tall forest trees, with a corresponding increase of bracken spores. Charcoal and pollen from tutu, coprosma and wineberry (Aristotelia) are markedly abundant. This zone shows the effect of early Māori forest clearance in the area. This is a very typical pollen signature from cores taken from the drier parts of eastern New Zealand. How long did deforestation take? The dates for deforestation from pollen records throughout New Zealand indicate little difference between the timing of clearance at northern or southern sites, and show it proceeded rapidly. Although that clearance was faster at some sites (e.g. Kohika, Bay of Plenty) compared with others (e.g. Lake Taupo), the process was well advanced throughout New Zealand within about 200 years. The timing, extent, and speed of deforestation in New Zealand was related to rainfall and topography rather than location, i.e. forests in drought-prone, flat to rolling areas were destroyed more quickly and more extensively than those in wetter, hilly areas.
Even though the main species occurring after deforestation are similar to those occurring after the smaller preceding fires or the last Taupo eruption, there are subtle differences in the vegetation responses following deforestation, which can be detected in the detailed Rotonuiaha pollen profile (see Wilmshurst et al. 1997). Before deforestation, there was always complete and rapid forest recovery following the smaller fires that are recorded. However after deforestation, and with repeat burning, there is no sign of forest recovery. Initial Māori forest clearance was a one-way process in most cases, where tall forest was permanently replaced by seral vegetation. The increased number and size of fires that continued after deforestation allowed seral taxa – particularly bracken – to persist on the landscape and prevent any effective forest redevelopment. In addition, the scale of forest disturbance caused by deforestation is unmatched by any other preceding disturbance recorded in the cores, including the last Taupo eruption. For these reasons, the smaller fire disturbances preceding deforestation are considered most likely to have been caused by lightning strike.
After deforestation of the Rotonuiaha catchment, soil stability was maintained once a dense cover of bracken and scrubland had established. Bracken rapidly stabilised soils after deforestation and was as effective at minimising soil erosion as forest cover. Bracken rhizomes can penetrate the soil to half a metre or more in depth and form extensive, dense underground networks. The tangle of numerous rhizomes in the soil would have rapidly intermeshed the catchment soils after deforestation, improving the strength and cohesion of the soils. Since rhizomes are not destroyed by fire, soil stability would have been maintained even after frequent fires. The relatively tall, dense canopy created by extensive bracken colonies (up to 3–4 m high) would also have protected soils from slope wash, by shielding the ground from rain-drop impact, particularly during high intensity rain storms that are common in Hawke's Bay. In addition to the soil protection afforded by rapidly establishing seral vegetation, the intact root systems of burnt forest trees would have remained in the soil and continued to provide significant soil cohesion. Relatively low rates of sedimentation after Māori deforestation are also recorded in other North Island sites where bracken was the dominant vegetation, e.g. Firth of Thames; Whangapoua and Whitianga harbours on the Coromandel Peninsula; Lake Poukawa, Hawke's Bay; Holden's Bay, Lake Rotorua and Kohika, Bay of Plenty.
Zones 3 and 4 are the Early and Late-European settlement period: As in most pollen diagrams the European horizon is clearly identified by the first appearance of pollen from introduced plants. The base of this zone is given an estimated date of c. ad 1870, when European settlement and farming began in the area. There are increases in the pollen of grasses and dandelions, and the appearance of the first exotics including thistle, pine, plantain, sorrel, willow, and clover. Bracken spores remain at high percentages and continue to dominate the pollen sum, but in fact, this curve overestimates the amount of bracken fern on the landscape. Most of the bracken spores in this zone are old ones, which have been washed into the lake from inwashed eroded soils during high-intensity rain storms. Once forest and fern/shrubland had been destroyed and replaced with pasture grasses, the soils in this hilly country of Hawke’s Bay became highly vulnerable to soil erosion. During high-intensity rain storms, water just transports these soils (and any old pollen they may contain) into the lake, where they settle out and become a new layer of lake sediment.
Zone 4: (Post ad 1940s) Pollen from native forest taxa is at its lowest percentage in the profile, except for beech pollen which remains at c. 5–10% throughout zones 4 and 3. This beech pollen is likely to be long-distance transported from higher-elevation forests on the inland ranges that were not burnt. The pollen in this zone shows the transition to a farmed landscape.
Travis Swamp
Our second example of a pollen diagram showing initial human impacts on the vegetation is from a swamp core in Travis Swamp, Christchurch. A 50-cm peat core was extracted from the swamp. This was sampled for pollen every 2.5 cm down the length of the core. The pollen diagram has been divided up into four zones, which represent different periods of vegetation.
In the lowest zone 1, labelled ’pre-deforestation’ , the dominant pollen types are from forest trees mataī, tōtara, kahikatea, beech and hīnau (among many other species). Even though beech tends to produce vast amounts of pollen and it is often over-representative of the actual abundance of these trees on the landscape, levels of 30% indicate that beech was present in the local area. There were also many shrubby plants of which mnuka, myrsine and coprosma were the most common.
In zone 2`, labelled ’Māori’, there is a sudden decline in the tall forest and shrubby species, at the same time as a corresponding increase of charcoal, bracken, tutu and some grasses. This is the deforestation signal that we can replicate in many sediment cores taken from bogs or lakes from throughout the driest parts of New Zealand, from the south of the South Island right up to Northland. The bottom of this layer at Travis Swamp has been radiocarbon dated, allowing us to pinpoint the timing of this forest clearance to about ad 1270–1300. The same date for this event has been established now for over 30 pollen diagrams.
In zone 3,labelled ’early European’, the first record of pollen from introduced plants appears – e.g. dandelion, plantain, pine, clover, and willow, which are all indicative of early European settlement and farming. In the last zone, 4 ’late European’, at the surface layers of the core, forest is reduced markedly, bracken declines, and grasses and European plants are dominant. This reflects the change from a fern and tussock landscape to one dominated by agriculture and pine plantations. The persistent level of beech pollen at 5%, even at the surface, represents wind-blown pollen from beech forests growing on the Canterbury foothills.
