Otari-Wilton's Bush (41°14′ S, 174°45′ E) is located just within Wellington city limits at the southern tip of the North Island of New Zealand, and encompasses approximately 100 hectares of native forest. The reserve is situated 70-280 meters above sea level and the soil is comprised of stoney colluvium of greywacke parent material. Average annual rainfall totals 1,240 millimetres and average daily temperatures range from 20°C in summer to 7°C in winter (Otari Native Botanic Garden and Wilton's Bush Reserve 2007). The vegetation is classified as coastal conifer-broadleaved forest, whose vertical structure is highly complex and similar to most tropical forests (Dawson 1988). It has a fairly continuous canopy, which is frequently interrupted by canopy gaps and canopy emergent tree species. A dense community of shrubs and tree ferns occurs beneath the canopy (Blick et al. 2008). Lianas and epiphytes are also abundant (Burns & Dawson 2005). Dysoxylum spectabile is the dominant canopy-forming species, alongside Melicytus ramiflorus, Corynocarpus laevigatus and Eleaocarpus dentatus. Macropiper excelsum and Geniostoma rupestre are the most common subcanopy shrubs. Emergent trees include Dacrydium cupressinum, Beilschmedia tawa and Knightia excelsa. Burns (2007) gives a detailed inventory of the woody plant community.
Data collection I - Plots
Thirty 30m x 30m plots were surveyed within the reserve (following Marjot1992). The plots encompassed a whole range of environmental conditions which were decided upon before data collection started as it was desired to cover plots facing different aspects (N, S, E and W). Once the locations were chosen plots were randomly selected within the location. A random number was generated using a calculator, if the number generated was even, plot would be placed on the right side of the track, if odd, plot would be placed on the left side of the track. After that, another random number was generated in order to decide how many steps would be walked in the chosen direction, were the centre of the plot would be placed. A compass was also used to assure that I walked in a straight line from the track to the desired location of the plot in order to avoid sampling bias. However, some restrictions were applied regarding the location of the plots, mainly due to safety reasons. Therefore, locations with slope >40 degrees were avoided of sampling.
In each plot abundance of sexually mature individuals was recorded. Sexually mature individuals are defined in this study as those capable of reproducing. Clues such as presence of fruit and/or flowers or evidence of such were used in deciding whether or not to include an individual in the research.
The majority (24) of the plots at Otari's were once farm land which was acquired by the Wellington City Council over 100 years ago (___________) in order to extend the reserve. Therefore, these plots have been undergoing secondary succession for over 100 years. The remaining six plots were located in old growth forest, which has never been cleared and are part of the original Wilton's Bush.
Data collection II – Individual species
Once the abundance of mature individuals was recorded, the second step was to collected information on individual plants. Plant species present in 5 or more of the plots were re sampled, only those found in at least 5 of the plots were used in order to provide more robust comparisons. Emergent species (Beilschmedia tawa and Knightia excelsa) and Podocarps were also not surveyed due to restriction in collection of leaves from these individuals.
In each plot, the tallest two individuals of each species had their height measured. Height was measured with the use of a hypsometer (Nikon Forestry 550), and was taken from the base to the uppermost branch of the tree. Once height was measured, each individual had six leaves collected, except Dysoxilum spectabile which has large compound leaves, and for that species leaflets were collected since they are likely to be functionally equivalent to leaves (Bongers and Popma 1990). Only leaves fully exposed to sunlight from the outer part of the branch were collected for trees, for shrubs, leaves on the top of the plant were collected as those leaves are not exposed to direct sunlight. Leaves were collected with the use of a scissor attached to a pole, placed into a paper bag, and taken to the lab for further measurements.
Data collection III – physical aspects of plots
Information physical aspect of each plot was also collected. Information on plot aspect was collected using a compass; slope was collected using a hypsometer (Nikon Forestry 550). A soil tester with 30 cm probes was used to measure ph and soil moisture in the upper layer (< 30cm) of the forest floor, measurement on soil moisture was given on a scale between 1 and 10, with 1 being the driest and 10 the wettest. Ph was given on a scale between 3.5 (acidic) to 8 (alkaline). Moisture and ph were measured
Sunlight was measured using a light meter and given in lux. Measurement was given on a scale between 0 (dark) to 2000 (light). Data on sunlight was collected on a sunny day of clear sky and each plot was surveyed twice on the same day (one measurement in the morning and another in the afternoon). Sunlight was collected at a height of 1.5m above the floor, therefore, reflecting the amount of sun received by shrub species. Sunlight was measured at the corners and centre of each plot and an average was taken for each plot. Averages for morning and afternoon for each plot were calculated for statistical analysis.
Each measurement (sunlight, moisture and ph) was taken at 5 different areas of the plot (4 corners and centre) and average value was then used in each plot for statistical analysis. Slope was taken from one end to another end of the plot in the direction the plot was facing, therefore, slope was measured in an area of 30 m.
Data collection IV – Lab
The fresh collected from individual plants had their area in cm2 measured using _______________, area meter scanner. Fresh leaf area was recorded to an accuracy of 2 decimal points, and once measurements on area were made, leaves were dried for 48 hours at a temperature of 65o (following Westorby et al.) and dried weight was taken using an electronic scale. Weight was measured in mg to an accuracy of 3 decimal points. With those measurements, leaf mass per area (LMA) in g/cm2, which is the inverse of SLA, was calculated and an average was given for each individual plant, based on the six collected leaves. LMA was used in this research as it holds vast amount of information of ecological information and is positively correlated with leaf lifespan, and negatively correlated with mass-based nitrogen content, photosynthetic capacity, transpiration rate and respiration rate (Husk et al. 2008 tree), therefore a lot can be said about a plant strategy based on its LMA.
35 different species belonging to 33 different genera. Out of 35 species 13 were found in 5 or more of the plots. Tawa and rewawera too tall to collect leaves. Podocarps (rimu and miro not sampled) therefore 9 species were used for further data collection and analysis. For each of the nine specie, two plants had their height measured and in each plant six leaves were collected for further analyses.. Total number of plant with height measured was 288. In each of the 288 plants, six leaves were collected resulting in a total of 1728 leaves.
Data was analysed using PASW statistics 18. LMA diversity among species was investigated by obtaining the PCA1 value of the environmental condition. Environmental conditions used to create the PCA1 axis were average sunlight, moisture, slope and soil Ph.
Factorial analyses in order to reduce the components into on axis PCA seeks a linear combination of variables such that the maximum variance is extracted from the variables. It then removes this variance and seeks a second linear combination which explains the maximum proportion of the remaining variance, and so on. This is called the principal axis method and results in orthogonal (uncorrelated) factors.
The main applications of factor analytic techniques are: (1) to reduce the number of variables and (2) to detect structure in the relationships between variables, that is to classify variables. Therefore, factor analysis is applied as a data reduction or structure detection method (the term factor analysis was first introduced by Thurstone, 1931).
By using PCA analysis it was possible to reduce four measured environmental conditions (light, slope, aspect and pH) into one axis.
PCA1 explained over 50% of the variance in environmental conditions. PCA1 was positively correlated with slope (.608) and sunlight (.807) and negatively correlated with moisture (-.970) and ph (-.314) Therefore higher PCA1 values describe plots which are steeper, lighter, drier and slightly more acidic than plots with low PCA value.
For each of the surveyed species, LMA was plotted against PCA1.
The two graphs above display the same data in different ways, as we move from north to south (1 to 7) PCA1 decreases, which was expected as north facing plots tend to have more sunlight and be drier than south facing plots.
Correlation between PCA1 and LMA.
Correlation between PCA1 and height
PCA 1 is positively correlated with leaf LMA and negatively or slight positively (only for Karaka and Rangiora) correlated with plant height for the 9 surveyed species.
It is also expected as in many cases, LMA is inversely correlated with plant height (reference)