Many texts give vague descriptions of the preferred growing conditions of NZ native orchids. These publications often use terms such as "prefers light regions" or "commonly found in dark conditions".  These are highly subjective descriptions, that are based more on anecdotal evidence than scientific measurement or investigation.  

My Ph.D. research investigated the light environments of selected terrestrial native orchids in New Zealand.  This study involved measuring the light environments of 480 plots distributed over 20 forest sites around the South and Stewart Islands. 

As part of this work I examined whether the preferred light environments of orchid species could be classified into simple groups (as used by various authors), or whether their preferences varied too widely to fit this simplistic model.  

A plant on the forest floor receives a combination of direct and diffuse light.  A plant receives direct light only when the sun is shining directly through an opening in the canopy (a sunfleck).  The amount of direct light energy reaching the forest floor will depend on the position of the sun (which varies both during the day and over the season), the position of canopy holes and prevailing weather conditions.  

The amount of diffuse light energy reaching the forest floor will depend on the overall canopy density as well as the source of obscuration (cloud vs canopy).   The overall light environment experienced by a plant is the combination of both direct and diffuse light energy (total energy received) as well as the relative proportions of each light type.  

Consequently, the forest floor light environment cannot be described by a single parameter.  
This makes it difficult to define simple light environments (e.g. "light" and "dark" conditions)
from forest measurements.

Reconnaissance was carried out at each site to locate orchid colonies.  Plots of 5m x 5m were laid out in various selected or random locations.  The forest canopy directly above each plot was photographed using a Nikor 8mm f/2.8 hemispherical lens which was tripod mounted and aligned with magnetic North.  

Any orchids found in this plot were identified and counted.  Images were analysed using a computer program (see thesis for details).  This information was combined with regional weather data to generate estimates of typical light environments during the growing season of each species.  

The analysis generated a total of 8 parameters per photograph, each characterising a specific aspect of the light environment.  Having multiple parameters per photograph made it difficult to group species into clusters preferring similar light environments.  Consequently, a statistical method called Multiple Discriminant Analysis (MDA) was used, which determined species clusters by examining patterns in all parameters simultaneously.  Analysis was performed for the 13 species which were most regularly found at the survey sites.

Results of the MDA analysis are shown the Figure.  Points clustered in close proximity represent species which were found under similar overall light environments.  Grey ellipses shown on the plot represent three distinct light environments, described here as high, medium and low.

Forest floor light environments can be described quantitatively, however they are sufficiently complex that they cannot be expressed completely using only one or two parameters.

However use of suitable statistical tools showed it was possible to classify the typical or preferred light environment of these species into three simple groups.  This is consistent with the classifications broadly used in literature.  Most significantly, the classifications were based on quantified light environments rather than traditional ad hoc estimates of light levels.

It is interesting to note that all four species in the low light grouping are Pterostylis. These species are not confined to low light conditions, but are considered to be well suited to this light environment, and able to out-compete other species in this niche.  Subsequent studies showed that the species P.banksii and P.graminea were able to survive in considerably higher light environments in the absence of competition.  This finding is consistent with occasional observations of Pterostylis colonies growing in surprisingly bright conditions.  

The "preferred" light regimens identified in this study (shown in the Figure) represent the realised niche for each species, i.e. conditions under which they are best able to outcompete other species.  Some of these orchid species may actually prefer an alternative light environment, but be unable consistently to maintain populations under these conditions due to overwhelming competition from other species.
 

References
Abernethy A.K. (2002); Light regimes as a control of terrestrial orchid distribution in
New Zealand.  Ph.D, University of Canterbury.