Do we see trees or forests?: Knowledge gaps

It would seem reasonable to assume that research aimed to fill a basic knowledge gap should receive more attention than projects aimed at evaluating complex untestable theory. However in the field of tropical forest ecology it not always clear that fundamental basic knowledge gaps are either being identified or filled quickly enough. An issue that has concerned me for many years is the persistent difficulty in accurately tracing and understanding the spatial distribution of tropical organisms, particularly trees and other plants.

Trees are immobile, long lived, highly visible elements of the tropical landscape. It would therefore seem reasonable to assume that we already know a great deal regarding their distribution. However trees form forests. Forests cause numerous logistical difficulties for field research.  Tropical forests can be hostile, disorienting, pathless environments. They also contain a complex set of natural resources. Illegal exploitation of some of these resources such as valuable timber or plants with interesting pharmaceutical properties (of various types) may obstruct scientific research. Many tropical forests have also provided refuge at times for members of radical political movements.

Tropical trees can be extremely difficult to identify to species. Flowers, fruits and leaves may be out of reach. Their taxonomy is complex and dynamic. Tropical families are unfamiliar to researchers trained in temperate zones. Many species have converged in terms of leaf shape and characteristics.

This has led to chronic ignorance of the distribution of tropical tree species. Only a small fraction of the total area covered by tropical forest of any description has been studied in any detail. Regional species lists rely on data from a few well known areas, but within any given region distribution patterns are poorly known. Many assumptions are made concerning the nature of anthropogenic influence on areas of forest. Some are justifiable, some tend to exaggerate human impact on naturally resilient forest ecosystems.

Given the poor state of our knowledge it is vitally important that the few reliable data sources that are available are used to their full capacity. However even the best global data sources can only provide a small part of the picture alone. Missouri Botanical Gardens (MOBOT) has one of the largest collections of tropical plant specimens of any herbarium. MOBOT has provided online access to most of this data. Yet even when the whole set of tropical tree collections with known coordinates from this source are mapped on to space large gaps are revealed in the empirically based knowledge they provide regarding tree diversity.

A PostGIS query can be used together with the graticule trick mentioned in the last post to map the density of collections within 0.1 degree (approximately 100 km2) grid squares in the MOBOT Vast data base. The query is provided as a model for counting incidences within grid squares. The data itself is stored locally and forms part of a larger set that we have collated over a period of several years.
CREATE TABLE meso.count with oids as
SELECT g.ncollects as Ncollects, a.*  from

grat01 a,

(SELECT d.gid, count( as ncollects
FROM meso.trees c, grat01 d
WHERE c.the_geom && d.the_geom
AND contains(d.the_geom, c.the_geom)
AND c.source like ‘MOBOT%’
group by d.gid) g

WHERE a.gid=g.gid;

If it is assumed that 100 collections per 100 km2 is a minimum needed to provide a reasonable picture of the distribution of tree species (there may be more than 400 species in each grid square in some regions) the map below shows that only a fraction of the area  of mesoamerica can be considered to be well enough typified by this data set to allow distribution patterns to be mapped through the data alone. This contrasts sharply with the floristic atlases available for most of Western Europe based on known occurrences.

Even when the grid squares with less than 50 collections (far fewer than needed) are added to the map there are many completely unknown areas when this data source is taken alone. Where collections have been taken the density is generally below 5 per 100 km2. Direct empirical evidence regarding tree diversity hot spots is simply not available. Comparisons  between Chiapas, Yucatan, Belize and Costa Rica cannot be made from the data alone.

Light green 1-2, dark green 2-5, blue 5- 50, red >50

Light green 1-2, dark green 2-5, blue 5- 50, red >50

In contrast to our poor knowledge regarding forest composition, our knowledge regarding the extent of both forests and deforestation is becoming ever more accurate and detailed as high resolution satelite imagery becomes available. It appears that we are looking at forests, but not seeing the trees.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s