Page 1 of 1

Tree diversity comparison over the northern temperate world

PostPosted: Fri Dec 26, 2014 10:58 am
by KoutaR
It is well known that of the temperate forest regions of the Northern Hemisphere, East Asia has the richest, eastern North America (NA) the second richest and Europe the poorest tree flora. Among the comparisons which have been made is a good one by Latham and Ricklefs (1993: Continental Comparisons of Temperate-Zone Tree Species Diversity. In Species Diversity in Ecological Communities: Historical and Geographical Perspectives. University of Chicago Press. Available online). They found 124 tree species in temperate Europe including the Caucasus, 729 in East Asia, 253 in eastern NA and 68 on the Pacific slope of NA. They omitted northern species with pan-continental distribution and southern species with more than 50% of their ranges extending into subtropical or Mediterranean areas. However, they define the regions only vaguely as “temperate” having used the maps of “Vegetation of the Earth” by Walter (1979) for delimitation. It can be seen from the maps that the study areas in eastern NA and East Asia extend to warmer regions than in Europe and western NA. Actually, the “warm temperate” region is almost missing from Europe because the climate south from the Alps is mediterranean (not mesic) with long dry summers. The southern North Carolina coast already has a higher mean annual temperature than any place on the southern coasts of Portugal or Greece, which are already fully mediterranean. Thus, comparing these whole regions is not necessarily “fair”. In addition, Latham & Ricklefs’ eastern NA is 1.5 times larger and their “Pacific slope” much smaller than the other regions. What would the results be if equally large and climatically comparable sub-regions were selected from each region? My hypothesis was that the differences between equivalent European and eastern North American sub-regions would be smaller than the differences between whole regions. Another question I was interested in: How would the Balkan Peninsula and the Caucasus rank in the comparison? These regions in western Eurasia are known to have richer tree flora than western and Central Europe, with many regional endemics. Could they even match the species diversity of eastern NA? Not finding answers to these questions, I decided to make my own study.

Selection of genera

I took the genera mostly from Latham & Ricklefs, but because they do not include mainly subtropical species I added genera from other lists and books. I may have missed some genera but they should only make a small portion of all the genera. It would have been too laborious to check whether each individual species is a tree or a shrub. The tree definitions and even the criteria for defining the maximum height may also differ between countries, so grouping species as either trees or shrubs could potentially include a systematic error. Therefore I decided to include those genera comprised mainly of trees and omit those genera comprised mainly of shrubs (or climbers or herbs, e.g. Salix) and then count ALL THE SPECIES in the selected genera. My tree definition was either 5 m in height or just defined to be “tree”. I mostly used the Flora of China descriptions and Wikipedia to find if a given genus includes mostly trees. I may have erroneously included shrub genera or omitted tree genera but as the errors should be distributed evenly between the regions due to the objective selection method, they should not affect the comparison. I excluded Sorbus, Crataegus and Malus because they have a lot of apomictic microspecies, the counting of which would not fit the purpose of my study.

Delimiting the study regions

My intention was to delimit climatically comparable sub-regions of similar size, one from each region: Western Europe, the Balkans, the Caucasus, East Asia, eastern NA and western NA. These sub-regions should extend to the west or east from a sea/ocean: east from the Atlantic Ocean, east from the Adriatic Sea, east from the Black Sea, west from the Pacific Ocean, west from the Atlantic Ocean and east from the Pacific Ocean, respectively. Each sub-region should contain one or more countries, states or provinces to ensure availability of species distribution information. Climate data is from At the eastern end of the Black Sea, Batumi (Georgia) is the only point with climate data, which compelled its choice as the reference point. I made the assumption that mean annual temperature is the single climatic parameter best correlating with species richness and calculated it as the mean of the mean daily maximum and minimum temperatures for each month (the mean of 24 values). Thus calculated, Batumi’s mean annual temperature is 13.5°C. Then I searched for points on all the mentioned coastlines with mean annual temperature closest to 13.5°C. I tried to eliminate the influence of altitude by starting from the coastlines. Then from each established point I drew a straight line inland along the same latitude, including all the countries, states or provinces along that line until their combined area was more than half of the area of the largest country, state or province (France). As a result of the differences in other climatic parameters (temperature range, precipitation etc) and landforms, there are differences in vegetation between the regions. All the sub-regions contain a number of different vegetation types and below I describe them briefly (even though most readers know at least two of them much better than I do).

Locations of the Eurasian sub-regions (in red). The 13.5°C points are marked in black. Note the northern positions of the European points due to the effect of the Gulf Stream.

Western Europe: France

Along the Atlantic coast of Europe, Biarritz in southwestern France was the selected point, thus France was the natural choice as the sub-region. France is also the largest of the sub-regions, its area without Corsica being 543,015 km2.

The potential natural vegetation in most of France is temperate deciduous forest, the climate having a strong oceanic influence with small temperature amplitude and abundant rainfall throughout the year. Almost the whole of France is located north of Biarritz but this is compensated by the fact that France also has another coastline with the Mediterranean Sea, with a partly different set of species and a different climate with hot dry summers. There are also two high mountain ranges, the Alps and the Pyrenees.

European beech (Fagus sylvatica) and European silver fir (Abies alba) in the Pyrénées National Park, about 100 km from Biarritz. Elevation approx. 1300–1400 m.

Species distribution data is mainly from the Euro+Med PlantBase (, which is still incomplete, so the GRIN database ( was used for the few missing families. The Euro+Med also has some clear errors; I corrected those, I noticed, and made notes on them in the results table (see below).


The 13.5°C point was Bar in Montenegro. As I advanced to the east, after Montenegro I encountered Albania, Serbia, Macedonia and Bulgaria, and reached the Black Sea coast but the combined area was still not large enough, so I added the country closest to the selection line, Bosnia and Herzegovina. The area of the sub-region is 318,825 km2.

Marsh in the Lake Skadar National Park, about 20 km from Bar. In the background white willow (Salix alba) forest strip along a creek. Elevation 6 m.

Generally the climate is more continental than in France. The distribution of rainfall is mediterranean with two maxima in spring and fall. Much of the region is mountainous, with deciduous and conifer forests. The Adriatic coast is mediterranean and there has also been some natural steppe in Bulgaria (now perhaps under cultivation?).

Distribution data source as for France.


The sub-region is composed of Georgia, Azerbaijan and Russian North Caucasus. The latter, in the Euro+Med PlantBase, is composed of the North Caucasian Federal District, Krasnodar Krai and the Republic of Adygea. The area of the sub-region is 410,600 km2.

The climate and vegetation vary greatly. Much of the region is mountainous with deciduous and conifer forests. The natural vegetation north of the Greater Caucasus range would be steppe, like the North American prairies, now mostly under cultivation. The eastern lowlands are dry, even semi-desert. In the southeastern corner of Azerbaijan adjoining the Caspian Sea, and along the Black Sea coast, there are humid regions, also called temperate rainforests with evenly distributed rainfall; the climate is wetter but the temperature amplitude slightly larger than along the Atlantic Coast of Europe.

Mtirala National Park, about 20 km from Batumi. The most abundant trees are oriental beech (Fagus orientalis), sweet chestnut (Castanea sativa), European hornbeam (Carpinus betulus) and black alder (Alnus glutinosa). The valley bottom is at about 300 m.

Distribution data source as for France.

East China

In my source for the climate data, there is no point close to 13.5°C along the coast, the closest being Shanghai (a city province south of Jiangsu Province, 15.9 °C) and Qingdao (Shandong Province, north of Jiangsu, 12.8°C). I decided to choose the border between Jiangsu and Shandong at the coast as my starting point. In addition to these two provinces, Henan was selected. The combined area is 426,700 km2. Species distribution data is from Flora of China ( In China, species delimitation is sometimes narrower than in the west. I tried to eliminate this difference by following the species concepts of the GRIN database in those cases where it differs from that of Flora of China. However, this removed only few species; it is possible that in some cases GRIN simply follows Flora of China.

Most of the sub-region is lowland; it would probably be covered by temperate deciduous forest by nature but the forest has mostly been cleared for cultivation. Western Henan has mountains more than 2000 m in height. The climate is more continental than in the western Eurasian sub-regions.

Locations of the North American sub-regions.

Eastern US

As for East China, there was no point close to 13.5°C along the Atlantic Coast, the closest being Ocean City (Maryland, 13.1°C) and Hampton (Virginia, 14.8°C). I chose the border between Maryland and Virginia at the Atlantic Coast as my starting point. In addition to Maryland and Virginia, West Virginia and Kentucky were selected, giving a combined area of 310,333 km2.

The climate and vegetation resemble those of the Chinese sub-region: an only slightly oceanic “east coast” climate, lowlands in the east, mountains up to about 1700 m; but much more is currently forested than in the Chinese counterpart.

Species distribution data is from the GRIN database.

Western NA: California

Along the Pacific Coast, the 13.5°C point was Monterey, so California was the natural choice as the sub-region. Its area is 423,970 km2. California extends to a warmer climate than the other sub-regions but this is compensated by the fact that southern California is very dry with low tree species richness.

Monterey pine (Pinus radiata) forest in Point Lobos State Reserve right on the coast, about 10 km from Monterey.

The climate along the coast is very oceanic like along the Atlantic coast of Europe but here mountains prevent moisture reaching far inland, where low elevations are consequently dry. The vegetation and landforms vary greatly and include humid giant forests in the northwest, drier forest types, high mountains and deserts. A distinctive feature is a summer dry season, which is a main reason for most forest types being dominated by conifers.

Species distribution data is from the GRIN database.


Species numbers by genera in 6 sub-regions are listed in the Excel file below.

Edit: This is a new version. One species of Acaciella, one Senegalia and one Vachellia added to California.
(34.5 KiB) Downloaded 49 times

The main results are summarized below.

Edit: New version. Changes to the original as above.

Again, note that the figures are not total tree species numbers but include shrubs and even lianas and herbs if they are in a genus mostly composed of trees (e.g. Cornus canadensis), and trees are omitted if they are in a genus mostly composed of non-trees (e.g. Salix alba).

Species richness is greater in the Balkans and the Caucasus than in France but the difference is not big, and all three are still well behind the eastern US. The Balkans exceeds the Caucasus due to the low conifer diversity of the latter. The reputation of the Caucasus as a center of endemism is actually based on relatively few taxa, notably Diospyros, Albizia, Gleditsia, Pterocarya and Zelkova. The same trend continues further east: northern Iran has more angiosperms than the Caucasus but only four conifers.

I did not find much support for my hypothesis that if climatically similar sub-regions are compared European tree diversity would be closer to that of eastern NA than if the whole regions are compared. If the mean of the species numbers of France, the Balkans and the Caucasus is used, the species ratio of eastern NA versus Europe is 1.9 as compared to 2.0 of Latham & Ricklefs. (Remember also that in the latter study, the eastern NA is larger, whereas in my study it is smaller than the European sub-regions.) By contrast, the East Asia versus eastern NA ratio is much lower in my study: 2.1 as compared to 2.9 of Latham & Ricklefs. I hypothesize that the reason could be that it is particularly East Asia where species richness quickly increases when going further south as species with tropical affinities join the flora: it is the only region in the Northern Hemisphere where mesic forest continues uninterrupted from the tropical to the temperate (and further to the boreal) latitudes. I made a small test comparing the species numbers in Fagaceae, a family of major importance in the northern temperate zone. In the table below, E US and E China are as above, SE US is Florida and Georgia, and SE China is Fujian and Jiangxi, probably the climatic counterpart but still not southernmost east China.

Fagaceae.jpg (35.29 KiB) Viewed 1435 times

A temperature increase of about 7°C only adds 2 Quercus species in the eastern US but in eastern China the species number explodes to 2.5-fold, thanks to the southern genera Castanopsis and Lithocarpus and the Quercus subgenus Cyclobalanopsis. Of course, another reason for the difference in the ratio East Asia versus eastern NA may be the different species calculation methods (tree species vs. species in tree genera). It is also possible that a smaller plant is regarded as a tree in Asia, which would increase the Asian result in Latham & Ricklefs’ counting method. Anyway, eastern China is in a class of its own. Western NA has unique conifer diversity.

I look forward to comments, corrections and ideas.

EDIT: I have extended the comparison to the southern temperate world here:

Re: Tree diversity comparison over the northern temperate wo

PostPosted: Fri Dec 26, 2014 12:16 pm
by dbhguru

 An absolutely outstanding piece of work. I am anxious for Lee Frelich to see it and comment. I'll make sure he knows. Another group that would be very interested is Arc of Appalachia in Ohio.

  Can you tell us a little about what pointed you in the direction of this research? Whatever the answer, we're thrilled and appreciative that you did it.


Re: Tree diversity comparison over the northern temperate wo

PostPosted: Fri Dec 26, 2014 12:47 pm
by KoutaR
Thanks, Bob!

I have read so often that the Caucasus is a diversity center and particularly the Caucasian writers always remember to mention Pterocarya, Zelkova etc. I wanted to know how the Caucasus really ranks in a global comparison. So I got the idea.

Appartently the Caucasus is some kind of diverstity center as it is included in the 21 diversity hotspots of the world.
The diversity may be greater in lower plants.

Re: Tree diversity comparison over the northern temperate wo

PostPosted: Fri Dec 26, 2014 6:20 pm
by Chris
I guess it comes down to the definition of temperate and how the regions are defined.

To me, a definition of a temperate forest would be were the majority of angiosperms are deciduous. In most cases, I think this would help to limit those subtropical/Mediterranean species from the comparison (I am assuming those "extra" trees in increasingly south China are evergreen).

Some of the regions have larger diversity of habitat (very tall mountains) and that leads to "inflated" numbers, far beyond what you could ever see within a small geographic area. For example, California has anything from subtropical xeric "trees" to alpine tree lines. Obviously, any given forest will have a much, much lower diversity. On the other hand, it is easily to find a given forest that would have a larger percentage of total species in the region in East US region.

Re: Tree diversity comparison over the northern temperate wo

PostPosted: Sat Dec 27, 2014 4:22 am
by KoutaR

Thank you for your comments!

Indeed, my sub-regions are quite southern, ~all of them including areas that are classified warm temperate or subtropical in different climate maps. The reason, I chose such southern and warm locations, was that I wanted to include the Caucasus and the Balkans. It would be interesting to make another comparison with more northern sub-regions, which could be something like the Netherlands & Germany, Massachusetts and westward, Washington & Oregon and a corresponding sub-region in northeast China.

You are right about habitat diversity. Eliminating the effect of habitat diversity would have been impossible for my resources, so I decided to accept habitat diversity as an intrinsic feature of a region, just like e.g. its history.

Re: Tree diversity comparison over the northern temperate wo

PostPosted: Sat Dec 27, 2014 12:39 pm
by Lee Frelich

Great idea, to compare species richness in areas with similar temperature and size, since we know that both of those factors are strongly related to species richness. I am not surprised to see that the relationships between continents you found are similar to previous studies, since there is a lot of evidence that geologic and evolutionary history contributes in a major way to the differences among occurrences of temperate forest on different continents. And as you say, in Asia there are more opportunities for suptropical species to evolve over time to survive and spread into temperate areas. The opportunities include the contiguous spatial arrangement of the gradient from subtropical to temperate at large (continental) and local (mountains) spatial extents, and millions of years for species to evolve. An then there is the relatively small impacts of glacial advances in eastern China compared to Europe. One of the best studies of tree species richness I have seen was done for Europe (Svenning et al. 2010, Geography, topography and history affect realized-to-potetial tree species richness patterns in Europe, Ecography 33: 1070-1080). This study clearly shows how history affects tree species richness on 50 x 50 km grid cells throughout Europe, with mountain ranges and large water bodies interfering with migration of tree species after departure of the continental glaciers, leading to many areas with fewer species than the climate could support. These impacts are clearly larger in Europe than N. America or Asia.


Re: Tree diversity comparison over the northern temperate wo

PostPosted: Sat Dec 27, 2014 1:25 pm
by Larry Tucei
Kouta-   Wow super posting!!  So the southern hemisphere would have similar comparisons by region?   Say New Zealand, South east Asia, Africa, Cental and South America where some of the largest trees in the world are.   I never thought about the similarities throughout the world, thanks for waking me up to think more globally.   Larry

Re: Tree diversity comparison over the northern temperate wo

PostPosted: Sat Dec 27, 2014 6:21 pm
by KoutaR

Thank you for the comments and article recommendation! I will read it.

I have always doubted the often-claimed hypothesis that much of the northern temperate angiosperm flora has an Asian origin. How would it be possible that Asia creates more new taxa than other continents? But then I read that there were a connection between the tropics and the temperate zone only in Asia for long time, at least in the Cretaceous. The Tethys Ocean separated Europe and Africa, and there was no connection between North and South America. See a map, for example, in Latham & Ricklefs, p. 302. As the angiosperms have a tropical origin this makes clear that the ~only place to give birth to temperate angiosperm taxa was in Asia.


Enlarging the study to the Southern Hemisphere would be interesting. Let's see if I have enough energy to make it. Till now, I have only searched for species congeneric to the northern ones. In western South America the 13.5°C point is somewhere in central Chile and in Australia in southern Victoria. In Africa no coastal place is so cool. In the 168 genera of my material, Chile only has two species of Podocarpus, one Persea and four Prosopis. Victoria has one Podocarpus, two Elaeocarpus, one Ficus, one Symplocos and one Trema. (Sources: and Apparently, the tropical zone has been a much better “insulator” between the northern and southern temperate floras than the vast oceans between continents. Continents have broken up, moved and merged but the tropical zone has always been there. I guess southern Australia would win Chile thanks to the large genus Eucalyptus. Also Acacia are numerous in Australia.

Re: Tree diversity comparison over the northern temperate wo

PostPosted: Thu Jan 01, 2015 5:01 pm
by KoutaR
I uploaded a new version of the Excel file above. I added one species of Acaciella, one Senegalia and one Vachellia to California. These genera were formerly included in Acacia (not in my table but in plant systematics).