Hunting Magnolias, part 1: Spring 2018

There’s been a rush of new tree selection guides for urban foresters recently, ranging from nursery catalogues to university databases, government advice and industry guidance, but whilst these are all making big steps forwards to diversify the species we use in our towns and cities, they all face a similar challenge: they all talk about plant species. This is a problem because whilst the species concept is very useful for botanists, nurseries and gardeners (allowing them to conveniently lump groups of plants together), it’s not very useful if you want to really get to grips with the genetic diversity and ecosystem resilience of our urban forests because the species concept is in fact relatively blunt and masks an astonishing range of diversity: plant species are often distributed across wide geographic ranges, with populations adapting over time to local conditions and evolving in some cases into remarkably different plants- some taller, some with bigger leaves, some with differently coloured flowers, some better adapted to certain stresses than others... What we think of as the English Oak, for example, is also found throughout most of France, Germany, Denmark, Austria, Hungary, northern Spain, all the way east to the Turkish / Iraq border changing shape and form along the way as it needs. In any urban forest there are a whole range of locally different climatic conditions so it makes sense that not all genotypes are going to be equally suitable to a given place.

This situation creates a conflict: we need to be able to select plants in a much more targeted way, recognising the true diversity of nature but there’s no getting away from the fact that grappling with this enormous diversity is a daunting task. But this is the challenge we have set ourselves: at the University of Sheffield we are researching the potential to harness big data and see if we can develop a tool that allows us to target particular populations of trees that can be used in urban forestry research.

Our research hypothesis is that there’s a huge amount of plant distribution data that is already coordinated by GBIF and herbaria, and there’s an ever greater level of precision in modelling existing and future climates, so it should be possible to bring these vast datasets together in a way allows us to understand a place from a plant’s perspective and find the closest match between a genotype in the wild and a designed situation in a town or city. After a great deal of planning in 2016 & ’17, this year we carried out two expeditions to test and refine our models.

Our purpose on this first expedition in the Spring was to recce locations of two Magnolia species (M. obovata and M. salicifolia) that we would be able to revisit later in the summer and study in greater detail: both species are distributed widely across Japan so we wanted to cover as many environmental and ecological gradients as possible, finding places where outlier populations grow.

During our planning we narrowed the study sites down to about 12 warm temperate, cold temperate and sub alpine habitats ranging from Ibaraki Prefecture on the east coast of Honshu island, down to Miyazaki Prefecture in southern central Kyushu island. Timing of the expedition was crucial: we wanted to visit before leaf break (which would make it difficult to find the Magnolias amongst all the other trees growing) but after the snows had melted. This gave a window of about 4–5 weeks that we could work in and although it meant that we might miss the trees flowering (which would have made it much easier to find target trees), we ended up with a list of nearly 200 trees that we could evaluate in the early summer as part of the planning for the next expedition.

We surveyed each location systematically, noting first the physical characteristics such as altitude, accessibility, gradients and aspect, before moving on to recording habitat summaries, noting key plant species and the successional stage of the habitat, and finally, identifying target trees that we could revisit later in the year, recording DBH, height, canopy spread and health.

Tea plantation in Ibaraki Prefecture at a low altitude

Veratrum nigrum rosettes alongside a stream at the base of Mt Yamizo

A map showing footpaths around Mt Yamizo

We found our first Magnolia kobus along streamsides at 400m but Magnolia salicifolia remained elusive

A waterfall on Mt Yamizo at 500m, with the banks covered in Kerria japonica

Simon Hannus recording young, leafless Magnolia obovata by the Cryptomeria plantation

Tochigi, Fukushima and Ibaraki Prefectures, seen from the peak of Mt Yamizo

Our next stop was the Japanese alps in Gifu, with a base camp at 1,200m in the Takayama River Basin

In the valleys we again found many Magnolia kobus in flower

Cherry trees still in flower in mid April at a cemetery, halfway up the mountain

We reach our study site, where Dr Yoshitake Shinpei is researching forest carbon cycling and climate change

Our study site, a birch and oak dominated secondary woodland on an east facing slope at 1,200m. Here we found many Magnolia obovata and a few Magnolia salicifolia

Having found a good number of target trees between 800–1300m in the alps, we decide to see if we can find any at higher altitudes and head towards sub-alpine habitats on Mt Norikura

The snows have only recently melted, with stunningly clear mountain streams and wetlands

The mountains are grazed at very low intensity: the landscape here is extremely patchy with many habitat niches and plant communities

Large swathes of Japanese skunk cabbage (Lysichiton camtschatcensis) amongst birches in stagnant ponds at 1800m

Climbing down the mountain having found no Magnolias, we were surprised to see a groundcover plant grown widely in the UK (Pachysandra terminalis) growing happily in wetlands at 1,400m

Driving south we came across a Wasabi farm

The hydrology was fascinating, diverting a stream across complex microtopographical changes

Our last study site on Honshu was the satoyama wetlands in Kaisho forest, east of Nagoya at 300m

The climate here is warm temperate, with species from cold temperate and sub tropical habitats

We found Magnolia stellata in some of the wetlands, hybridising naturally with M. salicifolia. A real challenge to identify pure M. salicifolia

Magnolia obovata, on the other hand, was very easy to identify and we found a good many specimens

Our next stop was Aburayama Forest, another warm temperate habitat to the south west of Fukuoka on Kyushu Island

We found many Arisaema species growing in part shade, both in Cryptomeria plantations and natural woodland

Perhaps my favourite plant of the expedition was this Lindera erythrocarpa, in flower and breaking into leaf

We found lots of M. obovata here, just coming into leaf. M. salicifolia was harder to find as it had finished flowering but the leaves had not yet broken- they were only found on exposed slopes near the mountain peak

Our last study site was the furthest south, near Shiiba in Myazaki Prefecture on Kyushu Island

Shiiba, a remote village in the only cold temperate part of Kyushu Island

This area sits at the meeting point of two geological formations, creating a highly heterogenous mix of habitats

Climbing to 1,100m we found M. salicifolia and Rhododendron pentaphyllum flowering along riverbanks

M. salicifolia in flower. We also found some trees of this species with pink tepals, very rare

A Magnolia obovata, tagged for easy identification later in the summer. Note the shallow roots. This tree was found amongst a commnity of Betula grossa and Stewartia monodelpha

Over the summer we studied the data and narrowed the list down to about 120 trees that we would want to study in more detail during the second expedition. Across the sites a few themes emerged: Magnolia kobus was not nearly as common as we expected, and tended to be restricted to the banks of streams and lakes, whilst the closely related Magnolia salicifolia tended to behave in a similar way to birches in the UK, typically as a pioneer species. On the other hand, Magnolia obovata was very widely distributed across a whole range of habitats from low to high altitude, varying its leaf characteristics (such as size, width and hariness) accordingly- we even found it self-seeded, growing out of the top of a retaining wall on a south facing slope in Fukuoka! Perhaps we can start to challenge the idea that Magnolias only like rich, moist soils and in fact can be quite a bit hardier than we expect- as long as you find the right genotype.