Things are not always better on the sunny side!

Chris Birkinshaw is an assistant curator in the Missouri Botanical Garden’s Madagascar Program, based in Antananarivo. He describes his observations on forest succession at Ankafobe, a site in the central highlands.

Anyone flying over Madagascar’s highly dissected central highlands will be struck at first by the vast grasslands that dominate this landscape.  But, those looking more carefully will also detect pockets of forest within the rich network of valleys.  These forests have a distinct fauna and flora but, perhaps because of their small size, they have attracted little interest from conservationists.  Consequently, in the last few decades, the majority have been degraded or entirely destroyed as their trees were cut for timber or charcoal and the relicts burnt by wild fires that rage over this landscape in the dry season.

The Ankafobe Forest, located some 135 km NW of Antananarivo, is currently being designated as new protected area by Missouri Botanical Garden’s Madagascar Research and Conservation Program.  It is one of the larger remaining areas of highland forest but, here too, the forest has been impacted by exploitation for timber and charcoal and burning by wild fires.

Efforts are underway to restore this forest to its former extent in the recent past.  This is no easy task because away from the current forest edge tree seedlings are subjected to harsh conditions: soils impoverished and compacted by annual burning, grasses that compete greedily for water and nutrients, an extended 7-month long dry season, and exposure to hot sunshine and strong desiccating winds.  Even when firebreaks are used to prevent wildfires from penetrating the grassland surrounding the forest, few tree seedlings naturally colonize outside of nurturing limits to the forest.

Few but not none.  A closer inspection of the landscape reveals some woody plants in the grassland on the less sunny south-facing slopes surrounding the forest (south is less sunny because Madagascar is in the southern hemisphere). Perhaps then the forest could be helped to expand by planting young trees preferentially on these slopes?

Ankafobe Forest South-facing on left.JPG

Vegetation is lusher on south-facing slopes (left) compared to north-facing slopes (right) at Ankafobe, a proposed conservation area in highland Madagascar.

To test this idea in 2017 we planted 25 nine-month old seedlings of each of four native tree species in grassland 20 m from the forest edge on both a south-facing slope and a north-facing slope.  The species were selected for this test are native to the Ankafobe Forest and were available at the local tree nursery when the experiment was installed.  After 12 months the survival and growth of these young plants were measured.

All four species survived well on the south-facing slope but only one species, Nuxia capitata, had good survival on the north-facing slope.  Mortality of Uapaca densifolia was total on the north-facing slopes.  Growth was sluggish on both the south-facing and north-facing slopes with the exception of Nuxia capitata on the south-facing slope that had a mean 12-month growth exceeding 20 cm.  These results suggest that south-facing slopes may provide the best results, at least at Ankafobe, for forest restoration endeavors.

South- facing North-facing
Species % Survival Average growth (cm) % Survival Average growth (cm)
Eugenia pluricymosa 72% 4.1 8% 3.0
Baronia taratana 88% 9.1 28% 12.4
Nuxia capitata 96% 21.5 100% 8.7
Uapaca densifolia 72% 10.5 0%

Aspect – the direction that a slope faces – makes a big difference for vegetation in the temperate zone, especially in dry places. But it is not often considered in tropical ecology. Directly or indirectly, the difference in sun exposure between the slopes at Ankafobe can make the difference between life and death for young trees growing in this hostile, water-stressed environment.

To read more blog posts about the restoration efforts at Ankafobe, please click here. You may also read a 2019 open access paper about seedling trials at this site here.

Homemade mycorrhizal inoculum improves seedling growth for some native Malagasy trees

MBG Madagascar’s Chris Birkinshaw and Dinasoa Tahirinirainy describe exciting, preliminary results from a forest restoration experiment in highland Madagascar.


Ankafobe Forest on Malagasy highlands - experiment located on grassy ridges

This sliver of riparian forest is one of the last vestiges of Madagascar’s highland forests. Decades of Missouri Botanical Garden research in Madagascar have shown that more than 80% of all plant species on the island exist nowhere else. Many are threatened with extinction due to habitat loss. Several previous posts have described forest restoration efforts at this site, home to the largest population of the endemic sohisika tree (Schizolaena tampoketsana) – a species that belongs to a family (Sarcolanaceae) that only exists on Madagascar.

A small number of forest restoration projects in Madagascar routinely inoculate the tree seedlings in their nurseries with a homemade mycorrhizal inoculum. While the nurserymen are convinced that this technique promotes growth and survival of tree seedlings, there seems to be no published data objectively demonstrating these positive outcomes. In an effort to provide the evidence to justify investment in this technique, we designed a simple experiment that will compare the survival and growth under four treatments of young plants of six native trees planted in grassland adjacent to the Ankafobe Forest on the central Malagasy highlands.

Table – Four experimental treatments to test the effects of mulch and mycorrhizal inoculum on native tree seedling growth in highland Madagascar

  Inoculated Not inoculated
Mulched Treatment 1 Treatment 2
Not mulched Treatment 3 Treatment 4 (control)

In our experiment, fifteen seedlings of each of six native tree species will be grown under each of the four treatments listed above. The mycorrhizal inoculum was made by filling a pit (150 cm long × 50 cm wide × 30 cm deep) lined with sacks with topsoil collected from around the roots of three native tree species, then growing maize and beans in this soil for three months before cutting these plants down and letting the substrate dry out for two weeks. The substrate remaining in the pit is the inoculum and was used by adding one tablespoon to each seedling container.

Making the inoculum

Beans and maize are grown in topsoil collected from a remnant forest to amplify local mycorrhizae populations. This enriched soil (i.e., inoculum) is then added to seedling containers.

The tree seedlings that received mycorrhizal enrichment were inoculated in November 2017, and all of the seedlings were otherwise grown under the same conditions in the nursery until January 2018 when they were planted out into an experimental plot at Ankafobe. Half of the tree seedlings were surrounded by a thick layer of grass-based mulch (~30-cm deep). The comparison of seedling performance with and without the addition of mulch is interesting because of the possibility that mulch helps to maintain a relatively cool and moist environment in which the mycorrhizae can flourish.

Table – Mean difference in tree seedling height (cm) between seedlings inoculated versus not inoculated with homemade mycorrhizae, after two months in the nursery (N = 30 seedlings per species).

Species Inoculated seedling height Non-inoculated seedling height t p1
Aphloia theiformis 29.5 ± 10.2 33.4 ± 6.5 -1.75 1.0000
Baronia tarantana 18.1 ± 6.1 11.4 ± 3.5 5.21 <0.0001
Brachylaena ramiflora 27.2 ± 6.0 31.8 ± 6.5 -2.87 1.0000
Craspidospermum verticillatum 43.0 ± 5.9 42.6 ± 3.7 0.37 1.0000
Macaranga alnifolia 34.8 ± 8.6 39.2 ± 5.2 -2.43 1.0000
Uapaca densifolia 23.0 ± 7.9 11.5 ± 2.6 7.58 <0.0001

1 t and p values are from a one-tailed student’s t-test asking whether inoculated seedling height was greater than non-inoculated seedling height. P values are adjusted for multiple comparisons with Bonferroni correction.

Although we plan to measure seedling survival and growth 12 months from the time when they were planted (i.e., in January 2019), we were interested to see that for two of the species the height of inoculated seedlings was significantly greater than the height of non-inoculated seedlings after a mere two months in the nursery. On average, inoculated seedlings of Baronia tarantana are 1.6× taller than non-inoculated seedlings; while the seedlings of Uapaca densifolia are a full 2× taller. For the other species there was no significant difference between the height of the inoculated and non-inoculated plants.

Experiment showing line of seedlings some with and some without mulch (1)

Tree seedlings are planted out in a field experiment at Ankafobe in January 2018. These seedlings are planted adjacent to a line of “green manure” (i.e., nitrogen-fixing Tephrosia shrubs planted to improve the degraded highland soil prior to planting native tree seedlings).


Can fungus help grow trees in Madagascar?

Thomas Timberlake and Cyprien Miandrimanana write from Madagascar about a field experiment using fungus to help tree seedlings survive.

One of the problems that has long bedeviled ecological restoration efforts in Madagascar is persuading young seedlings to grow at a pace of more than just a few centimetres per year. The site of Ankafobe in the central highlands is a prime example, with many five year old individuals, planted in the anthropogenic grassland surrounding the remaining forest fragments, still no taller than waist height. Clearly, the environment into which the seedlings are planted is in some way inhospitable.  One hypothesis to explain seedling underperformance  is that they are not managing to establish their normal symbiotic relationships with vesicular arbuscular mycorrhizae (VAM) fungi on which most higher plants depend.

Scaled visual comparison of VAM and VAM-less seedlings at Mitsinjo in Andasibe, Madagascar.

Scaled visual comparison of VAM and VAM-less seedlings at Mitsinjo in Andasibe, Madagascar.

In a VAM symbiosis, plants exchange a significant carbohydrate donation to the fungus in return for important nutrients, particularly phosphorus, and often increased drought tolerance. So if mycorrhizae propagules are absent in the savanna soil, this could well explain the slow growth rates and high mortality observed among planted tree seedlings at sites like Ankafobe.

In response to concern about poor seedling performance, various restoration projects in Madagascar have begun inoculating their nursery seedlings with VAM using a simple protocol pioneered by Mitsinjo, a restoration project in the eastern rain forest of Andasibe. Soil (presumed to contain mycorrhizal fungus) is gathered from underneath forest trees, mixed with sand in a sack-lined pit and then sown with rice and beans to act as hosts for the developing VAM. After three months of maturation, you have a sack-full of VAM inoculum, ready to be applied to the young germinating seedlings – one teaspoon per plant.

Many groups in Madagascar swear by the VAM protocol and the visual results can be compelling, but as yet there have been no experiments in the country to rigorously test whether this method is actually effective. This lack of clear evidence is what prompted us to work on a series of experiments testing and perhaps refining the VAM protocol.

We planted 480 native tree seedlings with and without VAM inoculation to test whether this method increases seedling survival and growth in the degraded savanna around Ankafobe. Digging into the solid laterite and planting the experimental seedlings was hard work but our efforts were rewarded one day with the sighting of a family of 10 young Tenrecs (Tenrec ecaudatus) who ventured bravely out of the security of the forest to observe the progress.

Planting complete, we took our “Time Zero” measurements and then a small sample of roots from both VAM and control seedlings to return to Antananarivo and check for the presence of mycorrhizae vesicles. The process of staining involved cooking up some rather nasty chemicals in our improvised laboratory – the kitchen – back in Tana.

Our next project will be to replicate our VAM study in Ananalava, a humid site on the east coast that contrasts with the drier climate of the Malagasy Highlands. Repeating our study in different environments will help generalize our results and recommendations for people working across this heterogeneous island.

Cyprien in our kitchen laboratory preparing an improvised stain to look for VAM vesicles.

Cyprien in our kitchen laboratory preparing an improvised stain to look for VAM vesicles.