Author: Mwihaki John (MJ) | July 2025
Arrival in Antananarivo (Tana) marked the start of Phase II logistics. The day focused on moving equipment out of storage, checking preserved materials, and arranging permits, transport, and accommodation for a multi-site itinerary covering eastern, central-highland, and northern forests. We reviewed survey forms, GPS track templates, and specimen labels, and synced field devices with offline basemaps and elevations to prepare for long days away from connectivity. The team also met with collaborators to finalize access and discuss anticipated road constraints, given seasonal rains and mixed road conditions. The evening was spent calibrating cameras for consistent voucher photos and setting up drying supplies so pressings could begin immediately on our first collecting day. This setup time ensured that both vouchering and habitat notes could proceed smoothly once we left the city.
At Vohimana, east of Andasibe, we began in degraded edge habitat before moving into intact interior forest. The morning transects documented how smaller tree-fern individuals persist in open disturbed areas while taller crowns thrive under closed canopy. Microhabitat notes emphasized light availability, soil moisture, and the proximity to drainage lines. We standardized photo angles for stipes, frond architecture, and crown silhouettes to compare forms across disturbance gradients. The afternoon included voucher pressings and tissue subsampling, with location and elevation recorded for each accession. Vohimana’s mosaic allowed us to capture early evidence of disturbance response within a single site, and we flagged several individuals with unusual characters for follow-up.
A second day at Vohimana focused on repeatable plots to quantify edge-interior contrasts. We resurveyed marked points from Day 1 to verify consistency in trait scoring and added canopy-adjacent stations where humidity and shade were stable through midday. Notes highlight substrate differences (rocky vs. loamy), understory associates, and the presence of recent disturbance indicators along footpaths. We completed additional vouchers and tissue sets where populations showed noteworthy variation in frond texture and apical bud protection. By day’s end, we consolidated labels, verified GPS tracks, and updated the specimen ledger so processing could proceed without data gaps. The combined Vohimana work established a baseline for comparing lightly disturbed versus mature forest structure in subsequent sites.
Maromizaha’s intact rainforest provided a reference setting: high humidity, rugged terrain, dense vegetation, and reliable moisture along valleys and streams. Cyatheaceae were most common in shaded drainages, where we documented microclimate stability and consistent spore release conditions. With intact canopy overhead, we focused on crown architecture, stipe scaling, and frond toughness—traits that may shift along future elevation transects. We mapped clusters along tributaries to understand patch persistence across seasonal flow changes. The site’s relative lack of fragmentation made it ideal for clean morphological comparisons and for assessing how streamside stability supports continuous recruitment. Voucher work emphasized high-quality duplicates and careful field notes for downstream molecular-morphological integration.
In the Central Highlands, Ambohitantely stands out as a small refuge where primary rainforest persists as gully patches within open grassland. Transects traced the edges where forest gives way to savannah, capturing how forest-dependent tree ferns persist at the limits of their ecological tolerance. We noted rockiness, exposure, and wind as likely filters, with individuals occupying narrow humid niches. This patchwork made collections physically demanding yet scientifically valuable, illustrating marginal conditions seldom seen at other sites. Ongoing threats—including deforestation and fragmentation—were visible at several access points. Because the site exemplifies a transitional environment, we prioritized detailed habitat notes to support later niche modeling and to test whether trait differences here exceed those in wetter forests.
The overland transfer to the northeast involved long stretches on mixed-quality roads and several administrative checks. We reorganized equipment for a steep elevation gradient campaign, ensuring desiccants, silica, and field presses were accessible at trailheads. The day included permit confirmations, fuel and food provisioning, and a review of contingency routes in case of road washouts. We also prepared elevation labels and waypoint templates in advance to accelerate voucher processing once in the park. Logistically, this day was critical to protecting fragile specimens and keeping the molecular pipeline intact in the field.
Marojejy, a UNESCO World Heritage site covering nearly 56,000 ha, offered an exceptional transect from lowland rainforest to cloud forest. We documented more than fifteen tree-fern species and recorded noticeable morphological shifts with altitude: crowns tended to compact and fronds became more leathery at higher, wind-exposed sites. Sampling proceeded in elevation bands, with standardized photos and trait measurements to support comparisons across microclimates. We prioritized high-quality vouchers and matched tissue sets, mindful of long hikes between bands. The gradient allowed us to test whether temperature, moisture, and exposure drive predictable form changes and to identify populations for future genetic work on adaptation along steep environmental filters.
Moving farther north required careful timing around weather and surface conditions. Several segments slowed to a crawl due to ruts and washouts. We used the downtime to reconcile labels, back up GPS logs, and pre-sort pressings for shipment versus hand-carry. Administrative follow-ups continued with partner institutions to keep access aligned with the evolving timetable. Although not a collecting day, the consolidation prevented data drift and protected the integrity of the previous days’ collections.
Initial northern surveys targeted shaded ravines and perennial streams where moisture persists through the dry season. Plot notes emphasize canopy height, liana density, and leaf-litter depth—factors shaping understory humidity and spore dispersal. Several patches displayed mixed-age stands, suggesting stable recruitment despite seasonal variability. We captured site-wide panoramas and close-ups of stipe surfaces to compare with Marojejy and highland sites. The day ended with voucher pressing and duplicate preparation for Malagasy and international herbaria, maintaining the chain of custody for downstream curation and sequencing.
Follow-up work expanded to side ravines where mist accumulation supported epiphytes and robust fern growth. We documented gradients from splash zones to calmer understory, noting how frond integrity and crown spread vary with water movement and wind. Tissue sampling proceeded for phenotypes not observed at Marojejy, and new quadrats were added to improve cross-site comparability. Evening checks focused on label accuracy and matching photo IDs to voucher numbers, reducing the risk of metadata drift as collections grew.
On the fifth day in the north, we expanded our survey into nearby ravines with different light and slope conditions. We re-measured a few plants to check accuracy and recorded any clear habitat patterns. Early notes show that the northern sites rely on moisture like the eastern forests, but the canopy and disturbance signs are different. These findings will help us later decide whether the species uses the same niche across distant areas or not.
The return leg provided a short revisit to Ambohitantely to confirm notes in a transitional gully and to capture additional context photos showing forest-savannah boundaries. We compared humidity logger readings with perceived microclimate and confirmed that exposed rims remain limiting for crown expansion. Extra duplicates were pressed from a gully patch to support future anatomical work. The transit highlighted how small refugia buffer edge populations and why detailed habitat data is critical for conservation.
The final day focused on consolidating pressings, refreshing desiccants, and preparing shipments. Field notes were transcribed into the master ledger, with elevation, coordinates, and habitat descriptors double-checked against GPS tracks. We also compiled a short list of candidate populations for priority sequencing. Back in Tana, we confirmed herbarium intake schedules and aligned on timelines for DNA extraction, morphological scoring, and data publication. These steps ensure that the expedition’s collections translate directly into curated references and peer-reviewed outputs.
Thanks to the funding and support from various institutions and individuals, this research was made possible.
