2.1 Site description
This study was conducted near the town of San Isidro de Peñas Blancas, San Ramón canton, Alajuela province, Costa Rica, on land adjacent to the Children’s Eternal Rainforest and the Texas A&M University Soltis Center for Research & Education (10°22’59.1”N, 84°37’06.4”W). All sites were located in the middle Chachagua watershed within 1.7 km from the Soltis Center. The elevation ranges 400-500 m above sea-level and the area is composed primarily of either mature premontane forest (Holdridge, 1966), or converted agricultural land. Average annual temperature is 24 ºC, while precipitation averages 5000 mm per year, with monthly rainfall maximums of 430 mm during the rainy season and 130 mm in the dry season. This qualifies the area as moist tropical forest (D’Odorico et al., 2010).
The study compared microclimate within and under mature forest canopy to microclimate within the canopy boundary layer of three crop sites. We further stratified the forest into two categories: closed canopy (mature forest), and open canopy (in the case of a clearing formed from a wind downburst or a natural forest gap) to account for different canopy structures from dense tall forest to dense low-stature crop and including two additional low-stature crops with more open canopies. The forest is a mix of primary and secondary tropical vegetation, predominantly composed of trees in the Sapotaceae (hibiscus), Moraceae (fig), and Malvaceae (milkwood) families (Miller et al., 2013). The closed canopy forest site was approximately 30 m in height and accessed by a 40 m micrometeorological tower.
The three crop sites were representative of even-aged monoculture plots of papaya (Carica papaya ), taro (Colocasia esculenta ), and sweet potato (Ipomoea batatas ) fields. These vegetation types were chosen due to their economic value to Costa Rica and proximity to the forest sites. Within crops, the average canopy heights were 1.8, 1.3, and 0.7 m, respectively for papaya, taro, and sweet potato. Papaya had a sparse open canopy with large intercanopy gaps. Taro was grown more densely with smaller gaps between canopies, and sweet potato was grown in a dense layer of continuous vegetation near the ground surface.
Measurements within forests were collected in both 2011 and 2014, while the crop site measurements were added during June to July 2014. Within-forest comparisons include measurements from both 2011 and 2014, while comparisons between forests and crops were restricted to 2014. Forest conditions did not change significantly between measurement years.
2.2 Microclimate measurements over the crops
Continuous measurements of air temperature (°C), relative humidity (%), and leaf wetness (%) were averaged every ten minutes. Sensors were mounted on an Onset Computer Corporation HOBO U30 datalogger and Remote Monitoring System (Bourne, MA, USA) placed in the center of the plots. An air temperature and humidity sensor (HOBO S-THB-M00x) was mounted at approximately average canopy height for the three crops. An S-LWA-M003 capacitive grid leaf wetness sensor (Onset Computer Corporation, Bourne, MA, USA) was mounted parallel to the ground at 1.62 m (papaya), 1.41 m (taro), and 0.95 m (sweet potato) outside the canopy. To record total precipitation (mm) every 10-minutes, a tipping-bucket rain gauge (HOBO S-RGA(B)-M002, Bourne, MA, USA) was installed near the top of each station ~1.9 m above the ground.
2.3 Leaf wetness duration over both crops and forest
Leaf wetness in the covered forest condition was measured using an LWS-L dielectric LWD sensor (Campbell Scientific, Inc., Logan, UT, USA), which is based on the same dielectric principles as the Onset LWA sensors. Rain (or more rarely, condensed fog) was collected on the leaf wetness sensor and a current proportional to the water amount was detected by the data logger. The two sensor types, Onset in the crop fields and Campbell in the forest, were manually calibrated at saturation and fully dry end points. For the crop threshold determination, saturation was 100% and dryness was 15%, based on previous work at this site (Aparecido et al., 2016). In wet tropical rainforests, some portions of the canopy can remain wet for a significant fraction of daylight hours (Aparecido et al., 2016). The forest LWD measurements were taken at five heights on the micrometeorological tower, at 5, 12, 24, 34, and 40 m. For the LWS-L, the wet threshold was set at 400 and the dry threshold at 125 (dimensionless units), according to the most consistent fit for the data based on methodology in Aparecido et al. (2016).