Biological Characterization of Natural Forests, Darién Nucleus, Valle del Cauca.

Study Extent

The sampling was carried out through the DryFlor methodology in the Dry forest's ecosystems in the state province of Valle del Cauca in the counties of Calima and Riofrío.

Sampling

The methodology consisted of the DRYFLOR method which provide guidance for monitoring of woody stems (trees and lianas) in long-term permanent dry forests sample plots with a standard protocol to monitor forest biomass, compositional shifts and dynamics, and thus relate these observations to floristics, soil and climate across dry forests. This method consists in different steps: Plot establishment, tree measurement, liana measurement, data recorded, tree bole length and total tree height and smaller trees protocol (Moonlight P. et al., 2022).

Quality Control

In the data structuring and validation process, information quality control was performed. The processes associated with this activity are listed below: 1. Review of documented information associated with biological data; 2. Data structuring in the Darwin Core (DwC) standard according to elements, definition and controlled vocabularies; 3. Validation and cleaning of the information recorded in the DwC, 3.1. Correction of typos, for this task OpenRefine was used; 4. Finally, the DwC database was validated using the Data Validator - GBIF.

Method steps

1. The Dry Forest method implemented consists of (Moonlight P. et al., 2022): Plot Establishment: Location: Each individual plot should: • be on reasonably homogenous soil parent type and soil type, with a homogeneous land use history. • have adequate access • have sufficient long-term security from human disturbance, including intensive grazing; • have sufficient long-term institutional support. Position: Within study sites, plots should be randomly located to avoid bias (e.g., the ‘majestic forest’ bias, Phillips, 1996). If possible, plots should be located far (i>50 m) from the edge of a forest patch or other disturbances such as paths. Orientation: N/S and E/W directions for the principal axes of the plot are the general standard but the eccentricities of the local site may prevent this. The bearings of the main axis, and the latitude, longitude, and elevation of the centre of the plot should be recorded. Note if true or magnetic north is used. Shape and Size: Four rectangular plots of rapid evaluation of dimension 2X50 m (Nominated: 1, 2, 3 and 4) were established.
2. Tree Measurement: The majority of trees in dry forests have diameters <10 cm at 130 cm and many stems split below this height. The usual RAINFOR protocol of only measuring trees ≥10 cm diameter at 130 cm is therefore insufficient for capturing species diversity, structure and carbon storage in dry forests. We therefore measure all trees with a stem diameter ≥5 cm at the usual POM (130 cm) OR a stem diameter ≥5 cm at an extra POM (30 cm). Measurements are taken for all stems ≥5 cm at 130 cm and/or 30 cm (Moonlight P. et al., 2022). The POM and extra POM are not the vertical height above the ground; rather, they should be measured as the straight distance along the trunk, even if it is leaning or bent. 30 cm and 130 cm refer to the distance along the stem from the point the stem leaves the ground. Stems should be measured at exactly 130 cm and 30 cm wherever possible. To aid fast yet accurate measurements, it helps to use a pole marked at both heights, pushed firmly into the leaf litter to the mineral soil next to the tree, to define POM (Swaine, et al. 1987; Condit, 1998). Note: If 130 cm OR 30 cm is not used as the POM or Extra POM for any reason, record the height of the POM or Extra POM (Moonlight P. et al., 2022).
3. Liana measurement Liana measurement Lianas present special measurement challenges for long-term plot studies. The RAINFOR protocol includes a range of methods that maximize long-term comparability across sites and through time at individual sites. We reproduce this protocol herein. Selecting the point of measurement (POM) for lianas is particularly tricky, and has not been well standardized making comparisons among different research groups difficult. Our protocol calls for each liana stem to be measured at three different points, to maximize comparability within the site for time-dependent analyses (growth, recruitment, mortality), across the whole DRYFLOR and RAINFOR dataset, and with other studies worldwide. This follows the protocol pioneered by Alwyn H. Gentry (Moonlight P. et al., 2022; Phillips & Miller, 2002). We include any liana or hemi-epiphyte that reaches 2.5 cm diameter at any point along the stem between 0 and 2.5 m above the ground. 0 is defined as the last rooting point. For lianas and hemi-epiphytes, record the diameter measurements at three points: 1. at 130 cm along the stem from the principal rooting point (=”d1.3length”) 2. at 130 cm vertically above the ground (i.e., 30 cm below the nail in plots where plants are tagged at 1.6 m), (=”d1.3height”) 3. At the widest point on the stem within 2.5 m of the ground (= “dmax”), including any deformity.
4. Data recording: In summary, the following data should be recorded for tree and liana: - Subplot number - Estimated X and Y coordinates from the bottom left-hand corner of the subplot - Tree and liana tag number - Identification - Diameter (mm) at 130 cm - POM, if different from 130 cm
5. Tree bole length and total tree height In addition, tree heights should be measured, to establish plot level diameter/height relationships for accurate modeling of tree-by-tree volumes for each plot, and test whether tree shape differs between stands in different environmental conditions (Feldpausch et al., 2011, 2012). The aim is to characterize the ‘ideal’ height/diameter curve as determined by climatic and edaphic conditions and not confused by the influence of damaged trees. Ideally, every tree in the plot should have its height measured. In practice, this may not be possible due to time restrictions. If this is the case the following procedure is recommended (Moonlight P. et al., 2022): • Estimate the heights of all trees in the plot. All heights should be estimated by the same person to limit differences of opinion. • Excluding trees coded as leaning, rotten, broken, forked below 5 m, fallen or resprouted, aim to sample 50 trees at random per plot for height measurement, while including the ten trees with the largest diameter in this sample.

This dataset contains 491 biological records from the Plantae kingdom. They are distributed in 1 phylo, 3 classes, 22 orders, 36 families, 56 genus and 47 species. 53.76% is classified at the species level and 40.12% at the genus level.

1. Actinidiaceae
   rank: family
2. Anacardiaceae
   rank: family
3. Annonaceae
   rank: family
4. Aquifoliaceae
   rank: family
5. Araceae
   rank: family
6. Araliaceae
   rank: family
7. Arecaceae
   rank: family
8. Bignoniaceae
   rank: family
9. Cannabaceae
   rank: family
10. Chloranthaceae
    rank: family
11. Clusiaceae
    rank: family
12. Cunoniaceae
    rank: family
13. Cyatheaceae
    rank: family
14. Euphorbiaceae
    rank: family
15. Heliconiaceae
    rank: family
16. Hypericaceae
    rank: family
17. Lauraceae
    rank: family
18. Leguminosae
    rank: family
19. Malvaceae
    rank: family
20. Melastomataceae
    rank: family
21. Moraceae
    rank: family
22. Myristicaceae
    rank: family
23. Myrtaceae
    rank: family
24. Ochnaceae
    rank: family
25. Piperaceae
    rank: family
26. Primulaceae
    rank: family
27. Proteaceae
    rank: family
28. Rubiaceae
    rank: family
29. Salicaceae
    rank: family
30. Sapindaceae
    rank: family
31. Sapotaceae
    rank: family
32. Simaroubaceae
    rank: family
33. Siparunaceae
    rank: family
34. Solanaceae
    rank: family
35. Symplocaceae
    rank: family
36. Urticaceae
    rank: family

The sampling was carried out through the DryFlor methodology in the Dry forest's ecosystems in the state province of Valle del Cauca in the counties of Calima and Riofrío.