Introduction

Specimen records are the primary source of species information for biodiversity research. However, the information available online is geographically or ecologically biased. The tropical Andes is one of the world's hotspots and is a very complex region as topography and climate variability have created a mosaic of ecosystems and species arrangements. Despite the high species diversity of the tropical Andes, the distribution and completeness of the flora here remains largely unexplored.


Objectives

This study aimed to identify spatial and environmental biases, as well as gaps in the available digital specimen records and to assess the completeness of the Colombian Andes flora.


Methods

All available records of plant species were downloaded from online databases. We characterized spatial coverage and plant collection density using a grid with five cell sizes. In addition, we performed sampling bias analysis for altitude, precipitation and temperature. We calculated the magnitude of the sampling bias using the Kadmon index and the completeness by rarefaction.


Results

Significant bias and gaps were found across environmental, topographic and spatial variables. Spatially, the latitudes 0°– 4°N and 8°–12°N had fewer records than expected, while 4°–8°N had more records than expected. At high altitudes, collection efforts were concentrated in areas above 2000 m, with a temperature regime from 5-15°C and rainfall of 688-1666 mm/year. Scale analysis revealed higher coverage (>90% of the study area) and completeness (60-68%) at high cell sizes (e.g., 100 x 100 km), although, high environmental variability was found in these cells. In contrast, lower coverage (


Implications

Recovering information from small collections that have not yet been stored in a database or digitized, and/or revising the information (already in the database) that has been rejected due to quality problems can increase the coverage of plant collections. There is also an urgent need to sample the flora of the Colombian national protected areas, were most of the ecosystems remain largely unexplored.