Revision of Luquillo Dataset 23: Grid Points Bird Counts from Tue, 2010-06-08 11:07

The Luquillo Grid Points Bird Count is a long-term observational study of the bird community at the Luquillo LTER on Puerto Rico. While Dataset 23 includes files for several sites, this description currently only covers the Luquillo Forest Dynamics Plot (aka El Verde).

 Ecological Level: Community

 Biome: Tropical

 Location: Puerto Rico (Lat/Long)

 Spatial Scale of data: The Luquillo Forest Dynamics Plot is a 16 ha grid and bird data is collected at 40- 25m radius sampling sites on this grid.

 Timespan: 1990-2008

 Frequency: 1-3 times per year

 Observational/Experimental: Observational

 Dataset Location:

 Data collection summary:1-3 times per year, the Luquillo Forest Dynamics Plot is surveyed for birds. Surveys are conducted at 40 permanently located points on the plot. A single observer records for 10 minutes all birds detected within 25 m radius, (though there are additional files available recording data on birds observed > 25 m from the observer, if desired). 

 Data Collected: Species, abundance

 Issues:

• A census consists of the survey of all 40 sampling points, however the sampling of different points may be separated by several weeks. A census, therefore, can consist of sampling spread across more than one month. I have created a file which links date with the census season (spring, summer, fall). 
• separate files exist for each year and for birds surveyed <25 and >25 m from observer.
• Luquillo pub on birds indicated that high percentage (need to track down info) of birds <25 m from observer were identified during standard census. 
• data before Oct 1990 is from a subset of the sampling points used for the rest of the time series (33 pts pre 10/90, 40 pts starting 10/90) 
• Number and timing of censuses varies across years. The "summer" census period is generally consistent across years: it generally starts during may or june.
• number of censuses per year: 1990-1; 1991-3; 1992-2; 1993-2; 1994-2; 1995 through 2008 - 1 census.
• 1997: the single census conducted that year was conducted late (July) and was incomplete (only 15 sites were censused). 
• There are typos for some of the sampling point identifiers:
• 10/22/1990: there are two sets of 6-12 and 6-15 points. One set was conducted at 1136 and 1119 am and the other at 817 and 836 am. The sampling points 3-12 and 3-15 are also missing, indicating that one set of 6-12 and 6-15 is probably 3-12 and 3-15. However, it is impossible from the datafile to distinguish.
• 3/08/1991: A sampling point is listed as o-3. No o-3 exists. However, 3-3 is missing for that census, indicating that o-3 is probably 3-3.
• 5/29/1992: Two 3-12 are listed. One was censused at 820 and the other at 834. Given that all other 3-X sites were done in order, the 834 census is probably actually 3-15. 3-15 is missing. 
• 11/30/1992: There are two 12-15 listed, one sampled at 838 and the other at 934 am. 12-24 is missing. There is no way to disentangle from the database which site is which.
• 3/3/1993, 3/4/1993, 3/11/1993: Duplicate entries for all sample sites exist. I am fairly certain these are indeed duplicates because time of census, weather conditions, and birds surveyed are identical for each pair of duplicates. Note: this was apparently fixed on the Luquillo website since I downloaded that file.
• 6/14/1993: there are 2 12-24 sampling points.
• 3/3/1994: Site 9-7 listed. 9-7 does not exist. 9-3 is missing, suggesting 9-7 is actually 9-3.
• 5/19/1998: Site 6-26 is listed. but 6-21 is missing, suggesting that 6-26 is actually 6-21.
• 5/16/2005: Site 13-24 is listed. 12-24 is missing, suggesting that 13-24 is actually 12-24.
• Weather Column:  Normally 3 letter codes, but some words are spelled out. CLOUDY & CLO, CLEAR & CLE.