This post addresses the following topics:
- Emberizid (sparrow family) trends in the month of January for the past five years using eBird data.
- Making eBird checklist submissions more meaningful.
- Links to understanding the eBird review and data quality process.
Do you notice more Eastern Towhees than usual this winter in the mocosocoBirds area? Based on January data, the numbers bear this out as illustrated in the tables below. All combined Morris and Somerset County data is from the eBird database.
The number of people using eBird has risen dramatically in the past few years. Total numeric accumulations will certainly favor the most recent years. With that in mind, the following tables use averages and percentages, which carry more meaning in a comparative scenario.
Here are descriptions of the terms used:
- Abundance – the average number of individuals per checklist. For example, 0.09 Eastern Towhees in 2014 means there was an average of .09 Eastern Towhees submitted per checklist. This may not appear like much but it is the highest number in the past five years.
- Frequency – the percentage of checklists where a species was observed. In January 2014, 6% of all checklists had Eastern Towhee, the highest percentage in the 5 year sample period.
- Checklist – eBird users, please take note. Only the following checklists are used in Abundance and Frequency reports (and consequently the Bar Graphs as well) according to eBird documentation.
- First, the checklist must be complete. In other words, you must answer Yes to the question: Are you submitting a complete checklist of the birds you were able to identify? Answering Yes satisfies the criteria for Frequency reports.
- Abundance requires a complete checklist that also has species counts, i.e., no Xs. It is hoped the reason is obvious. You cannot tally the non-numeric X.
All the tables are based on the combined data of Morris and Somerset Counties.
Frequency Table (% of checklists) for January of the years sampled:
| Frequency | 2014 | 2013 | 2012 | 2011 | 2010 |
|---|---|---|---|---|---|
| Eastern Towhee | 6.0% | 0.3% | 3.8% | 2.0% | 0.0% |
| Am. Tree Sparrow | 12.9% | 18.8% | 8.2% | 24.9% | 20.7% |
| Chipping Sparrow | 0.7% | 0.3% | 0.2% | 0.0% | 0.0% |
| Field Sparrow | 2.2% | 3.0% | 4.6% | 4.1% | 6.2% |
| Savannah Sparrow | 1.2% | 3.0% | 1.6% | 3.0% | 7.6% |
| Fox Sparrow | 2.8% | 3.2% | 3.4% | 0.0% | 2.8% |
| Song Sparrow | 28.1% | 28.5% | 20.5% | 39.6% | 25.5% |
| Swamp Sparrow | 4.0% | 3.8% | 2.2% | 4.6% | 6.9% |
| White-throated Sparrow | 31.1% | 27.5% | 33.8% | 56.3% | 42.1% |
| White-crowned Sparrow | 0.8% | 2.7% | 2.4% | 3.6% | 0.0% |
| Dark-eyed Junco | 32.9% | 38.3% | 33.8% | 68.5% | 46.2% |
Abundance Table (average per checklist) for January of the years sampled:
| Abundance | 2014 | 2013 | 2012 | 2011 | 2010 |
|---|---|---|---|---|---|
| Eastern Towhee | 0.09 | 0.00 | 0.06 | 0.02 | 0.00 |
| Am. Tree Sparrow | 0.74 | 2.44 | 0.40 | 1.17 | 1.20 |
| Chipping Sparrow | 0.02 | 0.00 | 0.00 | 0.00 | 0.00 |
| Field Sparrow | 0.06 | 0.08 | 0.33 | 0.07 | 0.09 |
| Savannah Sparrow | 0.07 | 0.10 | 0.07 | 0.04 | 0.35 |
| Fox Sparrow | 0.05 | 0.04 | 0.18 | 0.00 | 0.04 |
| Song Sparrow | 1.55 | 1.44 | 1.06 | 1.17 | 0.99 |
| Swamp Sparrow | 0.13 | 0.10 | 0.04 | 0.07 | 0.14 |
| White-throated Sparrow | 3.44 | 2.74 | 6.25 | 3.88 | 3.56 |
| White-crowned Sparrow | 0.02 | 0.12 | 0.07 | 0.04 | 0.00 |
| Dark-eyed Junco | 5.37 | 5.86 | 4.08 | 6.42 | 4.80 |
Here is another example of how submitting a complete checklist affects the numbers. 695 checklists were submitted to eBird from the combined counties of Morris and Somerset in January 2014. Only 598 met the Frequency criteria (complete checklist); only 588 satisfied the Abundance criteria (complete checklist and numeric species count). That means 14% and 15.4% respectively of the checklists were not used for these reports. This is the same data used for bar graphs , a valuable tool for determining date ranges of species in a particular region. Incomplete and X’d records will appear on the Range and Point Maps and on personal and regional lists but not on other reports requiring completed checklists and numerical species tallies. It is hoped that eBird users will embrace the concept of taking their checklists to the next level and make them useful for all types of data reports.
Checklists Table -total number of checklists submitted for January of the years sampled (Morris and Somerset combined). Note that more than 1/3 of 2011’s submitted checklists were not used.
| Checklists | 2014 | 2013 | 2012 | 2011 | 2010 |
|---|---|---|---|---|---|
| Total | 695 | 774 | 684 | 316 | 192 |
| % Freq not used | 14.0% | 22.4% | 27.3% | 37.7% | 24.5% |
| % Abun not used | 15.4% | 23.5% | 28.2% | 39.2% | 28.6% |
It is a convenient shortcut to X certain families, such as gulls, or abundant species, such as European Starlings. Unfortunately, as mentioned earlier, this attitude renders the checklist unusable for certain data reports and research, which is the crux of eBird. The exact amount is not necessary, but simply a number representing proportion is serviceable.
For example, a quick scan shows that there are 1,000 gulls on Lake Hopatcong and that the majority are Herring Gulls. Simply listing 600 Hering Gulls 375 Ring-billed Gulls and 5 Gull spp. is sufficient. The exact number is not important – just a relative proportion, or ratio. If they are too distant to identify, 1,000 gull spp. will suffice. In any case, it is the use of a numeral that is important, not that each separate individual was counted. So, please: Avoid the ‘X’ and say ‘Yes’ to complete checklists.
Here is a relevant article regarding eBird data entry from the Illinois Bird Forum from 2012 entitled The Trouble with ‘X’.
Click on the hyper-link for more information about How to make your eBird checklists more valuable.
Click on the hyper-link to learn more about Understanding the eBird review and data quality process.
mocosocoBirds 
(Hanging head, just submitted a Barnegat ebird list with a number of lazy X’s). Well, I shall try … but it does make Ebird more of a chore. I will view it as the price to pay for having such a useful free resource – payment in effort. Still, I wonder how the data will be read in the distant future. And yes, as a historian, I know that all the caveats and implicit understandings of the words and data we record go missing – and I know our successors will be pondering the numbers as though they were real, before finally deciding we who record them were just incompetent record-keepers – much as we view the ancients’ estimates of warring armies (the Persians sent 800,000 warriors to a battle??? No way, the logistics make that impossible, the chronicler was exaggerating and couldn’t count …)
Proportionate figures are good enough – better than nothing. I just hate to see all of that other good data, like Harlequin and Long-tailed Ducks, et al, go to waste.