Naturalist Participation Varies Between Arizona Counties

iNaturalist participation is unevenly distributed at the global, national, and local scale. I looked at county-level observations in Arizona to try to identify trends and potential causes of variability in total observations and percent of observations identified.

Introduction

Simple hypotheses for the total number of observations in each county could be that the iNaturalist observations vary proportionally to population (based on number of observers) or based on land area (based on availability of organisms to observe).  Alternately, a hypothesis could be that areas with more species, or areas with more parks and tourist attractions, might attract more attention from iNaturalist users. 

Simple hypotheses for the percent of observations identified could be that this varies with the population of a county, so that larger population counties would be more likely to have more experts able to identify local observations.  Although counties with more people might also have more total observations, so this hypothesis could suggest that the percent of observations identified is either consistent across counties or, if the identifier effect predominates, there could be a greater percent in counties with more people.  Conversely, rural counties with low populations that also have lots of observations from visitors (such as to national parks) might have a lower proportion of observations identified.

I queried the iNaturalist.org website in December, 2025 for total observations and Research Grade (RG) for each Arizona county.  Some taxa, like fish or reptiles, might me more abundant in one part of the state over another, so to ensure consistency, I chose a taxa of species that is relatively evenly distributed across Arizona:  flowering plants (angiosperms).  I also obtained county population and land area from publicly available information on the web.  I then calculated % observations ID’d to RG, total species observed, and total RG species confirmed for each county and averages across all counties.

Results

Arizona has 15 counties that vary by orders of magnitude in size, population, as well as other factors that contribute to iNaturalist participation such as number of national parks and tourists.  The largest county by population is Maricopa county, with more than 4.5 million people, whereas the smallest county has less than 10,000 people.  The largest county by land area is Coconino county, with more than 18,000 square miles, whereas Santa Cruz, the smallest county, only has a little more than 1,000 square miles. 

Figure 1 - Arizona Counties

iNaturalist observations of flowering plants vary from only about 3,000 in Greenlee county, to almost 300,000 in Pima county.  RG observations vary from less than 2,000 in Greenlee county, to more than 200,000 in Pima county, with corresponding proportions of observations identified ranging from a low of 46% in Navajo county to a high of 74% in La Paz county. 

Counties vary in documented biodiversity from a high of almost 1,900 species in Pima county, to a low of about 430 species in La Paz county. 

AZ Counties	angiosperm observations	RG angiosperm	% ID'd	species count	RG species	% RG species
Pima	298,000	211,000	71%	2,396	1,855	77%
Coconino	134,440	61,500	46%	2,197	1,489	68%
Cochise	122,500	79,550	65%	1,860	1,460	78%
Yavapai	87,000	48,620	56%	1,798	1,326	74%
Santa Cruz	47,000	28,000	60%	1,586	1,273	80%
Maricopa	217,000	156,000	72%	1,800	1,200	67%
Pinal	41,000	29,000	71%	1,288	976	76%
Gila	26,233	15,500	59%	1,240	938	76%
Mohave	36,000	26,000	72%	1,200	890	74%
Apache	20,468	9,575	47%	1,182	832	70%
Graham	12,700	7,780	61%	1,044	774	74%
Navajo	14,000	6,400	46%	1,000	714	71%
Yuma	18,800	13,800	73%	682	521	76%
Greenlee	3,300	1,895	57%	617	432	70%
La Paz	8,110	6,000	74%	517	429	83%
Average	72,437	46,708	62%	1,360	1,007	74%
Total	1,086,551	700,620		20,407	15,109

Table 1 Arizona Counties Angiosperm plant observations and % Identified

To investigate trends across counties I used Tableau Public to create graphs of total observations versus county land area and population.  Because Maricopa county has 3.5 million more people than the next largest county, I excluded it from the analysis as an outlier. 

While there is a trend toward more observations in larger geographic areas, it is not very strong; the county with the greatest number of observations is right in the middle of the distribution of sizes.  The trend toward more observations with more people is a bit better, but it is dominated by the 2^nd most populated county in Arizona, Pima county.  Removing that county from the analysis essentially eliminates the trend, as the next largest county by population (Pinal) is below the average.

 

Figure 2 Observations by Land Area and Population, excluding Maricopa county.

I then looked at whether the activity of identifies varies with land area or population, and again I did not find simple trends that explain the variation.  It does appear that counties with smaller areas are more likely to have a greater proportion of reported species and observations identified.  This could be explained by the greater tractability of small areas, where large counties have more possible habitats, more potential species, and thus pose greater difficulty to botanists to identify plants across the whole county.  While the largest county, Coconino, does have the largest number of species reported, the smallest county (Santa Cruz) has more than the average number of species reported.  Also, two counties that are in the middle of the distribution (Yavapai and Cochise), have some of the highest number of reported species. 

Figure 3 RG Observations %R by Land Area and Population, excluding Maricopa county.

The trend between population and proportion of identified species is even weaker, with almost no trend apparent.  This leads me to reject the hypothesis that more people in an area necessarily lead to more identifications.

Instead, my current hypothesis, is that observation count is driven by other factors outside of total population or land area.  For example, I think that tourists who visit natural wonders such as the Grand Canyon disproportionately increase the number of observations and species in Coconino county, which results in a corresponding decrease in the proportion of observations and species identified.  In contrast, a county that has a strong resident population of people interested in nature, such as Pima county (Tucson), has a large  number of observations and species, but also has  a large proportion of those observations identified to species.