Discussion

“Citizen science” efforts to monitor birds over the last half century have produced compelling documentation of large-scale changes in bird populations. Rosenberg et al. (2019) employed multiple standardized bird-monitoring datasets (e.g., BBS) to estimate a cumulative loss of nearly 3 billion birds across North America. They were able to cross-check their results with completely independent estimates of changes in numbers of migrating birds using NEXRAD, a weather-radar network. 

Similarly, the analysis employed the 2 Florida Breeding Bird Atlases to assess changes in bird distribution and used BBS results to cross-check the results. Standard methods and comparable effort are essential to effectively analyze results of atlases that are conducted decades apart. Confidence in observed changes in distribution relies on confidence that the atlas surveys were completed in the same way in the same places. Various changes between atlases, such as access to private or remote sites could have also resulted in changes in the presence or absence of a species that may not have reflected real changes. Given the scale of the project, the span of time in which it was conducted, and reliance on a multitude of volunteers, perfect replication of effort, location, and implementation of methods was not practical or possible. 

The Breeding Bird Survey – conducted annually since 1966 – is a highly standardized effort to measure changes in bird abundance during the breeding season across the continent. For most species changes in abundance as measured by the average number of birds counted on each survey route are reflected in changes in distribution. Although the BBS has fewer than 100 routes in Florida, the sign (positive or negative) of BBS regression slopes of abundance is remarkably consistent with changes in distribution for the species that showed the greatest positive (10 species) and negative (10 species) as measured by a comparison of the two Florida BBAs (Table R6) [hotlink to Table R6]. Of this sample of species, the American Coot was the only species that showed an increasing distribution according to the BBAs, but a negative slope of the BBS abundance regression for Florida. This discrepancy was likely caused by important differences between the two approaches to monitoring. BBS trend results for the American Coot in Florida show high abundance in the 1960s – 1970s followed by a sharp decline. These changes, which happened long before BBA I and II, undoubtedly drove the BBS slope to be negative. One of the ways that both Florida BBAs addressed the issue of comparability of effort was the emphasis on priority blocks. These blocks — one-sixth of a quad — were established to ensure that 1 block per quad would be well-covered and thus comparable between atlases. This approach had the drawback that some species were not detected at all in the priority blocks. Given the high correlation (R3₁₈₀ = 0.98) between priority block and quad results, it made sense to analyze the more inclusive quad level. 

The authors made a macro assessment of changes in breeding birds in Florida. The difference between BBA I and BBA II in the percentage of quads in which a species was recorded was used as a metric of change in distribution. Of the species that were recorded in both atlas efforts (n=198), 133 species of birds showed a decrease, 10 showed no change, and 55 showed an increase in the percentage of quads in which they occurred. Thus, according to a comparison of BBA I and BBA II results, over twice as many species decreased in their distribution as species that increased (ratio of negative to positive = 2.4). This result should be viewed circumspectly because of potential differences in effort between BBA I and BBA II. Using the BBS as a second way of assessing population changes in the breeding birds of Florida, 84 had negative slopes, 2 had slopes of 0, and 50 had positive slopes (ratio of negative to positive = 1.7). 

The results of changes in distribution comport with continental-scale changes. For example, throughout North America, 74% of grassland bird species have declined since 1970 and grassland birds have shown the largest proportional losses in abundance (Rosenberg et al. 2019). Eastern Meadowlark, Eastern Kingbird, Killdeer, and Northern Bobwhite all show declining distributions in Florida according to a comparison of BBA I and BBA II. Thirty years ago, Kale et al. (1992) observed that “fewer species, fortunately, have experienced range contractions” than range expansions, but that conclusion was not supported in comparing BBA I to BBA II. Instead, the contemporary results show that both the BBA II and BBS suggest more bird species are declining than increasing in Florida.

EXOTIC SPECIES

Kale et al. (1992) wanted to document recently introduced or escaped exotic species in Florida whether the record is “a one-time historical event or the beginning of an expanding population”. They noted that, “With rare exceptions, little is known about the breeding biology of these species in Florida. In some cases, the map itself represents all we know about the species in the state.” We took the same approach in BBA II. Some exotic species disappear shortly after introduction, some persist at low population levels (Red-whiskered Bulbul), and others exhibit population explosions (e.g., Eurasian Collared-Dove). From Bailey (1925) to Greenlaw et al. (2014) the number of established exotic (MLEE) bird species in Florida increased from 3 to 14 (Table 1 in Greenlaw et al. 2014), and 20 verifiable non-established, exotic (VNEE) species were recorded in BBA II (Tables R1 and R3). Some species in the pool of VNEEs will likely become established, but which species is difficult to predict. Twelve VNEE species recorded in BBA I were not recorded in BBA II, although this may be caused by variation in effort. James (1997) identified the Common Myna as having a “promising future in Florida” however, defying this prediction, the distribution of Common Myna has decreased slightly in the period between BBA I and BBA II.

Birders and ornithologists are encouraged to work to close the gaps in our knowledge of exotic species. For example, studies of Purple Swamphen (Pranty 2012), Tropical Kingbird (Pranty et al. 2016), Scaly-breasted Munia (Duncan and Duncan 2018), Red-masked Parakeet (Chatfield-Taylor and Epps 2020) and Mitred Parakeet (Epps and Chatfield-Taylor 2020) provide valuable documentation of species new to Florida. Given the potential implications to Florida’s avifauna there is no reason to wait until the next atlas survey period to assess the status of these populations.