Rapid morphological change in an urban bird due to COVID-19 restrictions

Abstract

The COVID-19 pandemic provided a natural experiment to test the impacts of human activity on urban-dwelling wildlife. Urban dark-eyed juncos differ in bill shape and size in Los Angeles in comparison to local wildlands. We measured juncos that hatched before, during, and after COVID-19 restrictions at a Los Angeles college campus. Birds that hatched during and soon after COVID-19 restrictions had bills that resembled those of local wildland birds. Yet, bills rapidly returned to pre-COVID-19 morphology in birds hatched in the years following pandemic restrictions. Thus, human activity (and lack thereof) underlies rapid morphological change in an urban bird.

As the world becomes more urbanized, wildlife face unique and growing challenges due to rapid environmental changes that are mismatched with habitat characteristics that native populations evolved in (1, 2). Nonetheless, some species have managed to adapt and even thrive in cities. These populations often undergo rapid phenotypic change in response to novel environments (3). Yet, the mechanisms and drivers underlying this change are hard to decipher because novel stressors, landscape, and resource changes co-occur in cities. COVID-19 restrictions acted as a natural experiment where human activity could be separated from other effects of urbanization (4). This extended period of low human activity, the “anthropause,” and subsequent rapid influx of human activity were associated with marked behavioral changes in wildlife, particularly birds (5). However, to our knowledge, morphological shifts in wildlife in response to the anthropause have not yet been documented.

Dark-eyed juncos (Junco hyemalis; hereafter: “juncos”) in California are an ideal system to investigate phenotypic responses to urbanization. These populations have undergone multiple changes, including in morphology, since sedentary urban populations have been established over the past 40 y (6, 7). Interestingly, bill shape and size differ in Los Angeles in comparison to local wildlands—Los Angeles juncos have shorter, thicker bills (7). This unique shift may be due to exploiting food waste, driving shifts in bill shape and size, unlike other populations (7). Indeed, other explanations for bill adaptation relating to urban song, heat, the built environment, and possible founder’s effect cannot fully explain this phenotypic change (7). Yet, it is hard to isolate the impact of human activity—and associated food waste—at the population level. That is, until the COVID-19 pandemic (4).

We asked whether human activity drove shifts in bill shape and size across the population by assessing juncos of known age before, during, and after the anthropause. At the University of California, Los Angeles (UCLA), we began a long-term study of individually marked juncos in 2018 and followed their breeding behavior. We also banded and measured juncos in local mountainous wildlands, which represent their historic habitats. In March 2020, COVID-19 lockdowns and subsequent university-level restrictions began a few days before the first junco nest on campus was expected to hatch (8). These restrictions remained mostly in place until Fall 2021. To confirm the observed lower human activity, we observed a ~sevenfold reduction in human activity in 2021 in comparison to unrestricted 2022 conditions (9). Simultaneously, most dining options were closed until Fall 2021, aligning with a reduction in organic waste at UCLA during the anthropause (SI Appendix). Thus, the urban junco population experienced remarkably lower human activity and food waste from March 2020 to September 2021, and all birds that hatched in 2020 and 2021 experienced low human activity in their early life. As such, we can isolate immediate and cumulative impacts of lower human activity—including food waste—on the junco population’s morphological response. This work highlights the rapid, substantial effects humans can have on animals, including how we can inadvertently shape the lives of other species with whom we share space.

Results

We found that the anthropause was associated with a shift in bill shape and size: Urban juncos that hatched during the anthropause had bill shapes and sizes that were different from preanthropause bills and similar to wildland-type bills (linear model results within-model pairwise comparisons: before-during P < 0.01, before- and after-wildland P < 0.001, during-wildland P > 0.05 for bill shape and size; Fig. 1). Birds that hatched postanthropause returned to a preanthropause baseline for bill size (during-after t = 2.83, P = 0.02; Fig. 1A). When considering anthropause stages (before, during, and after the anthropause), bill shape represented an intermediate phenotype during and after the anthropause (during-after t = 1.29, P = 0.56; during-wildland t = −1.22, P = 0.47; after-wildland t = −3.83, P < 0.001; Fig. 1B). But, year-by-year differences demonstrate an eventual return to preanthropause bill shape and size (Fig. 2). Year-by-year differences also reveal a temporal lag in population-level response: Birds that hatched in 2020, at the beginning of the anthropause, had similar bills as preanthropause birds (0.68 ≤ |t | ≤ 1.44, P > 0.05). Bill shape in birds that hatched in 2020 did not differ from juncos that hatched in 2019 or 2021 (Fig. 2B). Birds that hatched in 2021 and 2022, the second year of the anthropause and soon after the anthropause ended, respectively, had significantly higher bill length and length to depth ratio than those that hatched before the anthropause and juncos that hatched postanthropause in 2023 and 2024 (3.08 ≤ |t | ≤ 7.22, P ≤ 0.05). 2021 and 2022 cohorts’ bills were not significantly different from the wildland-type (0.48 ≤ |t | ≤ 1.82, P > 0.05). Bill traits returned and stayed at preanthropause, urban-associated phenotypes in 2023 and 2024 (Fig. 2).

Fig. 1.

Bill trait variation for urban junco cohorts that hatched before, during, and after the anthropause, i.e., campus closure due to the COVID-19 pandemic, compared to local wildland populations. (A) Residuals bill length regressed by tarsus length (bill size) and (B) residual bill length to bill depth ratio (bill shape) for birds that hatched before, during, and after the anthropause, and adult wildland birds across all years (SI Appendix). Hatch years for each group are in blue bars. Bars represent mean ± SE for each group. Gray represents Los Angeles urban juncos, while green represents southern California wildland juncos. Groups with the same colored triangles are not statistically significantly different from each other. Groups with different colored triangles are statistically significantly different (P ≤ 0.05).

Fig. 2.

Bill trait variation in cohorts of Los Angeles juncos by hatch year before, during, and after the anthropause, with a local wildland reference. (A) Residuals bill length regressed by tarsus length (bill size) and (B) residual bill length to bill depth ratio (bill shape) for cohorts of adult birds that hatched before (2017 to 2019), during (2020 to 2021), and after (2022 to 2025) the anthropause and for adult wildland birds (green) hatched across years. Bars represent mean ± SE for each group. Solid lines represent Los Angeles urban juncos, with bars representing mean ± SE for each given hatch year. Green dashed line and shaded area represent the mean ± SE for the local wildland population. Groups with the same colored triangle are not statistically significantly different from each other. Groups with different colored triangles are statistically significantly different (P ≤ 0.05).

Discussion

Morphological shifts in bills can be driven by adaptive evolution, nonadaptive evolution, and/or plasticity. Adaptive evolution, which requires heritable genetic variation to underlie trait variation, is a likely explanation for the observed shifts. First, bill shape within and across bird species has noted genetic underpinnings and have been found to be highly heritable (10). Second, selection driven by food waste may explain this temporal pattern in bill morphology. Functionally, bill shape typically reflects foraging niche (11). Human activity shapes indirect supplemental feeding through the production of human food waste. Red-winged starlings (Onychognathus morio) used more food waste when human activity was high at a university campus in South Africa, benefiting adult birds’ body condition (12, 13), which has been associated with reproductive investment and success in other birds (14). Alternatively, juncos may have been better able to access more spaces with preferred natural food items without humans present. Observed temporal lags in population-level traits appear to demonstrate cumulative fitness effects of the previous year’s human activity—and lack thereof—on differences in bill shape. Collectively, we speculate that food waste may drive the adaptive evolution of a more generalist bill shape in urban juncos, which rapidly shifts at a population-level in response to resource change. Future work should determine how bill shape varies with foraging, fitness, and genetics in juncos.

A second or contributing explanation for the change in bill traits could be that the lack of human activity could have also facilitated the recruitment of more nonurban juncos into the population. However, we believe recruitment was low as juncos had similar fear response during the anthropause as before, substantially lower on average than nearly all tested nonurban juncos (8). Nonetheless, future work should examine gene flow and population structure within and across populations (see SI Appendix for gene flow observations). Finally, bill shape may also vary due to plasticity in the urban environment; foraging on hard surfaces may shorten bills (15). Thus, early life and foraging substrate may also play a role in adult bill shape.

COVID-19 restrictions provided a natural experiment to examine the impact human activity has on wildlife. In just several years, urban juncos rapidly shifted in their morphology, with juncos that hatched during and soon after the anthropause exhibiting nonurban phenotypes. Yet, the return of human activity led to a rapid shift back to urban-associated morphology. These findings show how rapidly and strongly humans can affect other species. In turn, wildlife can profoundly affect humanity’s health and well-being. Our work highlights the need to understand the complex interactions between humans and other species.

Materials and Methods

We mist-net, measured, and banded birds during from 2018 to 2025 at UCLA in urban Los Angeles and nearby nonurban, forested montane sites in Southern California, 10 to 100 miles from campus. We measured each bird’s tarsus length, bill length, and bill depth and aged them. Collectively, across the years, we captured and measured 302 s-year juncos at UCLA (n₂₀₁₇ = 21; n₂₀₁₈ = 24; n₂₀₁₉ = 17; n₂₀₂₀ = 21; n₂₀₂₁ = 23; n₂₀₂₂ = 39; n₂₀₂₃ = 73; n₂₀₂₄ = 65; subscripts refer to hatch year cohort) and 132 adult juncos across wildland sites. We compared bill traits with linear models and within-model pairwise comparisons. Detailed methods can be found in the SI Appendix. Data can be accessed at (16).

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

.csv and .r data have been deposited in figshare (10.6084/m9.figshare.29803019) (16).