Wildlife populations that co-vary in time and space have been investigated by scientists since the birth of ecology (Chapman, 1928; Elton, 1949; Andrewartha and Birch, 1954; Krebs, 1985; Brewer, 1988). Synchronous population dynamics, defined as coincident changes in abundance (Liebhold et al., 2004), have been observed across taxonomic groups, including invertebrates (Sutcliffe et al., 1996), fish (Myers et al., 1997), birds (Michel et al., 2016), and mammals (Ims and Steen, 1990). Despite its historical foundation, the patterns, causes, and consequences of synchronous population fluctuations are still not well understood (Liebhold et al., 2004). Most examples of synchrony come from disjunct populations of the same species (Burrows et al., 2002; Post and Forchhammer, 2002; Bellamy et al., 2003; Krebs et al., 2013) or closely related species (Raimondo et al., 2004a; Raimondo et al., 2004b; Robertson et al., 2015).

Synchrony between populations from different taxonomic groups is thought to be driven by exogenous factors such as environmental stochasticity (Moran, 1953) and trophic interactions such as competition and predation (Liebhold et al., 2004). There is considerable evidence that exogenous factors can cause synchrony in closely related species (Cavanaugh and Marshall, 1972; Ranta et al., 1997; Kendall et al., 2000; Koenig and Liebhold, 2016). Two sympatric populations may co-occur without synchronizing until a catalyst, such as changes in weather patterns, creates conditions for synchrony (Moran, 1953). Additionally, factors such as changes in habitat quality and weather patterns (i.e., the North Atlantic Oscillation) can alter the amount of synchrony found among populations across space and time (Hurrell, 1995; Ranta et al., 1997; Ranta et al., 1998; Koenig, 2001; Allstadt et al., 2015). While these patterns appear to be clear for closely related species, they are unclear when synchrony occurs in populations of different taxonomic groups.

Much of our understanding of synchrony between populations from different taxonomic groups has centered on trophic interactions, via predator–prey dynamics (Ims and Steen, 1990; De Roos et al., 1991; Gurney et al., 1998; Spiller et al., 2016) that occur in relatively simple communities of predators and prey (Lack, 1954; Angelstam et al., 1984). While the concepts generated from simple systems have been broadly applied (Davenport and Chalcraft, 2012; Nordberg and Schwarzkopf, 2019), rarely have they been examined in complex predator–prey systems. Meanwhile, other trophic mechanisms like competition for resources (Koenig, 2001; Jones et al., 2003) and similarity in species’ reproductive strategies may also increase the opportunity for synchrony (Moran, 1953; Liebhold et al., 2004). Additionally, there is some evidence (Liebhold et al., 2004) but little understanding of how trophic interactions and exogenous factors may interact to influence synchrony between populations from different taxonomic groups (Bjørnstad et al., 1999).

Two species that may allow us to better understand the influence of the synergies of exogenous factors and trophic interactions in a complex food web are hispid cotton rats (Sigmodon hispidus) and northern bobwhite (Colinus virginianus; hereafter bobwhite; Staller et al., 2005; Morris et al., 2011). Cotton rats and bobwhite are dominant primary consumers in ecosystems throughout the southeastern United States. Both species serve as an important food source to several shared generalist avian, mammalian, and reptilian secondary consumers (Schnell, 1968; Barrett et al., 2001). Anecdotal evidence of synchronous population fluctuations in cotton rat and bobwhite populations have been observed, leading some to hypothesize a trophic interaction mechanism of synchrony (Errington and Stoddard, 1938; Schnell, 1968; Barrett et al., 2001; Staller et al., 2005). However, environmental conditions can also influence both populations, for example, bobwhite chick survival can be sensitive to rainfall during their restricted breeding season (Terhune et al., 2019). Alternatively, variation in seasonal temperatures has been correlated with cotton rat reproductive activity (Goertz, 1965) and abundance (Rehmeier et al., 2005).

Our objective was to understand the patterns and drivers of synchrony for two sympatric primary consumers in a complex food web. Specifically, we wanted to address the following questions: 1) Do bobwhite and cotton rats synchronize consistently across space and time? 2) What is the relative influence of trophic interactions vs. exogenous factors on the amount of synchrony between populations, and 3) does the strength of trophic interactions increase with increasing synchrony between populations? Using datasets from sites in their southeastern geographic ranges, we predicted that bobwhite and cotton rats would exhibit punctuated but inconsistent periods of synchrony because the species’ reproductive potential is influenced differently by environmental variation. In support of Moran’s theorem of synchrony (Moran, 1953; Stien et al., 2012), we predicted that periods of synchronous fluctuations would be more closely associated with exogenous factors because populations of these species can be sensitive to climatic variation (Eifler and Slade, 1999; Perez et al., 2002; Hernández et al., 2005; Rehmeier et al., 2005). Finally, we predicted that trophic interactions (e.g., predation and competition) would not vary with the relative amount of synchrony between populations because empirical evidence (Miller and Epstein, 1986; Post and Forchhammer, 2002; Raimondo et al., 2004a) and theory suggest that exogenous factors drive synchrony in unrelated species (Moran, 1953; Royama, 1992; Koenig, 2001).

In our examination of a complex predator–prey system, we found evidence of regional and site-specific synchrony between two taxonomically divergent primary consumers demonstrated by both substantial levels of concordance and high proportions of concurrent maxima and minima. Unlike many synchronous populations, we found negligible evidence that variation in the relative amounts of synchrony were tied to exogenous factors (Post and Forchhammer, 2002; Stien et al., 2012). We did find that trophic interactions between cotton rats and bobwhite were associated with relative synchrony between these populations, but these interactions appeared to weaken in years of greater synchrony (Lee et al., 2020). If populations were synchronized by mechanisms like predation and competition, we would have expected a positive association between synchrony and trophic interactions (Ims and Steen, 1990; De Roos et al., 1991). Accordingly, our research suggests that the patterns of fluctuation among taxonomically divergent primary consumers in our system cannot be easily explained by our current suite of theoretical constructs that focus solely on exogenous factors (Moran, 1953) or trophic predator–prey interactions (Hagen, 1952; Lack, 1954; Angelstam et al., 1984) and may be influenced by the interaction of these factors (Coulson et al., 2004; Stenseth et al., 2004).

The pattern of synchrony between bobwhite and cotton rats varied considerably across years, fluctuating from no relationship to moderate synchrony. Declines in relative synchrony were associated with increasing bobwhite density. When regional synchrony between the populations was highest (2003, 2005, 2006, 2007), environmental conditions were characterized by above average greenness across multiple seasons and summer precipitation. During those same years, bobwhite density which are known to be sensitive to temperature and precipitation (Table 1; Lusk et al., 2001), was low; and cotton rat density, sensitive to seasonal greenness, precipitation, and temperature (Table 1; Reed and Slade, 2006), was variable. Alternatively, during years of reduced synchrony (2002, 2004, 2010), environmental conditions were characterized by above average spring precipitation and below average winter and summer precipitation as well as reduced cotton rat densities and average bobwhite densities. These patterns suggest that differences in synchrony can potentially be partially explained by the subtly different ways climatic conditions influence the population densities of these taxonomically divergent species. Yet they do not conform well with Moran (1953), commonly used to explain patterns between populations, suggests synchrony occurs during extreme weather events that reduce population densities. Following this concept, we would have expected increases in synchrony when bobwhite and cotton rat densities were both at their lowest. Based on maxima and minima across sites, bobwhite and cotton rat populations fluctuated concurrently in more than 50% of the study years (SI 7); however, these fluctuations were not clearly linked to exogenous factors as indicators of population change. Across sites, cotton rats appear to deviate from synchronous fluctuations more frequently than bobwhite (SI 7), possibly due to their reproductive plasticity relative to bobwhite’s defined breeding season. Cotton rats’ natural history facilitates opportunistic breeding when environmental conditions are favorable (Linzey, 1998), while bobwhite are restricted to a defined breeding season (Stoddard, 1931). Accordingly, although exogenous factors appeared to influence the population dynamics of each species individually and the proportions of concurrent maxima and minima were substantial across sites, there was not a strong relationship between synchrony and exogenous factors. However, it is important to note that we did not observe the extreme weather events that commonly support Moran (1953).

While both populations were positively associated with one another, only cotton rats were influenced by the interaction of relative synchrony and bobwhite density (Table 4). Numerous observational studies have suggested that cotton rats positively influence bobwhite density through generalist predators switching their prey selection (Staller et al., 2005; Ellis-Felege et al., 2017; Palmer et al., 2019). Predation by birds of prey and mammals is the leading cause of bobwhite mortality at all life stages (Burger et al., 1995; Rollins and Carroll, 2001; Cox et al., 2004). Similarly, avian and mammalian predators account for the majority of mortalities of adult cotton rats (Morris et al., 2011; McCampbell et al., 2023). Moreover, seasonal predation pressure has been shown to considerably alter both cotton rat (Wiegert, 1972) and bobwhite populations (Carroll et al., 2007). We found that as synchrony between bobwhite and cotton rats increases, the positive association of bobwhite density on cotton rat density declines, possibly due to prey switching by predators when both species populations are at relatively high densities. This differs from the trophic interactions commonly attributed to synchrony in taxonomically divergent primary consumer species. The alternative prey hypothesis (APH) suggests that predators are selective and synchronize prey population densities by depredating their primary prey (i.e., numerically dominant) until its population declines before switching to an alternative prey and instigating its decline (Hagen, 1952; Lack, 1954; Angelstam et al., 1984). We have no evidence that the generalist predators in our system are selective (Godbois et al., 2003; Cherry et al., 2016; Rectenwald et al., 2021), and if the APH was occurring, we would expect an inverse relationship between prey densities. As cotton rat densities are reduced by predation, bobwhite density would increase to synchronize with cotton rat densities (Angelstam et al., 1984). However, we observed periods of elevated synchrony after years of both increasing and decreasing cotton rats.

Another explanation of trophic interactions is the shared predator hypothesis which posits that predators can synchronize prey species populations through indiscriminate predation of primary and alternate prey to cause simultaneous increases and declines (Norrdahl and Korpimäki, 2000). Synchrony seemed to occur after years of both increasing and decreasing cotton rats; however, these associations were inconsistent. Under the shared predator hypothesis, we would also expect that alternative prey (e.g., bobwhite) and primary prey (e.g., cotton rat) would consistently decline after spikes in the population of the numerically dominant prey, which should facilitate increased predator activity and encounter rates (Bety et al., 2002; Ježková et al., 2014); however, we did not see evidence of this pattern either (Ydenberg, 1987; Ims and Steen, 1990).

Without a clear mechanistic explanation of the patterns in our data, we propose an alternative climate- mediated predation framework for understanding the linkages of prey in complex predator–prey systems with numerous non-selective predators. First, in keeping with classic bottom-up theories of regulation (White, 1978; Hunter and Price, 1992), exogenous conditions influence the populations of each prey species via availability of resources (Meserve et al., 2001; Meserve et al., 2009). Next, population fluctuations of the numerically dominant prey changes predator communities. Finally, changes in the predator community alter the predation risk and populations of species with the same predators in a manner consistent with the shared predator hypothesis. While this hypothesis needs to be tested, there is substantial support for its components, 1) fluctuations of numerically dominant prey are closely linked to climate-induced changes in resources (Lima et al., 1999; Ernest et al., 2000), 2) fluctuations of numerically dominate prey are tied to changes in populations and communities of their predators (Post and Forchhammer, 2002; Beaugrand et al., 2009; Turkia et al., 2020), and 3) changes in predator communities alter the population demographics of species with the same predators (Rooney et al., 2006; Stoessel et al., 2019; Quéroué et al., 2021). This framework (Figure 7) is more consistent with many patterns of our data than existing constructs.

Although we used data from three sites across 15 years, our study had several limitations which should be noted. Most importantly, because of the coarse spatial and temporal resolution (i.e., annual) of our population and remote sensing data, the patterns in our data might change at more biologically relevant temporal and spatial scales that more closely track species’ interactions. Another important caveat was that cotton rat and bobwhite population data were collected in different areas of each study site, which may have inadvertently influenced our findings as the species were not experiencing the same microclimatic conditions. Moreover, while annual population density data were collected at times of peak density, our sampling occurred at slightly different seasons for each species, potentially limiting our ability to detected linkages between them. Both species have an average lifespan of approximately 6 months (Brennan, 1999; Curlee and Cooper, 2012; McCampbell et al., 2023), suggesting limited annual carryover of population densities; however, we did not investigate this assumption with the inclusion of time lags. Additionally, without data on predator density we were unable to fully investigate the drivers of bobwhite and cotton rat population fluctuations. Although cotton rats are the most commonly trapped small mammal at these sites, they are not the only small mammals in these systems. Fluctuations in other small mammal populations, most notably Peromyscus gossypinus (cotton mouse) and Mus musculus (house mouse) may have influenced our results and therefore these populations should be acknowledged and accounted for in future studies. Another concern was the variation in population data collection, especially bobwhite density, across sites. While two sites used the same methodology, the assumptions made regarding covey size at the Jones Center may have skewed the data and influenced the results. Lastly, our lack of strong effects may have been influenced by the removal or mesomammalian predators and supplemental feeding that occurred across all sites and throughout the study. Although we have identified factors associated with synchrony within our dataset, further investigation should aim to collect population and predator data on finer spatial and temporal scales to capture the within-year variation in population fluctuations.

While a climate-mediated predation framework may provide a better explanation of the fluctuation patterns in populations of taxonomically divergent primary consumers in complex food webs, there is considerable work that needs to be done before attributing mechanisms to these patterns. Specifically, we suggest two important steps to determine the applicability of this construct. First, there is a need to link survival, cause-specific mortality, predator populations, and specific resources (i.e., vegetation characteristics, food availability) to the annual fluctuations of primary and alternative prey species. To do this, contrary to the coarse resolution of this study, research will need to investigate these populations on finer spatiotemporal scales that allow for the inference of mechanism. Second, we advocate for manipulations of resources for the numerically dominant prey to determine if they lead to population increases, alter predator communities, and change the predation rate of other prey populations in the system. We present this paper as a foundation to generate a greater understanding of the integral drivers of synchrony within complex systems.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was reviewed and approved by Tall Timbers Research Station and Land Conservancy: IACUC permit GB-2001-01-15; the Jones Center at Ichauway: Georgia Department of Natural Resources Scientific collecting permit 1000528068; the private property in Baker County, Georgia: Georgia Department of Natural Resources Scientific collecting permit 1000650622.

KRH: data analysis, study design, manuscript writing. LC: study design, data collection, manuscript editing. SJ: study design, data collection. GM: data collection, manuscript editing. WP: study design, data collection, manuscript editing. BR: study design, data collection. DS: study design, data collection, manuscript editing. TT: data analysis, study design, manuscript editing. SW: study design, data collection, manuscript editing. RM: study design, data analysis, manuscript writing. All authors contributed to the article and approved the submitted version.