Received: 12 November 2019 Revised: 24 May 2020 Accepted: 26 May 2020 Published on: 18 August 2020 DOI: 10.1002/1438-390X.12060 ORIGINAL ARTICLE Outbreaking forest insect drives phase synchrony among sympatric folivores: Exploring potential mechanisms Andrew M. Liebhold1,2 | Christer Björkman3 | Alain Roques4 | Ottar N. Bjørnstad5 | Maartje J. Klapwijk3 1USDA Forest Service Northern Research Station, Morgantown, West Virginia Abstract 2Czech University of Life Sciences Prague, We explore a common feature of insect population dynamics, interspecific syn- Faculty of Forestry and Wood Sciences, chrony, which refers to synchrony in population dynamics among sympatric Suchdol, Prague, Czech Republic populations of different species. Such synchrony can arise via several possible 3Department of Ecology, Swedish mechanisms, including shared environmental effects and shared trophic inter- University of Agricultural Sciences, Uppsala, Sweden actions, but distinguishing the relative importance among different mecha- 4INRAE, UR 0633, Zoologie Forestière, nisms can be challenging. We analyze interannual time series of population Orléans, France densities of the larch budmoth, Zeiraphera griseana (Lepidoptera: Tortricidae), 5 Departments of Entomology and Biology, along with six sympatric larch-feeding folivores from a site in the European Pennsylvania State University, University Park, Pennsylvania Alps 1952–1979. These species include five lepidopterans, Exapate duratella, Ptycholomoides aeriferana, Spilonota laricana, Epirrita autumnata and Tel- Correspondence eiodes saltuum, and one hymenopteran sawfly Pristiphora laricis. We docu- Andrew M. Liebhold, USDA Forest Service Northern Research Station, ment that the highly regular oscillatory behavior (period 9–10 years) of Z. 180 Canfield Street, Morgantown, WV griseana populations is similarly evident in the dynamics of most of the sym- 26505. Email: [email protected] patric folivores. We also find that all of the sympatric species are phase syn- chronized with Z. griseana populations with half of the sympatric species Funding information exhibiting nonlagged phase synchrony and three of the species exhibiting The Czech Operational Programme – Research, Development and Education, 2 5 year lags behind Z. griseana populations. We adapt a previously developed Grant/Award Number: tritrophic model of Z. griseana dynamics to explore possible mechanisms CZ.02.1.01/0.0/0.0/16_019/0000803; responsible for observed phase synchronization. Results suggest that either Swedish University of Agricultural Sciences or Sveriges Lantbruksuniversitet; shared stochastic influences (e.g., weather) or shared parasitoid impacts are USDA Forest Service likely causes of nonlagged phase synchronization. The model further indicates that observed patterns of lagged phase synchronization are most likely caused by either shared delayed induced host plant defenses or direct density- dependent effects shared with Z. griseana. KEYWORDS folivore, interspecific synchrony, outbreak, population dynamics, trophic 1 | INTRODUCTION Other recent reports and predictions of increased damage from insect pests as a consequence of climate change are Recent reports of dramatic decreases in insect numbers equally alarming (Estay, Lima, & Labra, 2009; Weed, (Hallmann et al., 2017; Leather, 2018) have increased the Ayres, & Hicke, 2013). These apparently conflicting awareness of the importance of insects in ecosystems. reports may seem paradoxical but highlight the lack of 372 © 2020 The Society of Population Ecology wileyonlinelibrary.com/journal/pope Population Ecology. 2020;62:372–384. LIEBHOLD ET AL. 373 knowledge on the population dynamics of communities (Liebhold, Koenig, & Bjørnstad, 2004). Like interspecific of insects. There is a strong need to better understand synchrony, several different processes are capable of pro- connections, not only between species at different trophic ducing spatial synchrony; these include stochastic influ- levels, but also among species at the same trophic level. ences (e.g., weather) which are synchronous across space For example, several processes are known to produce (the “Moran effect”), dispersal of individuals among periodic behavior in insect folivore populations, but little populations and dispersal of predators among habitat pat- is known on how these processes may interact among ches (Liebhold et al., 2004; Walter et al., 2017). Unfortu- sympatric folivore populations. nately, all three of these mechanisms are capable of The ability of trophic interactions to produce periodic producing identical patterns of synchrony and therefore behavior is well known (Elton & Nicholson, 1942; Varley, it is characteristically difficult to dissect the cause of spa- Gradwell, & Hassell, 1974). For example, in the classic tial synchrony from spatiotemporal patterns alone. Anal- Lotka–Volterra model, both predator and prey ysis of interspecific synchrony has received less attention populations cycle with peaks in predator populations lag- and it is not clear if it is similarly difficult to identify the ging one-quarter-cycle length behind that of prey drivers of interspecific synchrony. populations (Bulmer, 1976). In nature, there is a plethora Here, we analyze patterns of interspecific synchrony of evidence of such oscillatory dynamics in predator and in a highly periodic system. Specifically, we quantify peri- parasitoid populations linked to oscillations of their hosts odicity and phase dependency among populations of six (Bayliss & Choquenot, 2002). One ubiquitous characteris- foliage-feeding insect species feeding sympatrically with tic of such predator–prey cycles is the existence of tempo- the larch budmoth, Zeiraphera griseana, in the European ral lags between cycles of predators and prey. Alps. Population oscillations of this insect are widely rec- While linkages between oscillating prey and predator ognized as one of the most regular of all animal species populations are well-studied, less is known about mecha- (Baltensweiler, Benz, Bovey, & Delucchi, 1977; Varley nisms driving linkages in the dynamics of sympatric et al., 1974). During population peaks, populations in consumer populations. Populations of consumer species optimal habitats reach very high densities, often more may track fluctuating patterns in shared food resources, than 300 larvae per kg of foliage (ca. 120,000 larvae per contributing to synchrony in their dynamics. For exam- tree), causing massive defoliation of European larch ple, several studies have documented that fluctuating (Larix decidua). The very regular oscillations in Z. abundance of shared seed or insect populations can drive griseana populations are hypothesized to result from a synchrony among populations of seed-eating or insectivo- combination of delayed density-dependent response by rous birds (Bock & Lepthien, 1976; Jones, Doran, & hosts and by parasitoids (Turchin 2003, Turchin Holmes, 2003). However, many consumer populations et al., 2003). The year following defoliation, host larches exhibit oscillatory dynamics, and little is known about produce needles shorter in length and larvae feeding on how cycles among populations of different consumer spe- this foliage perform more poorly resulting in decreased cies are related. fecundity in females (Benz, 1974). Moreover, needle flush One group of consumer species for which interspe- is delayed relative to the timing of egg hatch, causing cific synchrony is known to commonly occur is in extensive mortality in early-emerging budmoth larvae foliage-feeding forest insects (Kawatsu, Yamanaka, (Benz, 1974). Sympatric larch folivores have never been Patoèka, & Liebhold, 2019; Klapwijk et al., 2018; Miller & observed to reach larval densities similar to those of Z. Epstein, 1986; Myers, 1998; Raimondo, Liebhold, griseana populations but may peak at around 18 larvae Strazanac, & Butler, 2004; Raimondo, Turcáni, Patoèka, & per kg of larch foliage (Baltensweiler, 1991). While noth- Liebhold, 2004). Several explanations have been proposed ing is known about the induced effect of Z. griseana defo- to explain synchrony among these folivore populations. liation on foliage quality for other insects, it seems likely Shared influences of interannual variability in weather is that changes in the physical and chemical property of one of these proposed mechanisms (Miller & Epstein, larch foliage would also adversely affect sympatric 1986; Myers, 1998). Others have suggested synchroniza- populations of other insect species feeding on foliage. tion via shared trophic interactions, such as predators Thus, it is possible that Z. griseana exerts a synchronizing and parasitoids (Marcström, Kenward, & Engren, 1988; influence on sympatric larch-feeding folivores via direct Raimondo, Turcáni, et al., 2004) or shared host plants competition for foliage, induced changes in foliage qual- (Baltensweiler, 1991; Klapwijk et al., 2018). ity or from mutual numerical interactions of shared para- The phenomenon of interspecific synchrony is in stioids species. We apply a model to simulate these some ways analogous to the phenomenon of spatial interactions and compare patterns of phase synchroniza- synchrony, which refers to synchrony in the dynamics tion with observed patterns to infer underlying mecha- of spatially disjunct populations of the same species nisms driving interspecific synchrony. 374 LIEBHOLD ET AL. 2 | METHODS (ACFs) and wavelet spectra for each of the seven species using the “acf” function in base R and the “biwavelet” R Time