Mismatch managed? Phenological phase extension as a strategy to manage phenological asynchrony in plant–animal mutualisms

Species‐specific shifts in phenology (timing of periodic life cycle events) are occurring with climate change and are already disrupting interactions within and among trophic levels. Phenological phase duration (e.g. beginning to end of flowering) and complementarity (patterns of nonoverlap), and their responses to changing conditions, will be important determinants of species' adaptive capacity to these shifts. Evidence indicates that extension of phenological duration of mutualistic partners could buffer negative impacts that occur with phenological shifts. Therefore, we suggest that techniques to extend the length of phenological duration will contribute to management of systems experiencing phenological asynchrony. Techniques of phenological phase extension discussed include the role of abiotic heterogeneity, genetic and species diversity, and alteration of population timing. We explore these approaches with the goal of creating a framework to build adaptive capacity and address phenological asynchrony in plant–animal mutualisms under climate change.

• Predictions of phenological asynchrony due to climate change call for novel conservation strategies.
• We propose extending phenological phase duration as one approach for buffering impacts of asynchrony.
• Techniques to extend the duration of plant or animal activity timing include utilizing abiotic heterogeneity, genetic and species diversity, and alteration of population timing.
• Existing biodiversity conservation techniques may have the potential to address mismatch concerns if put into the context of phenological shifts.
• We call on restoration ecologists to propose and test effectiveness of strategies to address mismatch concerns.

INTRODUCTION
Considering phenological shifts (italicized terms-see Box 1) in management decisions may be critical for conserving species interactions, mitigating invasions, and maintaining ecosystem functions and services (Elzinga et al. 2007). Species-specific shifts in phenology are occurring with climate change and are already disrupting interactions within and among trophic levels (e.g. Schmidt et al. 2016). Species vary in their responses to temperature and moisture changes, and one or more environmental cues may influence whether or not timing will shift with climate changes (Cleland et al. 2007). Phenological asynchrony between mutualistic species is expected to decrease fitness and yield population declines (van Asch et al. 2007;Rafferty et al. 2013). With predictions of climate-induced asynchrony, and no clear solutions or management principles available for practitioners, we were motivated to explore strategies that may build the capacity of a system to respond to phenological shifts. We propose extending phenological phase duration in mutualistic partners as a mechanism to build adaptive capacity, and discuss techniques to achieve this goal.
The likelihood and consequences of phenological asynchrony in mutualisms will depend on the level of specialization, seasonality and duration of interactions, and the intimacy of the relationship (i.e. symbiotic or free-living; Rafferty et al. 2015). Phenological asynchrony in mutualisms has been most studied in transportation mutualisms (e.g. seed dispersal [Warren et al. 2011]), but studies have also predicted the potential for asynchrony due to climate change in nutritional and protection mutualisms (Rafferty et al. 2015). Mutualistic species are under strong selection to remain in synchrony, and phenological asynchrony is therefore evolutionarily unstable (Renner & Zohner 2018). On ecological time scales, rapid anthropogenic climate change may increase asynchrony if species respond to different factors and not enough time is available for selection to offset these shifts. Where asymmetry in phenological shifts is occurring or expected to occur, reducing impacts of timing asynchrony may buy time for species to adapt.
Adaptive capacity is defined as the capability of organisms or systems to adjust to potential stress, take advantage of opportunities, or respond and mitigate negative impacts of environmental change (IPCC 2018). Improving and maintaining adaptive capacity provides an ecological buffer that protects the system from collapse when change occurs (Gunderson 2000). In this article we explore the potential for adaptive capacity of a community to phenological shifts to be improved by extending the phenological phase timing of mutualistic species to allow for adjustments and changes in species interactions, and community restructuring as necessary.
Lengthened duration of phenological activity may allow for adaptive capacity by supporting both animal and plant mutualists. Animal species require sufficient and abundant resources throughout their life cycles, and continual plant resources over the active season (e.g. flight, foraging, and nesting seasons) are needed for animal populations to maintain stability and function (Russo et al. 2013). If the timing of an animal species becomes out of sync with a plant mutualist resource, other plant species are essential to supplement its needs during the period of time when resources are unavailable (Waser & Real 1979). Plant species that bridge temporal gaps and extend resource timing may also allow for facilitation among plants by supporting mutualist animal populations (Moeller 2004). Recent evidence suggests that extended phenological timing can aid mutualisms (e.g. Frankie et al. 2013;Hindle et al. 2015;Mola & Williams 2018). Hence, increasing or maintaining availability of partner resources across time may support the survival of species in mutualistic interactions as the climate changes.
We propose that managing ecosystems in the face of mismatch will require implementing strategies to maintain and extend the duration of phenological activity. An extension in the duration of partner resources may allow for increased survival in species that are undergoing phenological shifts. This duration could be across a community for generalist mutualisms, with species at the beginning of the season starting earlier and those at the end of the season ending later to extend the total season-wide availability of the resource. Increased phenological phase duration may allow for adaptive capacity by maintaining natural patterns of overlap and buffering the impacts of timing shifts in the short term.
The focus here is on plant-animal mutualisms. However, building adaptive capacity to phenological shifts is not limited to mutualistic interactions. All interactions may experience mismatch with climate shifts, impacting species either positively or negatively depending on the type of interaction. For example, plant-herbivore relationships are dependent on the overlap between plant resources and herbivore timing (van Asch et al. 2007), and asynchrony may benefit plant species while negatively impacting herbivores, while overlap will benefit herbivores and other consumers. Longer phenological duration can benefit nonmutualistic partners (e.g. deer herbivory [Pettorelli et al. 2005]), and therefore extending phenological phase duration may be a strategy to aid in short-term survival or balancing of ecosystem dynamics for any interactions.

Techniques to Extend Duration of Phenological Activity
We propose extending the duration of phenological activity as a management possibility where phenological mismatch is a concern. Phenological activity may be extended via plant resource timing extension, animal partner timing extension, or providing supplemental partners (Fig. 1). We propose three techniques that we predict could lead to an increase in phenological phase durations at local and landscape scales: (1) diversifying species and genotypes; (2) utilizing microclimate heterogeneity; and (3) alteration of population timing. These techniques are based on the idea that complementarity (patterns of nonoverlap) in the timing of mutualistic partner resources will yield overall longer resource availability (Fig. 2). In addition, techniques may work synergistically to yield additional extension in timing. The aim should be to maintain the ecosystem's natural patterns of synchrony as much as possible in the face of climate change.

Box 1 Glossary of important concepts and definitions.
Adaptive capacity: capability of organisms or systems to adjust to potential stress, take advantage of opportunities, or respond to and mitigate negative impacts of environmental change.
Phenology: the timing of periodic life cycle stages of organisms.
Phase duration: time from start to end of a particular phenological phase (e.g. flowering period).
Phenological complementarity: complementary timing in species growth and reproductive timing (e.g. complementary flowering species flower at different times of the year). Complementarity is used here to describe patterns of nonoverlap in species of the same functional group or guild.
Phenological overlap: overlap in species growth and reproductive timing. Extent of overlap depends on amount of time both species are active simultaneously. Overlap in a mutualism occurs when a species or ecological phase is concurrent in time with its interacting partner.
Phenological mismatch/asynchrony: when a species or ecological phase is not concurrent in time with its interacting partner. Extent of asynchrony depends on amount of time both species are present or active in the absence of the other. A complete mismatch occurs when the phenological phases of mutualistic partners are entirely out of sync with each other.
Phenological shift: a change in the timing of life cycle stages, resulting in timing that is earlier or later in the year.
Technique 1: Diversifying Species and Genotypes. Introduction of species and genotypes, with complementary phenology, can supplement resources during periods of diminished availability that may be created by a phenological shift (Timberlake et al. 2019). Many restoration projects are designed with a diverse array of plant species with complementary traits and hence incorporating phenology into trait selection can extend the duration of resources across the year. In fact, at this time the only strategy that we know has been implemented to extend duration of fruiting and flowering resources is the selection of a palette of species with diverse timing (e.g. early and late flowering species; Fig. 2A) in restoration projects (Box 2). This practice may benefit a system in multiple ways, as it may serve to extend timing as well as increase functional redundancy and improve quality of resources (e.g. nutritional value) in a system. While this will be useful for generalist mutualisms, it may also support some specialized mutualisms if specialization occurs at the genus or family level, or on functionally similar species across clades (e.g. bats and Piper fruit (Marinho-Filho 1991); thrips and dipterocarps (Appanah 1993)).
A longer window of timing may also be achieved within species by diversifying genotypes. Genotypes with slightly different timing can complement each other, and yield an overall extended phenological duration (Smith et al. 2015). The extent of phenotypic plasticity can also vary by genotype (Pigliucci 2001). Therefore, managing for natural genotypic diversity in timing could extend phenological phases, via diversity in both fixed and plastic timing traits. Planting genotypes from diverse source locations in one place could yield an overall longer resource duration due to the complementarity of both early and late genotypes.
Genotypes with different phenologies may interbreed or adapt to have similar timing based on site conditions (Ware et al. 2019) potentially making the extension of timing shortlived. It may also be difficult to predict the exact timing of genotypes once moved to a new location due to phenotypic plasticity (Monty & Mahy 2009). However, this technique may still aid in increasing adaptive capacity, as the presence of different genotypes may allow for adaptive evolution, or may permit an interacting species to coexist long enough to adapt to new conditions (Millar et al. 2007). Evaluating risks will be important before implementing this technique, as swamping the population with nonadaptive genotypes may be a concern. However, in the case of an uncertain future, increasing the genetic diversity of local populations is likely to be beneficial (Millar et al. 2007).
Technique 2: Utilizing Microclimate Heterogeneity. Areas with different abiotic conditions within a patch and across a landscape may yield differences in phenological phase timing and duration (Fig. 2B). The nature of these areas, their spatial configuration, and scale at which they vary will determine phenological duration at the local and landscape levels. Heterogeneity in microclimates can yield lengthened duration by creating patterns of complementarity in the timing in both plant and animal activity. Abiotic gradients and habitat heterogeneity can impact the timing of both animal and plant species distributed across a site (Hindle et al. 2015;Olliff-Yang & Mesler 2018). Management to conserve, maintain, and restore abiotic heterogeneity, and increasing connectivity across heterogeneous landscapes could extend phenology at both local and landscape scales.
Altering conditions to create microclimate heterogeneity (e.g. watering, shade structures, and earth moving) could extend phenological phase duration for both animal and plant species. Microclimate characteristics that can affect phenological timing include landscape positioning like slope, aspect, hilltop, and valley bottom (Weiss et al. 1988), as well as soil moisture and canopy cover (Heinrich 1976). Creating microhabitat heterogeneity can be as straightforward as placing wind shields, which can extend activity timing in alpine environments (Fukuyo et al. 1998). Nesting habitat and positioning can also be manipulated to influence animal timing (e.g. moving bees in trap nests  [Forrest & Thomson 2011]). The scale of implementation of this technique will be dependent on the species, and microenvironments must be present within the average foraging range of the animal mutualist to be effective.
Increasing connectivity across microclimates may connect animal and plant mutualists with various timings, effectively increasing the duration of phenological activity on the landscape as a whole. Elevation gradients influence phenology for both plants and animals due to precipitation and temperature gradients (Forrest et al. 2010), and many montane animal species rely on moving to track differences in resource timing due to variation in snowmelt and spring vegetation onset across elevation (Pettorelli et al. 2005). Implementation of this technique would include providing linkages and corridors, enhancing habitat heterogeneity in closely adjacent locations, or otherwise facilitating the movement of organisms across abiotic gradients (Dunwiddie et al. 2009), with the goal of increasing the chances that suitable partners are within reach of one another at the right time.
Technique 3: Alteration of Population Timing. Management that increases timing heterogeneity within populations may be used to extend the phenological phase timing of both plant and animal species, as complementarity in timing between individuals in a population will yield extended timing. Techniques to directly alter population timing may include direct manipulation of growth timing, as well as altering biotic and abiotic conditions that affect population timing (Fig. 2C).
Manipulating seasonal growth (e.g. via hormone or growth initiation treatments) can alter and extend phenological phase duration as individuals with different growth timing will experience different climactic conditions, which may lead to a variety of timing in one location. Planting on various dates could induce timing complementarity and yield extended duration of phenological activity in one growing season (Iannucci et al. 2008). Effectiveness of directly manipulating seasonal growth on lengthening phenological phase timing would require species that reproduce in the first growing season, and that are not strongly dependent on photoperiod cues. In addition, it is not likely that the effects of such treatments will last for multiple years without continued management.
Heterogeneity in both biotic and abiotic factors may foster a variety of phenological timing, leading to complementarity between patches, and an overall extended season at landscape scale. Grasslands with heterogeneous management practices create a mosaic of timing, supporting successive flowering Box 2 Point Blue's Climate Smart Planting design tool.
Point Blue Conservation Science in California was an early developer and adopter of "Climate Smart" restoration practices. A planting design tool created by Point Blue Conservation Science allows practitioners to (Fig. IA) select candidate plant species for restoration planting and then (Fig. IB) view how flowering and fruiting resources will likely be distributed through time once plants are established (Point Blue Conservation Science 2019). This allows project designers to select complementary flowering and fruiting species to provide resources across the full season. Areas restored using these metrics may reduce the impacts of phenological mismatch by supplementing plant resources for generalist animal species across the season.  (Kubo et al. 2009). Competition and disturbance can affect flowering time and duration (Rathcke & Lacey 1985), and density can influence both animal and plant reproductive timing (Schmitt 1983;Avila et al. 2016). Burning can lengthen flowering time in fire-adapted landscapes (Mola & Williams 2018), and heterogeneous fire severity and intensity ("pyrodiversity") across a landscape may further lengthen the flowering season (Tunes et al. 2017). Finally, directly planting or placing nesting habitat across heterogeneous microclimates can also extend timing of populations (see Technique 2 -Utilizing Microclimate Heterogeneity).

Concluding Remarks and Future Perspectives
One of the most conspicuous responses of organisms to climate change has been shifts in timing of phenological events (Menzel et al. 2006). These shifts are already causing interacting species to become less synchronous in time than they have been. It will be important to keep the timing of interactions in mind while assessing climate risks and planning for the future, as this will help us envision and plan for instabilities (Russo et al. 2013). Considering novel ways to buffer the impact of climate change on ecosystems is critical for management success. Extending the phenological phase duration of mutualistic partners may be one way to buffer the impacts of timing shifts on asynchronous mutualisms.
It is important to keep potential trade-offs in mind. For example, techniques to extend phenological overlap may reduce the strength of selection that an asynchrony would cause. While increasing short-term survival may buy time for adaptive evolution, weaker selection pressures would slow the rate of adaptive response. On the other hand, selection cannot act if either partner in the mutualism is extirpated due to rapidly changing climatic conditions. Decreased overlap is only one of multiple factors that climate changes will impact, and short-term survival may depend on other conflicting factors, both biotic and abiotic, that have stronger impacts and undermine the effectiveness of techniques discussed here (Visser & Gienapp 2019). In addition, extending the duration of an activity may increase the overlap in time with undesirable interactions (e.g. Douglas fir trees and spruce budworm [Chen et al. 2003]).
It is also possible that invasive and weedy species may be facilitated or hindered with extended phenological phase timing, or in the implementation of techniques to extend timing. Any newly introduced species would yield an invasion risk, and introducing or changing disturbance may create opportunities for invasion (Hobbs 1989). Invasive species are typically more phenotypically plastic than noninvasives (Davidson et al. 2011), which may allow for the flexibility to capitalize on empty niches in time (Wolkovich & Cleland 2011). Our strategy to extend phenology may be an effective way to fill empty niches in time with native species rather than allowing invaders to take advantage of timing gaps. As always, focused study of individual systems may be needed to determine costs and benefits of phenological restoration and conservation strategies.
The selection of plant species and genotypes with complementary phenology lies at the heart of farm and garden design, where humans act as consumers seeking to enjoy a colorful display or fresh foods throughout the season. Not surprisingly, gardens can become important resources for animal pollinators and dispersers across the urban to rural gradient, and enhanced resource availability through the year has contributed to species range expansions (Greig et al. 2017). The line between habitat restoration and "artificial" gardens may become increasingly blurred in the face of climate change, as new species are selected that can manage novel conditions and hence contribute to biodiversity conservation (Dunwiddie & Rogers 2017).
Techniques to maintain and extend resource duration on the landscape are in line with many already used to support adaptive capacity to climate change. Increasing genetic and species diversity in restoration and forestry practices is a top recommendation for conservation of biodiversity under climate change (Heller & Zavaleta 2009). Expanding the number of seed source locations to boost genetic diversity is one strategy suggested for conserving plant populations in an uncertain future (Millar et al. 2007). Restoration projects alter population density and disturbance via seeding density, grazing, mowing, and prescribed burning (Stromberg et al. 2007). Utilizing diverse topography and microtopography on the landscape has also been implemented to support establishment of plant species and aid ecosystem function (Biederman & Whisenant 2011). Incorporating timing as a dimension of conservation and restoration planning can further achieve goals by aiming for adaptive capacity to climateinduced asynchrony. The strategies outlined here may also reduce asynchrony due to factors other than climate change, such as effects of landscape simplification and habitat loss.
The extent of phenological mismatch that will occur with climate change and the resulting impacts on population demography is unclear, as there is still a paucity of evidence for the effects of asynchrony on population fitness (Renner & Zohner 2018). It is likely that the extent and impact of phenological phase shifts and any need for management intervention will be context dependent. Continued research is needed on techniques to extend phenology, the risks and impacts of different techniques, and the relative increase in adaptive capacity with extended phenology in different systems. We consider extending phenological phase duration as a tool to add to the arsenal of strategies being developed with the aim of mitigating climate change impacts. We hope that the ideas proposed here will inspire continued discussion and research on creative strategies to mitigate the impacts of phenological mismatch with climate change.