Introduction
Whilst droughts are natural phenomenon, their frequency and severity are
both increasing due to climate change (Mukherjee et al. 2018), driven by
reduced regional precipitation and increased global evaporation (Ault
2020, Dai 2011). Drought has devastating impacts on species and
ecosystems, causing increase in species mortality and extinction (Choat
et al. 2018, Harrison 2000), loss of biodiversity (Peterson et al. 2021,
Tilman and Elhaddi 1992), and declines in ecosystem function and
productivity (Atkinson et al. 2014, Ciais et al. 2005).
These pressures are occurring against a backdrop of increasing habitat
fragmentation, driven by a range of factors including the extraction of
resources, development of settlements, increasing transport links, and
proliferation of farming globally. Consequently, many populations are
increasingly isolated, impacting their risk of extinction (Crooks et al.
2017, Reed 2004). Corridors have been suggested as positive conservation
actions which can be used to reverse the negative effect of
fragmentation by promoting dispersal and maintaining gene flow, allowing
species to move or adapt in the face of environmental change. Previous
studies have shown that enhancing habitat connectivity by corridors can
reduce the likelihood of extinction and enhance species diversity in
patchy habitats (Chisholm et al. 2011, Damschen et al. 2019),
potentially buffering species loss against increasingly inhospitable
environments.
Extreme climatic events such as droughts have a devasting impact on the
viability of natural populations. Previous studies showed that droughts
reduced the survival rate of populations in habitat patches by inducing
ecological traps, a phenomenon that occurs when species settle in
maladaptive habitats with a poor habitat selection (Hale and Swearer
2016, Robertson and Hutto 2006). For example, Coho salmons
(Oncorhynchus kisutch ) that inhabit intermittent streams used
remaining pools as habitats when connectivity was lost, but during
drought years a lower survival rate was found in some pools than others
(Vander Vorste et al. 2020). Such phenomena have also been reported in
other ecosystems (Robertson and Hutto 2006). In this case, dispersal
might be key to maintain the dynamics of metapopulations and communities
(Hale and Swearer 2016, Hale et al. 2015, Wolfe et al. 2023), as
climatic extremes are likely to occur dynamically across time and space,
and the quality of habitats can deteriorate or recover when the climate
regime changes. Yet, there is relatively few data on the population
changes with temporarily and spatially dynamic drought stressors.
Corridor quality is one of the important physical properties determines
corridor effectiveness and dispersal success (Bennett et al. 1994, Habel
et al. 2020, Li et al. 2021). Specifically, high quality corridors can
promote dispersal, leading to greater movement rates to colonised
patches compared to poor quality corridors (Li et al. 2021). Hence
corridor quality may affect the stability and longevity of
corridor-connected metapopulations, by facilitating or impeding
movements between habitat patches. However, we currently lack empirical
evidence on how corridor quality promotes population persistence, and
how it interacts with the increasing severity and frequency of drought
stressors.
While historical data has successfully investigated the population
consequences of extreme drought events on habitat networks (e.g., Oliver
et al. 2013), quantitively measuring the effect of increasing drought
severity and frequency on population declines and – importantly – how
this stress might interact with corridors is difficult. However,
disentangling the effects of corridor quality on metapopulation
persistence in such natural systems (in the absence of high replication
and control treatments) is complex. Experimental microcosms provide one
alternative method to achieve this, as they allow landscape style
manipulations at an observable scale (Altermatt et al. 2015, Benton et
al. 2007). Indeed, such systems have previously been used to study the
effect of network modularity (e.g., Gilarranz et al. 2017), and the
impact of habitat configuration on metacommunities (e.g., Chisholm et
al. 2011, Wolfe et al. 2022).
Here we investigate the effect of habitat connectivity on the
persistence of metapopulations under drought stressors. We manipulated
both drought severity and the number of patches affected by drought in
landscapes connected by either good or poor quality corridors. We use 3D
printed “four-patch, four-corridor” microcosms containing the soil
Collembola Folsomia candida as experimental metapopulations and
use a fully factorial manipulation where we change the moisture of
corridors (good vs poor quality corridors), amount of water added to the
patches (severity of drought), and number of patches (from 1 to 4)
affected by reduced water availability (number of patches affected by
drought). Drought is a known abiotic stressor to F. candidapopulations, as they need a high moisture to survive and reproduce. By
monitoring the changes in population abundance in microcosms for a
relatively long-period (c.16 weeks, ~ 5 generations), we
can measure the effects of corridor quality interacting with drought
severity and number of drought patches on the time of extinction, the
maximum rate of population decline, and the variability of abundance
among patches.