Role of habitat heterogeneity and landscape connectivity in shaping gene flow and spatial population structure of a dominant rodent species in a tropical dry forest

Abstract

Evaluating how habitat heterogeneity and landscape connectivity influence finescale population processes of gene flow and spatial population structure is key for understanding animal dispersal. Liomys pictus is a heteromyid rodent that inhabits tropical dry deciduous and semideciduous forests, where it is a dominant, abundant species and a key element on the ecosystem as seed disperser. We evaluated how landscape features shaped the genetic structure and gene flow at a fine scale among L. pictus populations in a well-conserved natural environment. On the basis of L. pictus ecological characteristics (e.g. dominant species, female philopatry), we predicted that forested areas (irrespective of vegetation type) should facilitate dispersal if the forest vegetation acted as a kind of corridor, and that these patterns will be different between sexes. We tested these assumptions (104 individuals, six sampling localities, 14 microsatellite loci) after checking for discrete genetic groups using Bayesian clustering methods, by assessing the effect of landscape variables on patterns of gene flow combining Mantel and partial Mantel tests, least-cost path and circuit theories. L. pictus differentiated into genetic clusters delimited by clear landscape boundaries. Accordingly, resistance hypotheses results showed that precipitation and stream channels were the main environmental and landscape attributes influencing gene flow. Exchange of a majority of migrants was detected from the center into other sampling localities, indicating higher dispersal throughout deciduous and semideciduous forest corridors, whereas isolation by distance was found only for females. Our approach allowed us to elucidate animal environmental space use, identifying some of the landscape features linked to the species dispersal patterns, which can serve as a basis for the study and comparison with other tropical forests and species. It can also have conservation applications, for instance, preserving the forested corridors we identified as significant for dispersal can likely benefit other codistributed, threatened rodent species.

Functional connectivity and home range inferred at a microgeographic landscape genetics scale in a desert‐dwelling rodent

“…Most landscape genetics inference studies have assumed either a positive or negative linear relationship between landscape features and cost surfaces (Garroway, Bowman, & Wilson, 2011;Koen, Bowman, & Walpole, 2012), including examples with rodent populations (Chiappero et al, 2016;Howell et al, 2017;Mora et al, 2017;Ortiz et al, 2017), despite that nonlinear responses are expected to be more common (Marrotte & Bowman, 2017;. Here, we found nonlinear relationships relative to landscape resistance for both humidity and temperature, variables that influence genetic connectivity in different species like the northern quoll (Dasyurus hallucatus; Hohnen et al, 2016) and Liomys pictus (Garrido-Garduño et al, 2015). Temperature has been explicitly proposed as determinant for genetic connectivity in climate sensitive species, for example, the American pika (Ochotona princeps), a heat intolerant, and cool microclimate restricted species for which an increase in temperature adversely affects gene flow (Castillo, Epps, Davis, & Cushman, 2014).…”

“…According to our expectations, we identified that vegetation (Russo et al, 2016). Another example showed that forested areas in a tropical dry forest function as corridors for dispersal for the spiny pocket mice Liomys pictus, overcoming the potential limiting effect of roads across the landscape (Garrido-Garduño et al, 2015); forests also facilitate gene flow in chipmunks (Tamias striatus) inhabiting fragmented landscapes like agroecosystems (Kierepka et al, 2016). Moreover, D. merriami activity peaks exhibit a behavior tightly associated with predator avoidance (Daly, Behrends, Wilson, & Jacobs, 1992;Soltz-Herman & Valone, 2000),…”

“…In addition, rodents have been proposed as ideal for conducting LG research (Waits, Cushman, & Spear, 2015) because of their small body size, short generation times, and limited dispersal abilities. Accordingly, LG studies have focused on different rodent species living in a wide range of environments, including tropical dry forests (spiny pocket mouse, Liomys pictus; Garrido-Garduño, Téllez-Valdés, Manel, & Vázquez-Domínguez, 2015), savannas of South Africa (Natal multimammate mouse, Mastomys natalensis; Russo, Sole, Barbato, von Bramann, & Bruford, 2016), subantarctic forests and Patagonian steppes (Longtailed pygmy rice rat, Oligoryzomys longicaudatus; Ortiz et al, 2017), and even in urbanized areas (White-footed mouse, Peromyscus leucopus, Munshi-South, 2012;and Norway rats, Rattus norvegicus, Gardner-Santana et al, 2009). Interestingly, vegetation has been underlined as a variable facilitating gene flow in some of those systems, including the canopy cover across New York City for P. leucopus (Munshi-South, 2012), tropical dry forest corridors in L. pictus (Garrido-Garduño et al, 2015), and forest patches in chipmunks (Tamias striatus) inhabiting agroecosystems (Kierepka, Anderson, One of the most conspicuous desert rodents is kangaroo rats (genus Dipodomys, family Heteromyidae).…”

Functional connectivity in replicated urban landscapes in the land snail (Cornu aspersum)

“…In our study, LCP and CT performed equally well with the exponential distribution, and it was not possible to resolve competing assumptions (e.g., single or multiple optimal paths). Circuit‐theory‐based analysis is commonly highlighted as more informative than LCP about gene flow (Coulon et al., ; Garrido‐Garduño, Téllez‐Valdés, Manel, & Vázquez‐Domínguez, ; La Point, Gallery, Wikelski, & Kays, ; McRae et al., ), but comparisons have seldom been made in urban landscapes. Urban landscape characteristics could lead to different relative performance between LCP and CT‐based models.…”

Genetic population structure of the long‐tailed pygmy rice rat (Rodentia, Cricetidae) at different geographic scales in the Argentinean Patagonia

“…Regarding rodents, Garrido‐Garduño et al . () studied the genetic structure of the Painted spiny pocket mouse Liomys pictus and found that populations were clustered according to clear landscape boundaries, mainly defined by precipitation and stream channels. Furthermore, Gerlach & Musolf () found the same pattern for the bank vole Clethrionomys glareolus , where rivers and highways acted as strong barriers to gene flow.…”