Research Article |
Corresponding author: Mark Hassall ( mhassall49@googlemail.com ) Academic editor: Katalin Szlavecz
© 2018 Mark Hassall, Anna Moss, Bernice Dixie, James J. Gilroy.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
Hassall M, Moss A, Dixie B, Gilroy JJ (2018) Interspecific variation in responses to microclimate by terrestrial isopods: implications in relation to climate change. In: Hornung E, Taiti S, Szlavecz K (Eds) Isopods in a Changing World. ZooKeys 801: 5-24. https://doi.org/10.3897/zookeys.801.24934
|
The importance of considering species-specific biotic interactions when predicting feedbacks between the effects of climate change and ecosystem functions is becoming widely recognised. The responses of soil animals to predicted changes in global climate could potentially have far-reaching consequences for fluxes of soil carbon, including climatic feedbacks resulting from increased emissions of carbon dioxide from soils. The responses of soil animals to different microclimates can be summarised as norms of reaction, in order to compare phenotypic differences in traits along environmental gradients. Thermal and moisture reaction norms for physiological, behavioural and life history traits of species of terrestrial isopods differing in their morphological adaptations for reducing water loss are presented. Gradients of moisture reaction norms for respiratory rates and thermal reaction norms for water loss, for a species from the littoral zone were steeper than those for species from mesic environments. Those for mesic species were steeper than for those from xeric habitats. Within mesic species, gradients of thermal reaction norms for aggregation were steeper for Oniscus asellus than for Porcellio scaber or Armadillium vulgare, and moisture reaction norms for sheltering and feeding behaviours were steeper for Philoscia muscorum than for either P. scaber or A. vulgare. These differences reflect differences in body shape, permeability of the cuticle, and development of pleopodal lungs. The implications of differences between different species of soil animals in response to microclimate on the possible influence of the soil fauna on soil carbon dynamics under future climates are discussed. In conclusion a modelling approach to bridging the inter-disciplinary gap between carbon cycling and the biology of soil animals is recommended.
Aggregation, CO2 emissions from soils, feeding behaviour, future rainfall patterns, life history traits, norms of reaction, response curves, soil animals
In this review we draw attention to an important contribution that soil biologists, in particular those who study the biology of terrestrial isopods, can potentially make to the current debate as to how global climate change may influence components of the global carbon cycle. Currently the greatest uncertainty in modelling the global carbon cycle is not the fluxes across the ocean atmosphere interface or fluxes relating to net primary production but in modelling carbon fluxes within the soil. Globally ten times more carbon dioxide is emitted from soils than from all anthropogenic sources combined (
Carbon dioxide emissions from soils are mediated predominately by microbial metabolism (Fig.
The possible consequences of this positive feedback cycle could potentially be reduced if soil moisture were to decrease as a result of changes in rainfall patterns (Fig.
The microbially mediated emissions of carbon dioxide from soils are strongly influenced by soil animals acting as key system regulators (Fig.
A conceptual diagram illustrating some of the pathways by which changes in global climate could potentially impact on rates of carbon dioxide emissions from soils. Both changes in temperature and in the levels and patterns of rainfall have strong direct effects on the metabolism of bacteria and fungi but their ecology and metabolism are also regulated by the extent to which they are stimulated by soil animals. Both functional (e.g., behavioural and physiological) responses and numerical (both life history and population) responses of soil animals are affected by their microclimate. This is in turn affected by larger scale changes in temperature and rainfall. Therefore, as well as their direct effect on microbial metabolism, these climatic variables have a strong indirect effect by influencing the behavioural, physiological, life history, and population processes of soil animals such as isopods.
It is becoming increasingly apparent that the importance of interspecific differences in key traits not only leads to different species having different functional roles (
Traits of many soil animals are sensitive to relative humidity, making the soil fauna vulnerable to their activities being curtailed by changes in soil moisture (
In this review we examine the extent to which traits of different species of terrestrial isopods known to have different strategies for reducing water loss in the terrestrial environment, respond to temperature and relative humidity. Phenotypic responses to a gradient in the environment can be summarised as norms of reaction. A reaction norm is the set of phenotypes produced in a range of environments (
We investigated differences in reaction norms of physiological, behavioural, and life history traits in response to differences in temperature and relative humidity for a range of isopod species chosen to reflect differences in their eco-morphology (
A schematic representation of a typical thermal response curve for enzymes (simplified from
Case study 1. Here we compare differences in aggregating behaviour of three species of isopod as described by
Differences in the degree of aggregation, at 90% relative humidity, between three species of isopods differing in their adaptations to reduce water loss in the terrestrial environment, are shown in Fig.
The temperature response curve of P. scaber (Fig.
Armadillidium vulgare shows a broadly similar pattern of response (Fig.
The temperature response curve for aggregation of O. asellus (Fig.
Overall these results show that these three species have substantially different patterns of thermal response curves for aggregating behaviours, reflecting differences in their morphological adaptations to terrestrial life.
Aggregation of isopod species differing in desiccation resistance at different temperatures. Mean ± 1 SE aggregation indices (variance:mean ratio) at 90% relative humidity. a P. scaber (F 4,249 = 3.76, p < 0.01) b A. vulgare (F 4,249 = 1.97, P < 0.01) c O. asellus (F 4, 249 = 12.22, P < 0.001) d thermal reaction norms for aggregation expressed as quadratic response curves for: P. scaber (dashed line): y = -11.519 + 1.526× - 0.04×2; A. vulgare (solid line): y = -3.534 + 0.574× 0.016×2; O. asellus (dotted line): y = -5.890 + 0.814× – 0.018×2.
Many terrestrial isopods aggregate in shelter sites, particularly during the day, often under stones or pieces of wood where moisture from the soil maintains a more favourable relative humidity than in more open sites. They then emerge to forage at night when temperatures are lower and relative humidity is higher. Sheltering behaviour is thus of central importance in reducing mortality due to desiccation while also reducing the risk of being eaten by diurnal predators, such as insectivorous birds.
Case study 2. Effects of substrate moisture content and relative humidity on sheltering and feeding behaviour were investigated by
Eight replicate mesocosms were used for each rainfall treatment, with six identical control boxes for monitoring substrate moisture content. Each mesocosm contained twenty individuals of each species. The behaviours of all animals over a 72-hour period were classified according to their location in the mesocosm and activity categories, including sheltering and presence in the feeding area. It was not possible to observe movements of mouthparts so time spent in the feeding area was assumed to be proportional to time spent feeding (
The results in Fig.
Ph. muscorum also had the steepest gradient for its moisture reaction norm for feeding of 0.55 (Fig.
Overall Ph. muscorum showed significantly steeper moisture reaction norms than did either of the other species just as O. asellus, which similarly lacks pleopodal lungs, had a steeper gradient for its thermal reaction norm for aggregation than did either P. scaber or A. vulgare.
Moisture reaction norms for a) sheltering and b) feeding behaviours with changing sand moisture content (time spent in behaviour as percentages of total observed behaviours). Lines represent linear regression models: A. vulgare (solid line) (sheltering: y = 95.24 – 1.05×; feeding: y = 0.32 + 0.03), P. scaber (dashed line) (sheltering: y = 96.72 – 0.19×; feeding: y= 0.22 + 0.05×), Ph. muscorum (dotted line) (sheltering: y = 93.14 – 2.16×; feeding: y = 0.36 + 0.55x). Further regression statistics and number of observations (N), are given in Table
Regression statistics for moisture reaction norms for sheltering and feeding behaviours of three species of isopods, Armadillidium vulgare, Porcellio scaber, and Philoscia muscorum, in laboratory arenas. Abbreviations: N: number of observations; a and b: parameters of regression equation; †: significant differences at P < 0.001(t test).
Regression statistics | Sheltering | Feeding | |||||
N | a | b | t | a | b | t | |
A. vulgare | 1648 | 95.24 | -1.05 | -21.84† | 0.32 | 0.03 | 3.77† |
P. muscorum | 824 | 93.14 | -2.16 | -25.07† | 0.36 | 0.55 | 15.32† |
P. scaber | 1648 | 96.72 | -0.19 | -8.39† | 0.22 | 0.05 | 6.84† |
Comparison of gradients | Sheltering | Feeding | |||||
A. vulgare vs. P. muscorum | 18.99† | -30.19† | |||||
A. vulgare vs. P. scaber | -25.08† | -3.96† | |||||
P. muscorum vs. P. scaber | -44.43† | 28.36† |
Case study 3. Both thermal and moisture reaction norms for the key life history traits of growth and mortality were compared for O. asellus and Porcellio dilatatus Brandt, 1833 by
The 2×2 factorial experimental design for investigating both relative growth rates and mortality rates permits comparison of thermal and moisture reaction norms simultaneously. Growth rates of O. asellus increased significantly at 5 °C higher temperatures under the drier, 70% relative humidity conditions but did not grow significantly faster at the higher temperatures in the moister, 90% humidity (Table
Mortality rates (Table
Overall, 20% lower relative humidity resulted in steeper moisture reaction norms for mortality rates than for thermal reaction norms resulting from a 5 °C rise in temperature, with O. asellus being more susceptible to drier conditions than P. dilatatus.
Relative growth rates (mg g-1 day-1) of Oniscus asellus and Porcellio dilatatus. Three way ANOVA: temperature F1,72 = 5.15, P = 0.026; humidity F1,72 = 88.62, P < 0.001; species F1,72 = 41.53, P < 0.001. Reaction norms are derived from the differences in response to the two temperature and two humidity conditions.
Temperature °C | Relative humidity % | Moisture reaction norms | ||
---|---|---|---|---|
90% | 70% | |||
O. asellus | 13.5 °C | 34 ± 6 | 2 ± 0.1 | -32 |
18.5 °C | 32 ± 2.5 | 17 ± 9 | -15 | |
Thermal reaction norms | -2 | 15 | ||
P. dilatatus | 13.5 °C | 66 ± 0.5 | 15 ± 4 | -51 |
18.5 °C | 77 ± 8 | 28 ± 4 | -49 | |
Thermal reaction norms | 11 | 13 |
Mortality rates (numbers dying container -1 7days-1) of O. asellus and P. dilatatus. Mann Whitney U test for O. asellus: temperature NS, humidity U = 1851 P < 0.001; for P. dilatatus: temperature U = 2754, P = 0.016, humidity U = 2277, P < 0.001. Reaction norms are derived from the differences in response to the two temperature and two humidity conditions.
Temperature °C | Relative humidity % | Moisture reaction norms | ||
90% | 70% | |||
O. asellus | 13.5 °C | 0 | 0.73 ± 0.18 | 0.73 |
18.5 °C | 0 | 0.58 ± 0.12 | 0.58 | |
Thermal reaction norms | 0 | 0.46 | ||
P. dilatatus | 13.5 °C | 0.09 ± 0.04 | 0.78 ± 0.16 | 0.69 |
18.5 °C | 0.03 ± 0.03 | 0.18 ± 0.08 | 0.15 | |
Thermal reaction norms | 0.06 | 0.60 |
One of the drivers of differences between species of isopods in their behavioural and life history responses to differences in temperature and moisture is their different physiological adaptations to the terrestrial environment. In this section we compare responses of terrestrial isopod species, occurring in a wide range of biomes representing a gradient of moisture conditions.
Moisture reaction norms for differences in respiratory rate for O. asellus, P. scaber, and A. vulgare are shown in Fig.
This interspecific comparison is extended for species from a wider range of habitats for thermal reaction norms for the physiological process of water loss in Fig.
The range of morphological adaptations in body shape, respiratory surfaces of the pleopods and cuticle of isopods have been comprehensively described (
Gradients of moisture reaction norms for respiration of isopods differing in their resistance to desiccation. Reaction norms over the range 50–100% relative humidity for respiratory rates measured as rates of oxygen uptake (mm3 mm-2 body surface h-1) (
Similarly, gradients of thermal reaction norms for water loss were substantially higher for L. oceanica than for the more fully terrestrial species O. asellus and Ph. muscorum, both of which had steeper gradients for their thermal reaction norms than members of either of the Porcellio species or A. vulgare (Fig.
Thermal reaction norm gradients for evaporation rate (water loss) for isopods from biomes differing in availability of moisture. Evaporation rate (g g-1 h-1× 10-2) standardised to a temperature range of 3.5 °C (from
Given that these patterns of inter-specific differences in physiological traits for both thermal and moisture reaction norms reflect differences in morphology (
Sheltering behaviour is another tactic evolved in isopods to help reduce water loss by taking refuge in more humid shelter sites. Again, Ph. muscorum had a significantly higher gradient for its sheltering moisture reaction norm than either P. scaber or A. vulgare. Similarly, feeding behaviour moisture reaction norms of Ph. muscorum decreased more steeply than for either of the other two species. Thus within three species, all typical in mesic environments, there was a consistent trend in all of these behavioural traits, with species without pleopodal lungs being more sensitive to changes in microclimate than species better adapted to resist desiccation. This again indicates that there are significant inter-specific differences which suggest those species least adapted to the terrestrial environment might be more susceptible to potential changes in micro-climate resulting from changes in future patterns of rainfall.
It would be logical to predict that such a trend might also apply to growth rates but that was not supported by the data because thermal reaction norms were very similar for O. asellus and P. dilatatus while the gradient for moisture reaction norms at both temperatures were higher for P. dilatatus. P. dilatatus had higher growth rates under all four combinations of temperature and relative humidity, possibly because the microclimate actually experienced by P. dilatatus was moister than that experienced by O. asellus due to P. dilatatus behavioural trait of burrowing into the sand substrates thereby being subject to a higher relative humidity in its immediate microclimate.
These life history traits of growth and survivorship are important correlates of fitness and have a very strong influence on the population dynamics of isopods. Differences in the gradients of reaction norms of life history traits could thus result in differences in the way abundances of these species might change under future climates.
An important conclusion of this overview is that, due to inter-specific differences in morphological, physiological, behavioural and life history traits, different species of isopods are likely to respond very differently to predicted changes in global climate. Interspecific differences in response to changes in temperature (
If the abundance of some species declines, it is not yet known whether other species, less sensitive to changes in microclimate, will respond by expanding their realised niches. It is known that interspecific competition for high quality foods between different species of isopods does occur in the field (
The importance of species-specific responses to climate change has been highlighted for Collembola by
Our understanding of the global carbon cycle is predominantly encapsulated in models. Our understanding of terrestrial isopod biology, in contrast, is mostly based on results of empirical studies. There is a very important need to bridge the gap between these two contrasting approaches and methodologies in these completely different disciplines. Considering the whole series of symposium volumes on the Biology of Terrestrial Isopods from 1984 until 2018, models may qualify as an endangered species.
We know that carbon dynamics in the soil is the least well understood part of the global carbon cycle. Whilst, it is extremely complicated, models by definition are simplifications of reality. Models can never represent the full complexity of the real world, that is not their function, but what they have are both holistic and heuristic properties. Holistic in that it should be easier to appreciate emergent properties of a system as a whole from a model of it, rather than from detailed studies of individual components. Heuristic in that models should generate testable predictions that cannot be made on the basis of studying individual components in isolation.
Terrestrial isopods are soil animals about which we have a wealth of knowledge based on empirical studies on their anatomy, behaviour, physiology, life history, and ecology. What we do not have is integrative models of different aspects of their biology that can then be interfaced with those of other soil animals and ultimately with those of micro-organisms. This integrative approach would significantly contribute to our understanding of how global climate change will affect the soil component of the global carbon cycle.
Models do not necessarily have to be numerical, at least initially. A topographic map is a model of a landscape. At a glance, that piece of paper only tens of cm wide, shows us where hills, mountains, rivers and bridges are in landscapes at much larger scales. By using a map we can predict how best to travel from A to B. What is very urgently needed in soil biology is some comparably simplified maps of the interactions involving the soil animals under our feet. A starting place could be for the context of many future studies to be introduced using conceptual box and arrow models of the system under study (Fig.
The next stage could then be to parameterise these models using the extensive empirical data available in the literature, making assumptions we know to be simplistic but which enable us to make further quantitative predictions. Progressively testing these predictions experimentally to validate the models further could then increase our understanding of the system.
A problem will always be integrating models representing different levels of organisation which, from previous studies, range from molecules to communities. These may initially appear to be measured in different currencies but ultimately, while we are working within the paradigm of neo-Darwinian evolutionary theory, the answer to this problem may be in using fitness, or at least fitness correlates, to equate different traits at the individual level. Moving from the individual level to the ecosystem function level could then at least be based on a sound theoretical foundation.
The “functional traits” approach (
MH is very grateful to Stephen Sutton for initiating, stimulating, and encouraging his interest in the biology of terrestrial isopods. We also thank Martin Zimmer for suggesting an earlier version of this paper, NERC for a studentship for AM, and to the guest editors of this volume for inviting us to contribute to these symposium proceedings.