Species distribution models, use both geographical and environmental space to predict a species realised environmental niche. Below we explore what these terms mean.

Geographical Space

Species occurrences are usually plotted on a map, however it can also be visualised in a graph, where the occurrence records are defined by the values of the longitudinal coordinates on the x-axis and the latitudinal coordinates on the y-axis.

Environmental Space

Environmental space refers to the environmental niche that a species occupies. A species’ environmental niche is the set of environmental conditions in which a species can survive and persist. This concept was defined by the ecologist George Hutchinson in 1957.

Looking at only one environmental condition, and plotting the probability of a species to occur in relation to that condition, the result could for example be a bell-shaped curve. This curve shows the optimum value of the environmental variable for a species to survive. For example, if the species is affected by temperature, the probability of occurrence is highest at the optimum temperature, with decreasing probabilities towards the minimum and maximum values that the species can tolerate.

Add another environmental condition that is relevant for the species, and a two-dimensional graph is created, with one environmental variable on each axis. Each of these variables has its own optimum curve. The model will find the area where the combination of both variables results in the most optimum conditions for a species to occur. The centre of the circle represents the combination of environmental variables with the highest suitability for a species to survive. The gradient colour visualises a decrease in suitability of the environment from the centre of the circle outwards. 

Tolerances for three different environmental conditions can be plotted in three-dimensional space in which the suitability of the environmental conditions is then represented by a sphere or a cube of which the edges correspond to the minimum and maximum tolerable values of each environmental condition. However, a realistic species distribution model takes into account all environmental variables that are relevant for a species, and this results in an n-dimensional hyperspace that defines the environmental space of a species. 

Environmental Niche

Within the environmental niche concept, there is a distinction between the fundamental and realized niche of a species. The fundamental niche of the species refers to the environmental conditions where a species can occur, and assumes that the species only occurs there and nowhere else. However, a species is not necessarily found in every location that has suitable environmental conditions. This is because the distribution of a species is also influenced by biotic factors and can thus be limited by the presence of a predator, or the absence of food. The area that is actually occupied by a species is referred to as the realized niche. In geographical space, we refer to the realized niche as a species’ actual distribution, whereas the fundamental niche is referred to as its potential distribution.  

Two important factors that might affect the environmental niche of a species are:

  1. Source-sink dynamics: individuals of a species might move from the area of the fundamental niche, which is the source, to sink areas outside the fundamental niche. This is likely to occur for species with high dispersal abilities. For example, if the fundamental niche of a migrating species is defined based on the environmental conditions of areas where the animals breed, it is likely that this does not include areas where the species moves to outside the breeding season.
  2. Dispersal barriers: geographic features such as mountains, rivers or oceans might prevent a species from occurring in areas that have suitable conditions, just because the species cannot reach those areas.

Environmental and Geographical Space Models

Most algorithms, are designed to link the places of occurrences to the environmental conditions of those places, and thus function in environmental space. These models follow three steps:

  1. Take the occurrence data out of the geographical space.
  2. Use the data* from step1 to calibrate the model in the environmental space.
  3. Project the model back into geographical space.

Geographic models, however, are not based on the environmental conditions of occurrence locations, and as such the second step is skipped. 

Once you have modelled your potential species distribution with suitable environmental variables, you can overlay the actual occurrence points with the modelled distribution on a map. This is useful to identify areas where a species is likely to occur, but just has not been found yet. Those areas where a species is not occurring, but where the environment is suitable for it to survive can be used, for example, to prevent an invasive species from spreading to an area where it has not been observed yet.

It is important to realise that it is unlikely that the occurrence records that exist for a species reflect the entire range of environmental conditions that a species can occupy. The species distribution models are only built with the data available, and are thus only calibrated with the environmental niche that is represented by the occurrence records.The outcomes of these models should therefore always be viewed with caution and might not represent the full extent of the actual or potential distribution of a species. This is a general drawback of a modelling approach, as a model is always only a representation of the real world.


  • Guisan A & Zimmermann NE (2000) Predictive habitat distribution models in ecology. Ecological modelling, 135(2), 147-186.
  • Guisan A & Thuiller W (2005) Predicting species distribution: offering more than simple habitat models. Ecology letters, 8(9), 993-1009.
  • Pearson RG (2010) Species’ distribution modeling for conservation educators and practitioners. Lessons in conservation, 3, 54-89.
  • Pulliam HR (2000) On the relationship between niche and distribution. Ecology letters, 3(4), 349-361.

You can also view this information in Module 2 of our Online Open Course in SDM: https://app.bccvl.org.au/training


Lead partners: Griffith University, James Cook University

Thanks to: University of New South Wales, Macquarie University, University of Canberra

Funded by: National eResearch Tools and Resources Project (NeCTAR)

The BCCVL is supported by the National eResearch Tools and Resources Project (NeCTAR), an initiative of the Commonwealth being conducted as part of the Super Science Initiative and financed from the Education Investment Fund, Department of Education. The University of Melbourne is the lead agent for the delivery of the NeCTAR project and Griffith University is the sub-contractor.