Is local provenance important in habitat creation?


David M Wilkinson, Biological and Earth Sciences, Liverpool John Moores University, Byrom Street, Liverpool L3 3AF, UK (e-mail


  • 1Many habitat creation schemes specify that biological material of local provenance should be used in reintroductions. This has come to be the ‘text book’ approach. However, very little discussion of the theory underlying this idea has been published in the scientific literature. This paper aims to initiate this much-needed discussion.
  • 2A major reason for the use of local provenance is the claimed importance of conserving locally adapted genotypes, which are assumed to show high fitness. Using both genetic arguments and a consideration of Quaternary environmental change I argue that this reason will seldom be important.
  • 3I make tentative suggestions of when local provenance is likely to be important and when it can be given a low priority in habitat creation schemes.

During recent years there has been a growing interest in habitat creation as we attempt to recreate habitats that have been seriously damaged or even totally destroyed by human action. This often includes the reintroduction of species to the area being restored (here I concentrate on plant species but many of the arguments apply to other taxa). An increasingly important paradigm in such work is that it is often specified by authorities that seeds used should be of local provenance (Gilbert & Anderson 1998; Moore 2000). Indeed, I have been told of conservation planting schemes in Britain that have been abandoned due to a lack of ‘suitable’ local seeds. In this paper I challenge the idea that local provenance is an important consideration in many such schemes.

The concerns over non-local seed centre on what happens when the plants they produce interbreed with local members of the same species. One potential problem is that this will tend to lead to the destruction of interesting genetic information on the past history of a local population; for example, in Europe two species of oak, Quercus robur L. and Q. petraea (Mattuschka) Liebl., share two geographically distributed polymorphisms in their chloroplast RNA that have been related to the effects of isolation in different glacial refugia (Ferris et al. 1993). The information in such patterns could be confused by the introduction of non-local individuals.

The main concern is that hybrids between local and introduced individuals will be of lower fitness than individuals of ‘pure bred’ local provenance. There are two main mechanisms that could lead to this (Van Andel 1998; Keller, Kollmann & Edwards 2000). The most widely cited problem is that such hybrids may lose fitness due to the loss of local adaptation, as recently highlighted in the Journal of Applied Ecology by Keller, Kollmann & Edwards (2000). Another possibility is epistasis; that is, the disruption of co-adapted gene complexes. The effects of a gene at one locus on an individual’s fitness may depend on what alleles are present at another locus (Maynard Smith 1998), and the introduction of genes from a genetically distinct population may disrupt these beneficial gene complexes. I will consider each of these two mechanisms for reduced fitness in turn before making some tentative suggestions for when local provenance may be an important consideration.

Poor adaptation to local conditions

The idea that plants and other organisms are adapted to their local conditions seems so obvious that few people stop to ask what the local conditions are. Consider the following climatic date relating to different time scales.

101 years: the mean temperature for central England in August 1989 was 16·6 °C while in 1990 it was 18·0 °C (Cannell & Pitcain 1993).

102 years: during the last 130 years at Abisko in northern Sweden, the mean summer air temperature has varied between approximately 7 °C and 13 °C (Hofgaard 1999).

103 years: analysis of peat cores from Bolton Fell Moss in northern England covering the last 6000 years shows cyclic changes in bog surface wetness with a periodicity of ≈ 800 years (Barber et al. 1994).

104 years: the last glaciation ended only 10 000 years ago. As Colinvaux (1993) pointed out, this is only about 40 generations ago for many tree species.

Thinking about conservation problems in the context of Quaternary environmental change, as above, can provide an important time perspective on current conservation issues (Huntley 1991; Chambers, Mauquoy & Todd 1999; Wilkinson 1999, 2001).

In addition to temporal changes in climate, different microclimates (e.g. north- and south-facing slopes) can also create very different environmental conditions at the same site (Pope & Lloyd 1975; Rorison, Sutton & Hunt 1986).

Local conditions are not just determined by abiotic factors such as climate or topography, an important aspect is the identity of other species found in the community. Pollen analysis has shown that plant communities with no modern analogue have been extensive in Europe for much of the last 10 000 years, persisting as recently as 1000 years in many areas (Huntley 1990), which is only four generations of trees using Colinvaux’s (1993) criteria. When the extent of environmental variation is appreciated, then the idea of adaptation to very local conditions appears much less important, especially for long-lived plants such as trees. Yet in Britain great effort is often made to use tree seeds of local provenance in conservation schemes (Gilbert & Anderson 1998). A mature oak tree may have germinated during the ‘little ice age’ with annual temperatures 1·5 °C colder than those of the mid-twentieth century mean (Lamb 1982). In what sense are its acorns now locally adapted? It makes more sense to think of successful germination as a lottery with winning seeds being those lucky enough to have the correct genetic make up for that year; ecological fitting (sensuJanzen 1985) not local adaptation. This view would suggest using seeds from a variety of sources to increase the genetic diversity from which the environment can select (Van Andel 1998).


Epistasis is most likely to be a problem in situations where members of a population only receive genes from other members of the local population. In plants this is more likely to be found in insect-pollinated rather than wind-pollinated plants. For example, Chuine, Belmonte & Mignot (2000) have demonstrated a lack of local genetic variation in several species of wind-pollinated tree species. Evidence for an epistatic effect has recently been found in three weed species, Agrostemma githago, Papaver rhoeas and Silena latifolia (Keller, Kollmann & Edwards 2000), all of which are insect pollinated (Firbank 1988; Proctor, Yeo & Lack 1996). A recent study of local provenance in hawthorn Crataegus monogyna shows better performance by plants of local provenance (Jones, Hayes & Sackville Hamilton 2001). However it is unclear if this was due to a true local provenance effect or purely that the local plants were the only ones from a suitable upland site. Matching habitats may be more important than matching local seed sources that may come from dissimilar habitats.

The all-pervasive nature of environmental change (described in the previous section) probably means that epistasis is not a widely used method of adaptation to local conditions. This is likely to be particularly the case for long-lived plants. For example, it is hard to envisage how adaptation to local conditions by co-adapted genes could evolve in European trees that have been forced to migrate repeatedly in response to glacial/interglacial cycles over a time span of < 50 generations (Huntley & Webb 1989; Wilkinson 1999). Exceptions could be relatively short-lived plants that have a long history of isolation in an area of unusual environmental conditions (e.g. soils), such as the relict arctic alpine plants of Upper Teesdale in the northern Pennine Hills of England, which have survived at the site for in excess of 10 000 years (Turner et al. 1973). The genetic make up of such populations is also likely to be of historical interest, possibly the main reason to avoid the introduction of non-local plants.


My aim in this paper is primarily to raise questions that challenge the uncritical acceptance of the importance of local provenance. However, based on the above discussion, it is possible to make a number of tentative suggestions. First, local provenance is most likely to be unimportant for the fitness of long-lived individuals (this applies to both local adaptation and epistatic arguments). Epistasis is most likely to cause problems in the use of non-local seeds for isolated populations with very limited input of genes from other populations (in plants this suggests that local provenance should be of less concern in wind-pollinated plants). Natural selection will tend to remove any less-fit hybrids; one possible exception to this is when new genetic material is being constantly introduced, as in the repeated sowing of wildflower strips (Keller, Kollmann & Edwards 2000). In such cases, limiting the number of introduction events may be valuable.

The key message of this paper is that the ever-changing nature of the environment means that a strong emphasis on the importance of local provenance is often hard to justify. Many conservationists also fail to realize that geographical distance and genetic distance need not be correlated (Moore 2000) which can effectively make it impossible to know if local seed is genetically similar to the plants of a particular site unless molecular genetic evidence is available (which will seldom be the case in most conservation schemes). So, not only is local genetic provenance often hard to justify on theoretical grounds it may also be practically very difficult to achieve.